473

LIVING AND DWELLING UNDER THE ROMAN ROOF
IN THE EASTERN EMPIRE:
DOMESTIC ARCHITECTURE IN PROVINCIAL SYRIA
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF SOCIAL SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY

LIVING AND DWELLING UNDER THE ROMAN ROOF
IN THE EASTERN EMPIRE:

PLAGIARISM
I hereby declare that all information in this document has been obtained and presented in accordance with academic rules and ethical conduct. I also declare that, as required by these rules and conduct, I have fully cited and referenced all material and results that are not original to this work.

iv
ABSTRACT
LIVING AND DWELLING UNDER THE ROMAN ROOF
IN THE EASTERN EMPIRE:
DOMESTIC ARCHITECTURE IN PROVINCIAL SYRIA

This thesis examines the domestic architecture and the practices of daily life in the eastern extent of the Roman Empire. It documents the Roman period houses in their urban, social, cultural, and architectural contexts. The discussion is elaborated in reference to Antioch on the Orontes, Apamea on the Orontes, Zeugma, Palmyra, and Dura-Europos; all of which were in the Roman province of Syria. The study covers a time span between the Hellenistic period and the end of antiquity, roughly between 300 BC and 650 AD, and it illustrates the continuities, changes, and transformations in the design and use of the residential buildings throughout the Roman period with reference to the political, economic, and social dynamics of the region. Within this geographical and chronological perspective and through a comparative framework, it elaborates the common, disparate, and enduring features, site-specific issues, and cultural and architectural assimilations within the domestic contexts of the eastern Roman world. By doing this it also aims to place the local evidence into the wider imperial context and to provide new contextual insights and a more inclusive framework for the current debates on Roman domestic architecture.
Keywords: Roman Domestic Architecture, Domestic Space, Roman House, Roman East, Roman Syria.
v
ÖZ
DOĞU ROMA İMPARATORLUĞUNDA YAŞAM VE BARINMA:
SURİYE EYALETİNDE KONUT MİMARİSİ


Bu tez Roma İmparatorluğu’nun doğu bölgelerindeki günlük yaşam pratikleri ve konut mimarisini incelemeyi amaç edinir. Çalışmanın temel odağı kent konutları olup, antik Suriye eyaletine bağlı beş farklı şehire yoğunlaşır. Bu şehirler konut mimarisi için nitelik ve nicelik bakımından daha doyurucu bir arkeolojik veri sunan Asi’deki Antakya, Asi’deki Apamea, Zeugma, Palmyra ve Dura-Europos’dur. Bu kentlerden bugüne kadar elde edilmiş olan arkeolojik ve epigrafik veriler ışığında Roma dönemi evleri sosyal, kültürel ve mimari bağlamları ile birlikte belgelenir ve tartışılır. Tezin çalışma alanı Helenistik dönemden antik çağın bitimine kadar olan uzun bir zaman dilimini kapsar. Bu şekilde, Roma hakimiyeti boyunca, bölgenin politik, ekonomik ve sosyal dinamiklerine de bağlı olan konut mimarisindeki, devamlılık, değişim ve dönüşümler izlenir. Bu coğrafi ve kronolojik çerçeve doğrultusunda ve karşılaştırmalı bir bakış açısıyla doğu Roma kentlerindeki konutlarda tespit edilen ortak özellikler ve farklılıklar, bölgeye özgü olgular ile kültürel ve mimari alanlardaki özümlemeler tartışılır. Bu tartışmalar doğrultusunda, bu tez, bölgesel verileri daha geniş bir bağlama oturtmayı ve Roma konut mimarisi ile ilgili güncel tartışmalara yeni perspektifler ve daha geniş çerçeveler sunmayı hedefler.
Anahtar Kelimeler: Roma Konut Mimarisi, Konut Mekanları, Roma Evi, Doğu Roma, Suriye Eyaleti.

vii
ACKNOWLEDGMENTS
First and foremost, I am extremely grateful to my advisor Prof. Dr. Lale Özgenel for her academic guidance and insight throughout the research. This thesis has been completed thanks to her encouragement, friendship and always positive attitude. I am also indebted to Prof. Dr. Suna Güven and Assoc. Prof. Dr. Charles Gates for their valuable critics and encouragement from the very early stages of the study to the end. I also express my gratitude to the examining committee members Prof. Dr. Burcu Erciyas, Assoc. Prof. Dr. Inge Uytterhoeven and Assist. Prof. Dr. Pelin Yoncacı Arslan for their inspiring comments and suggestions. Thanks to the enthusiasm and sincerity of all the committee members the jury was a truly memorable occasion.
I am thankful to a number of people who helped me in accessing the resources used in the thesis. First of all, I am indebted to Prof. Dr. Catherine Saliou who was so kind to mail me an off-print copy of her article from Paris. I am also thankful to Nihal Uzun, the assistant librarian at BIAA, for sending me resources when the library was closed to visitors during the Covid-19 pandemic. I am also thankful to many friends who helped me to access resources in the libraries abroad.
I am extremely grateful to my husband Ayhan for his endless love and support, not only during the preparation of this thesis but in every moment of my life. He always trusts me more than I do and this thesis could not have been completed without his encouragement and support. I am also thankful to my dearest son Çınar for his patience, understanding and sensibility throughout this period. Last but not least, I am grateful to my parents and all the members of my large family who always make me feel so lucky to have them.
viii
TABLE OF CONTENTS
PLAGIARISM ............................................................................................................ iii
ABSTRACT ................................................................................................................ iv
ÖZ ................................................................................................................................. v
DEDICATION ............................................................................................................ vi
ACKNOWLEDGMENTS .......................................................................................... vii
TABLE OF CONTENTS .......................................................................................... viii
LIST OF FIGURES ................................................................................................... xiii
CHAPTERS
1. INTRODUCTION .................................................................................................... 1
2. THE ROMAN EAST: GEOGRAPHICAL, POLITICAL AND SOCIAL CONTEXTS .......................................................................................................... 10
2.1 Antioch on the Orontes: The Provincial Capital of Syria ................................ 10
2.1.1 The Population and Community of Antioch on the Orontes ..................... 15
2.2 Apamea on the Orontes: The Provincial Capital of Syria Secunda ................. 16
2.2.1 The Population and Community of Apamea on the Orontes .................... 19
2.3 Zeugma: A Bridge on the Euphrates ................................................................ 20
2.3.1 The Population and Community of Zeugma ............................................. 25
2.4 Palmyra: A Commercial Oasis in the Desert ................................................... 26 2.4.1 The Population and Community of Palmyra ............................................. 30
2.5 Dura-Europos: A Fortress on the Euphrates .................................................... 33
2.5.1 The Population and Community of Dura-Europos ................................... 36
3. ANTIOCH ON THE ORONTES: LEISURE AND LUXURY IN THE EASTERN CAPITAL .............................................................................................................. 39
3.1 Urban Layout and Development of a Provincial Capital ................................. 39
3.2 Domestic Architecture in Antioch on the Orontes ........................................... 45
3.2.1 The Atrium House (The Roman Villa) ..................................................... 48
3.2.2 The House of the Calendar and the House of Cilicia ................................ 51
ix
3.2.3 The House of the Drinking Contest .......................................................... 53
3.2.4 The House of Dionysus and Ariadne ........................................................ 55
3.2.5 The House of the Buffet Supper and The House with Floral Pavements (Building B) ....................................................................................................... 57
3.2.6 The House of the Boat of Psyches ............................................................ 63
3.2.7 The House of Menander ............................................................................ 66
3.2.8 The Constantinian Villa and The House of Ge and the Seasons............... 71
3.2.9 The Yakto Complex (The Villa of Ardabur) ............................................ 73
3.3 Overview of Domestic Architecture in Antioch on the Orontes ...................... 78
4. APAMEA ON THE ORONTES: DOMESTIC SPLENDOR IN LATE ANTIQUITY ......................................................................................................... 86
4.1 Urban Layout and Development of a Later Provincial Capital........................ 86
4.2 Domestic Architecture in Apamea on the Orontes .......................................... 90
4.2.1 The Building with Triclinos (L’Édifice “Au Triclinos”) .......................... 92
4.2.2 The House of the Consoles (La Maison aux Consoles) ............................ 98
4.2.3 The House of the Pilasters (La Maison aux Pilastres) ............................ 104
4.2.4 The House of the Console Capitals (La Maison des Chapiteaux A` Consoles) .......................................................................................................... 109
4.2.5 The House of the Deer (La Maison du Cerf) .......................................... 113
4.2.6 The House with the Bilobed Columns and The House with the Trilobed Columns (La Maison aux Colonnes Bilobées & La Maison aux Colonnes Trilobées) ......................................................................................................... 117
4.3 Overview of Domestic Architecture in Apamea on the Orontes ................... 121
5. ZEUGMA: TRANSFORMATION IN BETWEEN ............................................ 128
5.1 Urban Layout and Development of the Twin Towns on the Euphrates ......... 128
5.1.1 Urban Layout and Development of Apamea on the Euphrates............... 128
5.1.2 Urban Layout and Development of Seleucia on the Euphrates/Zeugma 130
5.2 Domestic Architecture in Zeugma ................................................................. 136
5.2.1 The House of the Fountain ...................................................................... 139
5.2.2 The House of the Helmets ....................................................................... 142
5.2.3 The House of the Bull ............................................................................. 144
5.2.4 The House of Dionysus and Ariadne ...................................................... 147
5.2.5 The House of Poseidon ........................................................................... 151
5.2.5.1 Part A of the House of Poseidon ....................................................... 152
x
5.2.5.2 Part B of the House of Poseidon ....................................................... 158
5.2.6 The House of Euphrates .......................................................................... 164
5.2.7 The House without Mosaics .................................................................... 170
5.2.8 The House of Synaristosai (The House of Zosimos) .............................. 172
5.2.9 The Late Imperial Peristyle House .......................................................... 179
5.3 Overview of Domestic Architecture in Zeugma ............................................ 181
6. PALMYRA: INHABITING THE DESERT ........................................................ 186
6.1 Urban Layout and Development of a Desert City .......................................... 186
6.2 Domestic Architecture in Palmyra ................................................................. 191
6.2.1 Domestic Structures in the Hellenistic Quarter ....................................... 192
6.2.2 Roman Residential Buildings in the Western Urban Quarter ................. 193
6.2.2.1 Houses 38, 39 and 45 ........................................................................ 193
6.2.2.2 House F ............................................................................................. 194
6.2.2.3 The Peristyle House .......................................................................... 199
6.2.3 Residential Buildings in the Eastern Urban Quarter ............................... 201
6.2.3.1. The House(s) of Achilles and Cassiopeia ........................................ 201
6.2.4 Residential Buildings within the Monumental Centre of the City .......... 204
6.2.4.1 The House to the South-east of the Theatre ...................................... 204
6.2.4.2 The Houses to the East of the Temple of Nabu................................. 208
6.3 Overview of Domestic Architecture in Palmyra ............................................ 209
7. DURA-EUROPOS: FROM SHOPS TO PALACES ........................................... 213
7.1 Urban Layout and Development of a Fortress City ....................................... 213
7.2 Domestic Architecture in Dura-Europos ........................................................ 218
7.2.1 The Citadel Palace ................................................................................... 222
7.2.2 The Redoubt Palace/The Strategeion ...................................................... 225
7.2.3 Insula C7 ................................................................................................. 229
7.2.3.1 House C7A ........................................................................................ 230
7.2.3.2 House C7A2 ....................................................................................... 232
7.2.3.3 House C7B ........................................................................................ 232
7.2.3.4 House C7B2 ....................................................................................... 233
7.2.3.5 House C7C ........................................................................................ 233
7.2.3.6 House C7C2 ....................................................................................... 234
xi
7.2.3.7 House C7D ........................................................................................ 235
7.2.3.8 House C7E ........................................................................................ 235
7.2.3.9 House C7F/The House of the Frescoes ............................................. 236
7.2.3.10 House C7G ...................................................................................... 238
7.2.3.11 House C7G2 ..................................................................................... 239
7.2.3.12 House C7G3 ..................................................................................... 239
7.2.4 Agora/Sector G........................................................................................ 239
7.2.4.1 Insula G1 ........................................................................................... 241
7.2.4.2 Insula G3 ........................................................................................... 245
7.2.4.2.3 House G3C ..................................................................................... 249
7.2.5 Insula B2 ................................................................................................. 250
7.2.5.1 House B2C ........................................................................................ 252
7.2.5.2 House B2A ........................................................................................ 253
7.2.6 The House of Lysias (Insula D1) ............................................................ 255
7.2.7 The House of the Large Atrium/The House of the Cistern ..................... 260
7.2.8 House E4/The House of the Parthian ...................................................... 265
7.2.9 The Christian House/ The House of the Christian Church ..................... 269
7.2.10 The Roman Palace/The Palace of Dux Ripae ....................................... 273
7.3 Overview of Domestic Architecture in Dura-Europos................................... 280
7.3.1 Textual Evidence: PDura 19 ................................................................... 281
7.3.2 Overview of the Archaeological Remains .............................................. 283
8.CONCLUSION: LIVING AND DWELLING AS ‘ROMAN’ IN THE EASTERN PROVINCE OF SYRIA ...................................................................................... 291
8.1 Living as a ‘Roman’ in the Eastern Province of Syria ................................... 291
8.2 Dwelling as ‘Roman’ in the Eastern Province of Syria ................................. 293
8.2.1 Residential Buildings as Part of the Urban Fabric .................................. 294
8.2.2 Architectural Layout and Vocabulary: Form, Function, and Meaning in the Domestic Space ................................................................................................ 299
BIBLIOGRAPHY .................................................................................................... 320
APPENDICES
A. FIGURES ............................................................................................................ 363
B. CURRICULUM VITAE ..................................................................................... 517
xii
C. TURKISH SUMMARY/TÜRKÇE ÖZET .......................................................... 519
D. THESIS PERMISSION FORM /TEZ İZİN FORMU ......................................... 548
xiii
LIST OF FIGURES
Fig. 1 Map of the Roman East at its greatest extent. Five cities studied in the thesis are marked. (Edwell, 2008: 21) ............................................................... 363
Fig. 2 Geographical map showing the location of the five cities studied in the thesis (www. earth.google.com) ........................................................................ 364
Fig. 3 Aerial photo of modern Antakya taken in 1930s. The city lays between the Orontes River and Mount Silpius (Kenfield and Moss, 2014: fig. 1) ..... 364
Fig. 4 The Yakto Complex, The Megalopsychia mosaic Dunbabin (1999: fig 194) ......................................................................................... 365
Fig. 5 The Yakto Complex, Details from the topographical border of the Megalopsychia mosaic (Photos taken by the author in the Hatay Archaeology Museum) ............................................................................ 365
Fig. 6 Antioch on the Orontes, Urban Plan (Becker and Kondoleon, 2005: xv) .... 366
Fig. 7 Antioch on the Orontes, Remains of the walls and towers in 1977, Engraving by L. F. Cassas, (Foss, 2000: 22) ............................................................ 367
Fig. 8 Antioch on the Orontes, Medina Gate in 1977, Engraving by L. F. Cassas, (Downey, 1961: 768, fig.21) ................................................................... 367
Fig. 9 Antioch on the Orontes, Urban scenes (Lassus, 1935: Plates XII-XVIII) .... 368
Fig. 10 Daphne, Excavation photo of the House of the Buffet Supper in 1937 (http://vrc.princeton.edu/archives/items/show/15449) ............................ 369
Fig. 11 Daphne, Plan of the excavated residential sector (Levi II, 1947: Plan III). 369
Fig. 12 The Atrium House, Plan (after Fisher, 1934: Plate IV) .............................. 370
Fig. 13 The Atrium House, Computer generated reconstruction photograph of the mosaic pavement in room 82 (Becker and Kondoleon, 2005: 18, fig. 2) 371
Fig. 14 The House of the Calendar (highlighted section), Plan of the excavated part (The attached building to the north-east is the House of Drunken Dionysus) (http://vrc.princeton.edu/archives/items/show/14645) ........... 371
Fig. 15 The House of the Calendar, Excavation photo of room 2 and the nymphaeum beyond (http://vrc.princeton.edu/archives/items/show/14109) ............... 372
Fig. 16 The House of the Calendar, Mosaic pavement of room 2 (Photo taken by the author in the Hatay Archaeology Museum) ............ 372
Fig. 17 The House of Cilicia, Plan of the excavated part (http://vrc.princeton.edu/archives/items/show/15781) ............................ 373
Fig. 18 The House of Cilicia, Excavation photo (http://vrc.princeton.edu/archives/items/show/15613) ............................ 373
xiv
Fig. 19 The House of the Drinking Contest, Plan (Gruber and Dobbins, 2010: fig. 1) ......................................................... 374
Fig. 20 The House of the Drinking Contest, Axonometric view (Gruber and Dobbins, 2010: fig. 3) ......................................................... 374
Fig. 21 The House of the Drinking Contest, Mosaic pavement of triclinium A (https://artmuseum.princeton.edu/collections/objects/29551) ................. 375
Fig. 22 The House of the Drinking Contest, Mosaic pavement of portico B (Photo taken by the author in the Hatay Archaeology Museum) ............ 375
Fig. 23 The House of the Drinking Contest, Excavation photo with Mountain Casius in the background (http://vrc.princeton.edu/archives/items/show/17412) ............................ 375
Fig. 24 The House of Dionysus and Ariadne, Restored Plan (Levi I, 1947: fig. 56) .............................................................................. 376
Fig. 25 The House of Dionysus and Ariadne, Excavation photo towards the harbor (Levi I, 1947: fig. 57) .............................................................................. 376
Fig. 26 The House of Dionysus and Ariadne, Mosaic pavement of the reception room (Photo taken by the author in the Hatay Archaeology Museum) ... 377
Fig. 27 The House of Dionysus and Ariadne, Mosaic pavement of the portico (Photo taken by the author in the Hatay Archaeology Museum) ............ 377
Fig. 28 The House of Dionysus and Ariadne, Excavation photo from triclinium towards the northern room (http://vrc.princeton.edu/archives/items/show/15764) ............................ 377
Fig. 29 The House of the Buffet Supper, Plan of the 3rd c. AD phase (After Stillwell, 1941: figs.32) ................................................................. 378
Fig. 30 The House of the Buffet Supper, Excavation photo of rooms C1, C2, C3 and the portico in front (http://vrc.princeton.edu/archives/items/show/16135) ............................ 379
Fig. 31 The House of the Buffet Supper, Mosaic pavement of room C2 (Photo taken by the author in the Hatay Archaeology Museum) ............ 379
Fig. 32 The House of the Buffet Supper, Mosaic pavement of room C3 (Photo taken by the author in the Hatay Archaeology Museum) ............ 379
Fig. 33 The House of the Buffet Supper, Mosaic pavement of the portico (Photo taken by the author in the Hatay Archaeology Museum) ............ 380
Fig. 34 C300 Complex, Excavation photo (http://vrc.princeton.edu/archives/items/show/16346) ............................ 380
Fig. 35 The House of the Buffet Supper, Plan of the 5th c. AD phase (After Stillwell, 1941: figs.33) ........................................................................................... 381
Fig. 36 Narrow streets between the house and the entrance doors in modern Antakya (Photos taken by the author) .................................................................... 382
xv
Fig. 37 The House with Floral Pavements, Opus sectile floor pavement of room B13 (http://vrc.princeton.edu/archives/items/show/15499) ............................ 382
Fig. 38 The House with Floral Pavements, Plan of the late 5th c. AD phase (http://vrc.princeton.edu/archives/items/show/15783) ............................ 383
Fig. 39 The House of the Boat of Psyches, Restored plan by Levi (1947: fig. 63) 384
Fig. 40 The House of the Boat of Psyches, Restored plan by Stillwell (1961: fig. 16) .......................................................................................... 384
Fig. 41 The House of the Boat of Psyches, Plan by Kondoleon (2000: fig.5) ........ 385
Fig. 42 The House of the Boat of Psyches, Mosaic pavement of room 1 (Levi II, 1947: Plate XXXVa) ................................................................. 385
Fig. 43 The House of the Boat of Psyches, Mosaic pavement of room 3 (Photo taken by the author in the Hatay Archaeology Museum) ............ 386
Fig. 44 The House of the Boat of Psyches, Mosaic pavement of corridor 4 (Photo taken by the author in the Hatay Archaeology Museum) ............ 386
Fig. 45 The House of the Boat of Psyches, Mosaic pavement of room 6 (http://vrc.princeton.edu/archives/items/show/13638) ............................ 386
Fig. 46 The House of Menander, Plan (Stillwell, 1941: plan VII) ......................... 387
Fig. 47 The House of Menander, Colored plan with five different sections (Dobbins, 2000: 50) ................................................................................. 388 Fig. 48 The House of Menander, Excavation photo of the southeast suite with court 1 and room 3 in the foreground (http://vrc.princeton.edu/archives/items/show/16987) ............................ 388
Fig. 49 The House of Menander, Excavation photo of courtyard 1 with two layers of mosaic pavement (Levi I, 1947: fig.27) .................................................. 389 Fig. 50 The House of Menander, Excavation photo of the late pool with rooms 11, 12 and 2 behind (http://vrc.princeton.edu/archives/items/show/16982) . 389 Fig. 51 The House of Menander, Court 17 (Levi II, 1947: Plate XLVIIc) ............. 390
Fig. 52 The Constantinian Villa, Plan of the excavated part (Levi, 1947: fig.85) .. 391
Fig. 53 The Constantinian Villa, Excavation photo (Levi I, 1947: fig. 86) ............ 391
Fig. 54 The Constantinian Villa, Mosaic Pavement of Room 1 (Photo taken by the author in Louvre Museum) ...................................... 392
Fig. 55 The House of Ge and the Seasons, Plan of the excavated part (Levi 1, 1947: fig. 139) ............................................................................ 392 Fig. 56 The House of Ge and the Seasons, Plan of room 1 in the earlier phase (http://vrc.princeton.edu/archives/items/show/14648) ............................ 392
Fig. 57 The Yakto Complex, Plans of the 3rd c. AD and 5th c. AD phases (Lassus: 1938 figs. 4 and 5) ..................................................................... 393
Fig. 58 The Yakto Complex, Plan of the 5th c. AD phase (Lassus, 1938: Plan IX) 394
xvi
Fig. 59 The Yakto Complex, Mosaic pavement of corridor 5 (Lassus, 1938: fig. 8) ............................................................................... 395
Fig. 60 The Yakto Complex, Central panel of opus sectile pavement of room 52 (Lassus, 1938: fig. 32) ............................................................................. 395
Fig. 61 The Yakto Complex, Excavation photo of room 10 with nymphaeum at the back (Lassus, 1938: fig. 39) ..................................................................... 395
Fig. 62 The Yakto Complex, Mosaic pavement of Room A (Photo taken by the author in the Hatay Archaeology Museum) ............ 396
Fig. 63 The Yakto Complex, Mosaic pavement of Room B (Photo taken by the author in the Hatay Archaeology Museum) ............ 396
Fig. 64 Parthian Palace Plans .................................................................................. 397
Fig. 65 Apamea on the Orontes, The google earth view of the site (www.earth.google.com) ......................................................................... 398
Fig. 66 Apamea on the Orontes, The Acropolis Hill (Modern Qal’at al Mudiq) (https://tr.m.wikipedia.org/wiki/Dosya:Qalat_el-Mudiq.jpg).................. 398
Fig. 67 Apamea on the Orontes, Urban Plan (The location of the excavated houses are highlighted) (After Balty, J., 1984: fig.3) .......................................... 399
Fig. 68 Apamea on the Orontes, The Colonnaded Street (https://www.stevensklifas.com/apamea-syria-hellenistic-ancient-city) . 400
Fig. 69 The Grand Colonnade of Apamea on the Orontes, Reconstruction in Cinquantenaire Museum in Brussels, Belgium (https://commons.wikimedia.org/wiki/File:Great_Colonnade_at_Apamea,_Syria,_reconstruction_-_Cinquantenaire_Museum_-_Brussels,_Belgium_-_DSC09011.jpg) ...................................................................................... 400
Fig. 70 The Building with Triclinos, Plan (After Balty, J. Ch., 1984: fig. 2) ........ 401
Fig. 71 The Building with Triclinos, Capital of the pilaster in the north gallery of the peristyle (Balty, 1969: Plate XLI.1) ........................................................ 402
Fig. 72 The Building with Triclinos, Marble wall decoration of room Q (Balty, 1969: Plate XLIV.2) .................................................................... 402
Fig. 73 The Building with Triclinos, Schematic drawing of the mosaic pavement in room AB (Balty, 1969: 110, fig.3) .......................................................... 402
Fig. 74 Scene from the mosaic pavement of room AB (Balty, 1969: Plate XLII.1) ...................................................................... 402
Fig. 75 The Building with Triclinos, Amazon mosaic in room T (Dulière, 1968: Plate I) ............................................................................ 403
Fig. 76 The Building with Triclinos, Staircase in room M (Balty, 1969: Plate XXXIX) .................................................................... 403
xvii
Fig. 77 The Building with Triclinos, Hunting mosaic in room (https://commons.wikimedia.org/wiki/File:Musée_Cinquantenaire_Mosaïque_de_la_ Chasse_01.jpg) ....................................................................... 404
Fig. 78 The Building with Triclinos, Inscription on the threshold of room A (https://spacedmm.com/pics/437/16-10%20226.JPG) ............................ 404
Fig. 79 Insula plan with three adjoining houses (After Balty, J. Ch., 1984: fig. 1) 405
Fig. 80 The House of the Consoles, Plan showing the construction phases (Balty, J., 1984: fig. 1) ............................................................................. 405
Fig. 81 The House of the Consoles, Plan (After Balty, J. Ch., 1984: fig. 2) .......... 406
Fig. 82 The House of the Consoles, Western façade with the entrance door (http://apamee.org/fr/le-site-approche-topographique) ........................... 407
Fig. 83 The House of the Consoles, Details from the entrance door (https://www.romeartlover.it/Apamea2.html) ......................................... 407
Fig. 84 The House of the Consoles, Peristyle courtyard (http://monumentsofsyria.com/wpid899-2005-09-04-sl-20-apamea-house-of-consoles-jpg/) ...................................................................................... 408
Fig. 85 The House of the Consoles, Opus sectile floor pavement of Room A’ (Balty, J., 1984: Plate X.1-2) ................................................................... 408
Fig. 86 The House of the Consoles, Marble and painted plaster wall coverings of Room A-A’ (Balty, J., 1984: Plate XIV.1-2)........................................... 409
Fig. 87 The House of the Consoles, North gallery with the apse at its end (Balty, J., 1984: Plate III.2) ..................................................................... 409
Fig. 88 The House of the Pilasters, Plan (After Balty, J. Ch., 1984: fig. 2) ........... 410
Fig. 89 The House of the Pilasters, Plan showing construction phases (Gisler and Huwiler, 1984: fig.1) ............................................................ 410
Fig. 90 The House of the Pilasters, Reconstituted plan of the initial construction (Gisler and Huwiler, 1984: fig.2) ............................................................ 411
Fig. 91 The House of the Pilasters, Room A .......................................................... 411
Fig. 92 The House of the Pilasters, Column bases between rooms L and T (Gisler and Huwiler, 1984: Plate XXX.1) ............................................... 412
Fig. 93 The House of the Pilasters, Niches on the west wall of room T (Gisler and Huwiler, 1984: Plate XXX.2) ............................................... 412
Fig. 94 The House of the Console Capitals, Plan (After Balty, J. Ch., 1984: fig. 2) ............................................................. 413
Fig. 95 The House of the Console Capitals, Plan showing construction phases (Baratte, 1984: fig.1) ............................................................................... 414
Fig. 96 The House of the Console Capitals, Polylobed basin in the peristyle courtyard (After Baratte, 1984: Plate XXXIV.1) .................................... 414
xviii
Fig. 97 The House of the Console Capitals, A console capital found in the courtyard (Baratte, 1984: Plate XLV.2) ................................................................... 415
Fig. 98 The House of the Console Capitals, Late cistern in the peristyle courtyard (Baratte, 1984: Plate XLIII.1) .................................................................. 415
Fig. 99 The House of the Console Capitals, Later wall built with column shafts (Baratte, 1984: Plate XLI.2) .................................................................... 415
Fig. 100 The House of the Deer, Plan (Donnay-Rocmans and Donnay, 1984: fig.2) .......................................... 416
Fig. 101 The House of the Deer, Details of the column bases in the peristyle (Donnay-Rocmans and Donnay, 1984: figs.7, 8) .................................... 416
Fig. 102 The House of the Deer, Schematic drawing of the mosaic of room A (Donnay-Rocmans and Donnay, 1984: fig. 5) ......................................... 417
Fig. 103 The House of the Deer, Green marble top found in room A (Donnay-Rocmans and Donnay, 1984: Plate LII.2) ................................ 417
Fig. 104 The House of the Deer, Mosaic pavement of room C (Donnay-Rocmans and Donnay, 1984: Plate LIV.2) ............................... 418
Fig. 105 The House of the Deer, Mosaic pavement of room F (Donnay-Rocmans and Donnay, 1984: Plate LV.2) ................................ 418
Fig. 106 The House of the Deer, White marble table top found in room F (Donnay-Rocmans and Donnay, 1984: fig. 6 and Plate LII.1) ................ 418
Fig. 107 The House of the Deer, East wall of room A with windows opening to courtyard D (Donnay-Rocmans and Donnay, 1984: Plate LI.3) ............. 419
Fig. 108 The House of the Deer, Plan showing later constructions (Donnay-Rocmans and Donnay, 1984: fig. 1) ......................................... 419
Fig. 109 The House with the Bilobed Columns, Two excavation photos taken in 1934 and 1970 (Mayence, 1935: fig. 10& Balty, 1972: Plate II) ............ 420
Fig. 110 The House with the Trilobed Columns, Plan (Raepsaet-Charlier: 1984: fig. 1) ............................................................. 420
Fig. 111 The House with the Trilobed Columns, Mosaic pavement of the western portico (Raepsaet-Charlier: 1984: Plate LXI) .......................................... 421
Fig. 112 The House with the Trilobed Columns, Opus sectile pavements of rooms O&M (Raepsaet-Charlier: 1984: Plate LXIII) ......................................... 421
Fig. 113 Asia Minor, Late antique house plans with apsidal audience halls .......... 422
Fig. 114 Seleucia and Apamea on the Euphrates, Reconstruction model (Görkay, 2020a: fig. 3) ............................................................................ 423
Fig. 115 Apamea on the Euphrates, Urban plan (Abadie-Reynal, Ergeç and Bucak, 1999: fig. 17) ................................... 423
Fig. 116 Zeugma, Aerial photos taken in 2000 showing the territory of the ancient city (Başgelen, 2000: 5-6) ........................................................................ 424
xix
Fig. 117 Seleucia and Apamea on the Euphrates, Urban plan showing the borders and fortifications of the city in the Hellenistic and Roman periods (Görkay, 2020a: fig. 1) ............................................................................ 424
Fig. 118 Zeugma, Belkıs Tepe as seen from the modern road approaching the site (Photo taken by the author at the archaeological site) ............................. 425
Fig. 119 Roman period coin depicting Belkıs Tepe and the temple (Görkay, 2015: 32) .................................................................................. 425
Fig. 120 Zeugma, Urban plan (Locations of the excavated houses studied or mentioned in the thesis are highlighted) (After Aylward, 2013a: Plate 2) .............................................................. 426
Fig. 121 The House of the Fountain, Plan of the excavated part (Tobin, 2013: Plate 77) ............................................................................ 427
Fig. 122 The House of the Fountain, Mosaic panel of room 11D (Photo taken by the author in the Zeugma Mosaic Museum).................. 427
Fig. 123 The House of the Helmets and The House of the Bull, Plan of the excavated part (Tobin, 2013: Plate 10) ..................................................................... 428
Fig. 124 The House of the Helmets, Peristyle courtyard (Tobin, 2013: Plate 26-C) ........................................................................ 428
Fig. 125 The House of the Bull, Room 2M (Aylward, 2013: Plate 38-A) ............. 429
Fig. 126 The House of the Bull, Room 2K (Tobin, 2013: Plate 36-A) ................... 429
Fig. 127 The House of the Bull, Entrance of room 2L from room 2K (Tobin, 2013: Plate 34-C) ........................................................................ 429
Fig. 128 The House of Dionysus and Ariadne and the The House of Danae, Sheltering Structure (Photo taken by the author at the archaeological site of Zeugma) .............................................................................................. 430
Fig. 129 The House of Dionysus and Ariadne (on the left) and the The House of Danae (on the right), View towards south (Photo taken by the author at the archaeological site of Zeugma) .......... 430
Fig. 130 The House of Dionysus and Ariadne, View towards north overlooking the river (Photo taken by the author at the archaeological site of Zeugma) .. 430
Fig. 131 The House of Dionysus and Ariadne and The House of Danae, Plan of the excavated part (Önal, 2005: fig. 1) .......................................................... 431
Fig. 132 The House of Dionysus and Ariadne, Reconstruction plan (Photo taken by the author at the archaeological site of Zeugma) .......... 432
Fig. 133 The House of Dionysus and Ariadne, Peristyle courtyard (D3/13) (Photo taken by the author at the archaeological site of Zeugma) .......... 432
Fig. 134 The House of Dionysus and Ariadne, Three-dimensional reconstruction of the peristyle courtyard with the impluvium (Görkay, 2020b: fig. 1) ....... 433
xx
Fig. 135 The House of Dionysus and Ariadne, Room D2/16 and the entrance door of room D1/19 (Photo taken by the author at the archaeological site of Zeugma) .......... 433
Fig. 136 The House of Dionysus and Ariadne, Mosaic pavement of room D1/19 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 434
Fig. 137 The House of Dionysus and Ariadne, Mosaic Pavement of room D2/16 (Ergeç, 2000: 21) ..................................................................................... 434
Fig. 138 The House of Dionysus and Ariadne, Three-dimensional reconstruction of room D1/19 (Görkay, 2020b: fig. 6) ........................................................ 435
Fig. 139 The House of Dionysus and Ariadne, Rooms D11 and D12 (Photo taken by the author at the archaeological site of Zeugma) ..................................... 435
Fig. 140 The House of Dionysus and Ariadne, Peristyle courtyard and room 6 in the upper level (Photo taken by the author at the archaeological site of Zeugma) .......... 436
Fig. 141 The House of Dionysus and Ariadne, The two columns behind the southern portico of the courtyard and the cistern on the south wall (Photo taken by the author at the archaeological site of Zeugma) .......... 436
Fig. 142 The House of Dionysus and Ariadne, Reused column drums in the wall of room D20/5 (Photo taken by the author at the archaeological site of Zeugma) .......... 436
Fig. 143 Plan of the contiguous houses in Trench 8 (The House of Poseidon, The House of Euphrates and The House without Mosaics) (After Önal, 2013: Plan 2) ....................................................................... 437
Fig. 144 The House of Poseidon, Plan with mosaic pavements (Önal, 2013: plan 4) ................................................................................. 438
Fig. 145 The House of Poseidon, Reconstruction of peristyle A6/P9 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 438
Fig. 146 The House of Poseidon, Fountain in peristyle A6/P9 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 439
Fig. 147 The House of Poseidon, Three-dimensional reconstruction of the fountain (Barbet, 2005: Plate XVIII) ..................................................................... 439
Fig. 148 The House of Poseidon, Court A2/P4 with the impluvium (Önal, 2013: fig. 21) ................................................................................ 440
Fig. 149 The House of Poseidon, Three-dimensional reconstruction of court A2/P4 (Görkay, 2020b: fig. 4) ............................................................................ 440
Fig. 150 The House of Poseidon, Three-dimensional reconstruction of triclinium A1/P3 (Görkay, 2020a: fig. 9) ................................................................. 441
Fig. 151 The House of Poseidon, Mosaic pavement of triclinium A1/P3 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 441
xxi
Fig. 152 The House of Poseidon, Three-dimensional reconstruction of room A5/P5 (Görkay, 2020d: fig. 20) .......................................................................... 442
Fig. 153 The House of Poseidon , Mosaic pavement of room A5/P5 (Önal, 2013: 35, fig. 33) .......................................................................... 442
Fig. 154 The House of Poseidon, Wall paintings of room A11/P6 (Önal, 2013: fig. 38) ................................................................................ 443
Fig. 155 The House of Poseidon, Mosaic pavement of room A14/P38 (Photo taken by the author in the Zeugma Mosaic Museum).................. 443
Fig. 156 The House of Poseidon, Mosaic pavement of room A13/P37 (Photo taken by the author in the Zeugma Mosaic Museum).................. 444
Fig. 157 The House of Poseidon, Three-dimensional reconstruction of room A13/P37 (Görkay, 2020e: fig. 58) ........................................................... 444
Fig. 158 The House of Poseidon, Peristyle courtyard B6/P13 (Barbet, 2005: Plate. VII) ........................................................................ 445
Fig. 159 The House of Poseidon, Reconstruction of peristyle courtyard B6/P13 in the Zeugma Mosaic Museum (Photos taken by the author) .................... 445
Fig. 160 The House of Poseidon, Rooms B1/P10, B2/P11, B3/P12 and B4/P23 (Önal, 2013: fig. 53) ................................................................................ 446
Fig. 161 The House of Poseidon, Room B1/P10 towards south (Önal, 2013: fig. 92) ................................................................................ 446
Fig. 162 The House of Poseidon, Three-dimensional reconstruction of room B2/P11 (Görkay, 2020e, fig. 63) .......................................................................... 447
Fig. 163 The House of Poseidon, Mosaic pavement of room B2/P11 (Photo taken by the author in the Zeugma Mosaic Museum).................. 447
Fig. 164 The House of Poseidon, Wall paintings of room B2/P11 (Önal, 2013: fig. 88) ................................................................................ 447
Fig. 165 The House of Poseidon, View from room B2/P11 towards rooms B3/P12 and B4/P23 (Önal, 2013: fig. 87) ............................................................ 448
Fig. 166 The House of Poseidon, Doors opening to room B9/P27 from the courtyard (Önal, 2013: fig. 95) ................................................................................ 448
Fig. 167 The House of Poseidon, Statue of Mars as found in the store room (Önal, 2013: fig. 99) ................................................................................ 448
Fig. 168 The House of Poseidon, Latrine (Önal, 2013: fig. 105) ........................... 449
Fig. 169 The House of Euphrates, Plan with mosaics (Önal, 2013: 90) ................. 449
Fig. 170 The House of Euphrates, Corridor C3/P21 towards south (Önal, 2013: fig. 113) .............................................................................. 450
Fig. 171 The House of Euphrates, Drawing of graffiti on the wall of corridor C3/P21 (Önal, 2013: fig. 114) .............................................................................. 450
xxii
Fig. 172 The House of Euphrates, Mosaic pavement of triclinium C1/P19 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 450
Fig. 173 The House of Euphrates Impluvium court C2/P20 (Önal, 2013: 109, fig. 132) ...................................................................... 450
Fig. 174 The House of Euphrates, Three-dimensional reconstruction of impluvium court C2/P20 (Görkay, 2020c: fig. 5) ...................................................... 451
Fig. 175 The House of Euphrates, Paintings on the eastern wall of impluvium court C2/P20 (Önal, 2013: 109, fig. 133) ......................................................... 451
Fig. 176 The House of Euphrates, Mosaic pavement of room C8/P17 (Önal, 2013: fig. 122) .............................................................................. 451
Fig. 177 The House of Euphrates, Rooms C4/P16, P2/C15 and P15/C5 (the latter two provided passage between C2/P20 and C6/P22) (Önal, 2013: fig. 112) .............................................................................. 452
Fig. 178 The House of Euphrates, Reconstruction of peristyle courtyard C6/P22 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 452
Fig. 179 The House of Euphrates, View from room A11/P6 of the The House of Poseidon towards peristyle C6/P22, with rooms C13/P26 and C12/P25 in the background (Önal, 2013: 120, fig. 151) ............................................. 453
Fig. 180 The House of Euphrates, Wall paintings of room C13/P26 (Photo taken by the author in the Zeugma Mosaic Museum) .................. 453
Fig. 181 The House without Mosaics, Section of the eastern part of the building (Önal, 2013: fig. 33) ................................................................................ 454
Fig. 182 The House without Mosaics, Wall paintings on the southern wall of room D6/P30 (Önal, 2013: fig. 172) ................................................................. 454
Fig. 183 The House of Synaristosai, Schematic plan of the house(s) in the 2nd c. AD (Abadie-Reynal, 2012a: Plate II) ............................................................. 454
Fig. 184 Rural house dated between 2nd – 6th c. AD in Northern Syria (Tchalenko, 1953: Plate V) ...................................................................... 455
Fig. 185 The House of Synaristosai, Schematic plan in the first half of the 3rd c. AD (Abadie-Reynal, 2012a: Plate III) ............................................................ 456
Fig. 186 The House of Synaristosai, Plan showing walls of subsequent constructions (Abadie-Reynal, 2012a: Plate I) .............................................................. 456
Fig. 187 The House of Synaristosai, Partly hypothetical axonometric drawing (Abadie-Reynal, 2012a: pl. VII) .............................................................. 457
Fig. 188 The House of Synaristosai, Three-dimensional reconstruction towards south-east (Abadie-Reynal, 2012a: Plate VIIIa) ...................................... 457
Fig. 189 The House of Synaristosai, Excavation photo of room P4 (Abadie-Reynal, 2012a: fig. 15) .............................................................. 458
xxiii
Fig. 190 The House of Synaristosai, Three-dimensional reconstruction of room P4 towards south (Abadie-Reynal, 2012a: Plate IX) .................................... 458
Fig. 191 The House of Synaristosai, Excavation photo of the eastern part with room P1 in the foreground (Abadie-Reynal, 2012a: fig. 4) .............................. 459
Fig. 192 The House of Synaristosai, Excavation photo of Room P9 (Abadie-Reynal, 2012a: fig. 21) .............................................................. 459
Fig. 193 The House of Synaristosai, Three-dimensional reconstruction of room P2 towards south-east, (Abadie-Reynal, 2012a: Plate X) ............................. 460
Fig. 194 The House of Synaristosai, Mosaic pavement of room P13 (Photo taken by the author in the Zeugma Mosaic Museum).................. 460
Fig. 195 The House of Synaristosai, Excavation photo of rooms P8 and P13 (Abadie-Reynal, 2012a: fig. 36) .............................................................. 461
Fig. 196 The House of Synaristosai, Three-dimensional reconstruction of rooms P8, P13 and the upper floor (Abadie-Reynal, 2012a: Plate XI) .................... 461
Fig. 197 The House of Synaristosai, Three-dimensional reconstruction of rooms P13-P11 (Abadie-Reynal, 2012a: Plate XIIa) ......................................... 462
Fig. 198 The House of Synaristosai, Excavation photos of room P8 and the apse (Abadie-Reynal, 2012a: 51, figs. 35&39) ............................................... 462
Fig. 199 The House of Synaristosai, Excavation photo of the western part of the building (Abadie-Reynal, 2012a: fig. 56) ................................................ 463
Fig. 200 The House of Synaristosai, Three-dimensional reconstruction of court P3 (Abadie-Reynal, 2012a: Plate XIIb) ........................................................ 463
Fig. 201 The House of Synaristosai, Three dimensional reconstruction of court P17 (Abadie-Reynal, 2012a: Plate XIII)......................................................... 464
Fig. 202 The House of Synaristosai, Schematic plan of the later constructions (Abadie-Reynal, 2012a: Plate V)............................................................. 464
Fig. 203 The Late Imperial Peristyle House, Plan of Trench 7B with the house in 7B (Tobin, 2013: Plate 96) ............................................................................ 465
Fig. 204 The Late Imperial Peristyle House, Excavation photo of the courtyard towards southeast (Tobin, 2013: Plate 103A) ......................................... 465
Fig. 205 Palmyra, Urban plan showing different urban quarters (Zuchowska, 2011: 142) .......................................................................... 466
Fig. 206 Palmyra, Plan of the Hellenistic quarter, the structures marked with gray are a set of four houses with courtyards (Al As’ad, Schmidt-Colinet, 2000: fig 3) ................................................ 467
Fig. 207 Palmyra, Urban plan of the south-western quarter (Grassi, 2008:6) ........ 467
Fig. 208 Palmyra, Grand Colonnade with Tetrapylon at the back (https://www.pinterest.com/pin/292804413265457383) ......................... 468
xxiv
Fig. 209 Palmyra, Consoles on the columns that carried statues (https://www.mediastorehouse.com/worldinprint/ruins-colonnade-1224471.html) .......................................................................................... 468
Fig. 210 Palmyra, Temple of Bel (https://www.khanacademy.org/humanities/ancient-art-civilizations/palmyra/a/temple-of-bel-palmyra) ...................................... 468
Fig. 211 Palmyra, Temple of Baalshamin (https://www.khanacademy.org/humanities/ancient-art-civilizations/palmyra/a/temple-of-baalshamin) ....................................... 468
Fig. 212 Palmyra, Urban plan showing the location of the houses that are examined in the thesis (After Zuchowska, 2011: fig. 1) .......................................... 469
Fig. 213 House Plans by Gabriel (1926: figs. 3, 5& Plate XV) .............................. 470
Fig. 214 Plan of Houses 39, 40 and a third house (Krencker, 1932: Plate 19) ....... 470
Fig. 215 House F (Gawlikowski, 2007: figs. 6, 14, 13) .......................................... 471
Fig. 216 House F, Excavation photo (Gawlikowski, 2007: fig.12) ........................ 472
Fig. 217 House F, Plan of the latest phase (7th-9th c. AD) (Gawlikowski, 1996: fig.2) ...................................................................... 472 Fig. 218 The Peristyle House, Plan (Grassi, Rocca and Piacentini, 2015: fig. 2) ... 473 Fig. 219 The Peristyle House, Excavation photo (Grassi, Rocca and Piacentini, 2015:fig. 3) ............................................. 473
Fig. 220 The Peristyle House, Reused inscribed block and small altar (Grassi, Rocca and Piacentini, 2015: figs. 11& 18) ................................ 474
Fig. 221 The House(s) of Achilles and Cassiopeia, Plan (Gawlikowski, 2007: fig. 4) ..................................................................... 474
Fig. 222 The House(s) of Achilles and Cassiopeia, Plan (Delplace, 2013: 44) ...... 475
Fig. 223 The House(s) of Achilles and Cassiopeia, Peristyle courtyard 17 (Delplace, 2013: fig. 2) ............................................................................ 475
Fig. 224 The House to the Southeast of the Theatre, Plan (Gawlikowski, 2007: fig. 5) ..................................................................... 476
Fig. 225 The House to the Southeast of the Theatre, Aerial view of the remains of the house (in the foreground), the Theatre and the Colonnaded Street (in the background) (https://www.bbc.com/turkce/multimedya/2015/08/150824_ gallery_palmyra) ...................................................................................... 476
Fig. 226 The Houses to the East of the Temple of Nabu, Plan (Saliby, 1996: fig.7) ................................................................................. 477
Fig. 227 Dura-Europos, Aerial photo of the ruins of the ancient city (Archives from the French-Syrian Mission to Dura-Europos (MFSED) .......................... 478
xxv
Fig. 228 Dura-Europos, Isometric reconstruction of the Hellenistic city by Rostovtzeff (1938, 11-fig. 5) ................................................................... 478
Fig. 229 Dura-Europos, Isometric reconstruction of the Roman city by Perkins (1973: fig. 8) ............................................................................................ 478
Fig. 230 Dura-Europos, Urban plan (Baird, 2014: 9) ............................................. 479
Fig. 231 Dura-Europos, Hellenistic agora (Brown, 1944a: fig. 9) .......................... 480
Fig. 232 Dura-Europos, Agora in the 3rd c. AD (Brown, 1944a: fig. 78) ............... 480
Fig. 233 Dura-Europos, The military camp (Hopkins and Rowell, 1934: Plate III) .................................................... 480
Fig. 234 Dura-Europos, Urban plan The houses examined in the thesis are highlighted (After Baird, 2014: fig. 1.4) ................................................. 481
Fig. 235 The Citadel Palace, Aerial view of the remains in 1929 (Pillet, 1931: Plate XXIX) ....................................................................... 482
Fig. 236 The Citadel Palace, Plan (Pillet, 1931: Plate IV) ...................................... 482
Fig. 237 The Citadel Palace, Hypothetical reconstruction plan by F. Brown (Rostovtzeff, 1938: fig. 9) ....................................................................... 482
Fig. 238 Ai Khanoum, Hellenistic residential buildings ......................................... 483
Fig. 239 The Redoubt Palace, Plan (Pillet, 1933: Plate III) .................................... 484
Fig. 240 The Redoubt Palace/Strategeion, Three-dimensional reconstruction (Rostovtzeff, 1938: Plate V) .................................................................... 485
Fig. 241 The Redoubt Palace/Strategeion, Plan (Hoepfner and Schwander, 1986: fig. 216) ............................................ 485
Fig. 242 Insula C7, Plan (Baird, 2014: fig. 6.2) ...................................................... 486
Fig. 243 Insula C7, Housing Plots (Saliou, 2005: fig. 5) ........................................ 486
Fig. 244 Insula C7, Plan (After Baird, 2014: fig 6.2) ............................................. 487
Fig. 245 Insula C7, Plan of the latest phase of occupation (After Baird, 2014: fig 6.2) ...................................................................... 488
Fig. 246 House C7A, Courtyard with cistern/cesspool (A), cooler (B) and trough (C) (Baird, 2006: fig. 70) ............................................................................... 489
Fig. 247 House C7A, Cooler next to door from courtyard to room 7 (Baird, 2006: fig. 71) ............................................................................... 489
Fig. 248 House C7B2, The door between room 2 and room 4, the plaster cornice above the door (A) and bench on the floor (B). (Baird, 2006: fig. 78) ... 489
Fig. 249 House C7C, View of room 14 from the courtyard (A is the entrance of room 4, B and E are the column and central pier against the wall, C is the storage jar and D is the 'flag-pole base') .................................................. 490
xxvi
Fig. 250 House C7C, Graffiti from room 4 (Hopkins, 1934: Plates XXXIII&XXXIV) .............................................. 490
Fig. 251 House C7C2, Niches in the place of the blocked door of room 3 (Baird, 2006: fig. 87) ............................................................................... 490
Fig. 252 House C7F, Fresco on the wall of room 8 (Rostovtzeff, 1935: fig. 82) ... 491
Fig. 253 House C7G, View from court towards the entrance of room G4 (Baird, 2006: fig. 95) ............................................................................... 491
Fig. 254 Houses C7G2 and C7G3, Excavation photo from the west (Baird, 2006: fig. 100) ............................................................................. 491
Fig. 255 Insula G1, Plan (After Brown, 1944c: fig. 80) ......................................... 492
Fig. 256 Houses G1A and G1B, Excavation photo taken from the east (Brown, 1944c: Plate XII) ....................................................................... 492
Fig. 257 House G1A, Plan (Brown, 1944c: fig. 61) ................................................ 493
Fig. 258 House G1B, Plan (Brown, 1944c: fig. 65) ................................................ 493
Fig. 259 Insula G3, Plan (After Brown, 1944c: fig. 79) ......................................... 494
Fig. 260 Houses G3B-G3C-G3G, Early phase of occupation (Brown, 1944c: fig.30) ............................................................................ 495
Fig. 261 Houses G3B-G3C-G3G, Later phase of occupation (Brown, 1944c: fig.31) ............................................................................ 495
Fig. 262 Insula B2, Excavation photo from south-west (Allara, 2002: fig. 31) ...... 496
Fig. 263 Insula B2, Plan (Allara, 2002: fig. 35) ...................................................... 496
Fig. 264 House B2C, Plan (Allara, 2002: fig. 43) ................................................... 497
Fig. 265 House B2C, Excavation photo from north-west (Allara, 2002: fig. 44) ... 497
Fig. 266 House B2A, Plan (Allara, 2002: fig. 96) .................................................. 498
Fig. 267 House B2A, Excavation photo from north-west (Allara, 2002: fig. 97) ... 498
Fig. 268 The House of Lysias (Insula D1), Plan (Baird, 2014: fig. 6.6) ................. 499
Fig. 269 The House of Lysias, North entrance. (Baird, 2006: fig. 410) ................. 500
Fig. 270 The House of Lysias, Plan with different sections of the house (Pontbriand, 2012: 84-fig. 10) ................................................................. 500
Fig. 271 The House of Lysias, Plan with different parts of section A (Pontbriand, 2012: 84-fig. 11) ................................................................. 501
Fig. 272 The House of Lysias, View of Courtyard 1 from north-east (Baird, 2014: fig.6.7) ............................................................................... 501
Fig. 273 The House of Lysias, View of Courtyard 1 towards south (Baird, 2014: fig.6.9) ............................................................................... 502
Fig. 274 The House of Lysias, Fanciful reconstruction of south side of the courtyard 1 by Herbert Gute (Baird, 2014: fig. 6.12) .............................................. 502
xxvii
Fig. 275 The House of Lysias, Stables (Pontbriand, 2016: 123-fig. 6) .................. 502
Fig. 276 The House of Lysias, Sections (Pontbriand, 2016: figs. 7-8) ................... 503
Fig. 277 The House of Lysias, Axonometric reconstruction (Pontbriand, 2016: fig. 10) ...................................................................... 503
Fig. 278 The House of the Large Atrium/Insula D5, Plan showing housing units in early phases of occupation (After Baird, 2014: fig: 6.3) ......................... 504
Fig. 279 The House of the Large Atrium/Insula D5, Plan showing housing units in latest phase of occupation (After Baird, 2014: fig: 6.3) .......................... 505
Fig. 280 The House of the Large Atrium/Insula D5, Staircase F11 (Baird, 2006: fig. 184) ............................................................................. 506
Fig. 281 The House of the Large Atrium/Insula D5, Central door from court F1 to room F4 (Baird, 2006: fig.186) ............................................................... 506
Fig. 282 The House of the Large Atrium/Insula D5, Niche on the south wall of room F4 (Baird, 2006: fig. 182) ........................................................................ 506
Fig. 283 The House of the Large Atrium/Insula D5, Terracotta tiles (Pillet: 1933, Plate VI-3, 4) ..................................................................... 507
Fig. 284 The House of the Large Atrium/Insula D5, Friezes from room F4 (Baird, 2006: figs. 135, 136, 139, 140).................................................... 507
Fig. 285 The House of the Large Atrium/Insula D5, Courtyard D1 (Baird, 2006: fig. 162) ............................................................................. 507
Fig. 286 House E4, Plan (Baird, 2014: fig. 3.12) ................................................... 508
Fig. 287 House E4, North-east corner of courtyard 14 (Baird, 2006: fig. 199) ...... 508
Fig. 288 Insula M8, Plan (Kraeling, 1967: plan II) ................................................. 509
Fig. 289 The Christian House, Plan (Baird, 2006: 500-fig. 392) ............................ 509
Fig. 290 The Christian House, Graffiti on the south wall of room 4B (Goldman, 1999, figs. 13b&14b) ............................................................ 510
Fig. 291 The Christian House, View of rooms 4B and 4A with windows on the back wall (Güney, 2012: fig. 37) ..................................................................... 510
Fig. 292 The Christian House, Plan of the building after its conversion into a church (Kraeling, 1967: fig. 1) ............................................................................ 511
Fig. 293 The Christian House, Isometric Drawing of the building after its conversion into a church (Kraeling, 1967: plan III) ................................................... 511
Fig. 294 The Roman Palace/The Palace of Dux Ripae, Plan (Detweiler, 1952: fig. 7) .......................................................................... 512
Fig. 295 The Roman Palace, Longitudinal Sections (Detweiler., 1952: fig. 8) ...... 512
Fig. 296 The Roman Palace, Apse of room 2 during excavation (Detweiler, 1952: Plate III. 1) .................................................................. 513
xxviii
Fig. 297 Roman Palace/Palace of Dux Ripae, Painted Ceiling Patterns of Rooms 3, 6, 19 and 61 on the left and of rooms 12 and 13 on the right (Detweiler, 1952: Plates VIIII& IX) ........................................................................... 513
Fig. 298 The Roman Palace, Apse 21 during excavation (Detweiler, 1952: Plate V. 2) ................................................................... 513
Fig. 299 The Roman Palace, Isometric reconstitution (Detweiler, 1952: fig. 2) .... 514
Fig. 300 The Roman Palace, Reconstructed Plan of the northern section (Downey, 1993: fig. 5) ............................................................................. 514
Fig. 301 Dura-Europos, Isometric reconstruction of an insula (Hoepfner and Scwander, 1986: fig. 222) ............................................... 515
Fig. 302 Dura Europos, Geophysical map of insula M3 (Benech 2010, fig. 6a) .... 515
Fig. 303 Dura Europos, Remains and reconstructions of roof construction (Baird, 2006: figs. 48, 49, 50, 51) ............................................................ 516
Fig. 304 House M7W, Panels of wall paintings with dining scenes (Baird, 2006: figs. 381& 382) .................................................................. 516
1
CHAPTER 1
INTRODUCTION
This thesis examines the domestic architecture and daily life practices in the eastern extent of the Roman Empire. It documents the Roman period houses in their social, cultural, and architectural contexts concerning the political, economic, and social dynamics of the geography and the period. The scope of the study is limited to the urban context and focuses on a sample of five ancient cities of the eastern Roman province of Syria -Antioch on the Orontes, Apamea on the Orontes, Zeugma, Palmyra, and Dura-Europos -that present a more remarkable and promising archaeological sample in terms of both quantity and quality (figs. 1, 2). Covering a time span between the Hellenistic period and the end of antiquity -roughly the period between 300 BC and 650 AD- it observes the continuities, changes, and transformations in the design and use of the residential buildings throughout the Roman period. The archaeological evidence is essentially examined through a reading of architecture and the function of the spaces are interpreted according to their size, location, decoration, entrance patterns, circulation and communication schemes as well as to the finds and in-situ features like fountains, basins, niches and alike.1 With this geographical, chronological and methodological perspective and through a comparative framework, the thesis puts the common, disparate, and enduring features, site-specific issues, and cultural and architectural assimilations in the domestic contexts of the eastern Roman world, into a discussion.
Roman domestic architecture and art found a place in the historiography of Roman studies from the 18th century onwards, thanks to the rediscovery of the ancient cities of Pompeii, Herculaneum, and slightly later, of Ostia. The earlier studies on the subject
1 This thesis, written in an architectural history program in the department of architecture, examines the archaeological data in an architectural framework.
2
mainly concentrated on the documentation, description, and classification of the extensive and elaborate housing sample provided by the excavations conducted in the three cities. Much of the effort in the earlier scholarship was indeed allocated to typological studies that were based on the articulation of spaces around the atrium, a space which was accepted as the hallmark of the ‘Roman House’. Accordingly, the houses with atrium was considered as an indicator of a distinctive, constant and universal Roman identity, the representational elements of which were employed and constructed by the socially elite and/or wealthy families. In the late 20th century, the accumulation of new archaeological evidence from both Italy and the provinces, the study of architectural and decorative elements and artefacts in a contextual approach and their multidimensional interpretation challenged the admitted function of the atrium itself and its role in the definition and study of Roman houses.2 The renaissance seen in social sciences and the new insights and theories put forward in such disciplines such as sociology, archaeology, art and architectural history as well as the interdisciplinary approaches widened the scholarly horizons and developed new theoretical frameworks concerning the studies on Roman domestic architecture.3 The flourishing engagements with the social theories put Roman private life and family on stage and increased the attention to the ‘Roman domus’, as a social unit, instead of the
2 The novel perspective of Andrew Wallace-Hadrill was epoch-making in that manner. By throwing light to the excavated houses in Cosa and to the stratigraphic evidence coming from Pompeii, he argued that the essence of the Roman house might be the gathering of rooms around an open or covered central court, rather than the presence of an impluviate atrium (1997: 223). Andrew Wallace-Hadrill and John Clarke are also pioneering figures who attempted to assess architecture and decor as a coding of social iconography and paved the way for further discussions on the ‘public’ and ‘private’ aspects of the Roman house (Clarke, 1991 and 2003; Thébert, 1993; Wallace-Hadrill, 1994; Grahame, 1997; Cooper, 1997; Tuori and Nissin, 2015). Recent research on cubiculum provided useful insights to these discussions (Riggsby, 1997; Nissinen, 2009; Anguissola, 2013; Nissin, 2015). Penelope Allison and Joanne Berry, also made critical assessments on the use of Roman house by looking into material evidence; they studied the distribution of household artefacts in Pompeian houses to question the room functions and discussed the social conditions of domestic life in Pompeii (Allison, 1993, 2001, 2004 and Berry, 1997).
3 For a discussion and interdisciplinary approach to the archaeology of domestic architecture see Kent, 1993.
3
‘Roman house’ as a static physical entity.4 Discussions that are centered on such concepts as ethnicity, identity, and gender brought a multidimensional investigation and understanding of Roman houses and households.5 The increase of attention to Later Roman periods and the recent excavations conducted in Roman provinces particularly in Britain, North Africa, Syria, Spain and Asia Minor uncovered a considerable number of houses that provided new chronological, geographical and contextual perspectives. The recent scholarship, with the ever-increasing archaeological evidence profoundly challenged the problematic idea of the ‘Roman House’ as a static and formally definable entity and illustrated how complex and complicated were its social and architectural realities.
The 20th c. French sociologist and philosopher Henri Lefebvre, a pioneering critic of everyday life, starts his analysis of society from the humans and defines them as social beings who produce their own life, their own consciousness, and their own space (1991: 68-169). Thus, space, according to him, is not just a thing nor a container but a social reality and a set of relations and forms. It is a social product and a means of production as well. According to Humphrey, dwelling is about the active projection of the social and individual being by means of the artefact (1988: 18). The Roman period houses, accordingly, cannot be understood as only the architectural reflection of the domestic practices and traditions; it should also be considered as a social space where social relationships were continually produced and reproduced. The house was also an expression of the relationship of the Romans with the cosmos, as Knights argued, since it embodied symbols and features that linked the world of the mundane with the cosmological world of the spirits (1994: 102-131). This relationship with a shared cosmological framework, as embodied and expressed in the very fabric of the house, defined and guided the household to perpetuate a common sense of cultural identity and was conceived as the proof of participation in the Roman society (Grahame, 1998:
4 For an introductory study on the history of private life in the Roman period see Veyne, 1987, for studies on Roman family and household see Rawson, 1986; Gardner and Wiedemann, 1991 and Dixon, 1992.
5 See Hales, 2003 and Allison, 2007.
4
163; Hales, 2003: 1). The study of the Roman period houses and the households, as such, not only provides valuable information on the domestic practices, private art, and architecture but, as the ‘microcosm’ of the larger Roman society, also provides a lens to examine larger social and cultural issues related to the Roman world.
Although the recent studies that articulated the social use of Roman domestic architecture opened new avenues of interpretation and provided new perspectives for further studies, the current state of scholarship is still confined mostly to the western context. A great majority of the studies still focus mainly on the ample evidence that have come from the Campanian sites, or poured in from the cities and countryside of the Roman west. The ever-increasing archaeological data coming from the eastern provinces, on the other hand, have not yet found enough space in the historiography of Roman domestic architecture. This thesis, in this regard, aims to reduce this gap by taking the eastern extent of the empire, the province of Syria in particular, as its research focus. It aims to set the local evidence into the wider imperial panorama and to provide new contextual perspectives and a more inclusive framework to study Roman domestic architecture.
The eastern provinces of the Roman Empire were established in a distinct territory that had a geopolitical significance, multi-layered history, and peculiar cultural topography.6 The region carried profound importance for the empire in political, military, and economic terms and also had a considerable impact on the social, intellectual, and religious spheres as well as on the urban context, art, and architecture of the Roman realm. The continuous occupation of the region since prehistoric periods and its domination by successive civilizations had culminated into a rich cultural and architectural heritage that became more sophisticated with the addition of Roman elements. So, the Roman East, as the meeting place and the melting pot of the eastern and western cultures, constitutes a very exciting and fruitful context for Roman studies.
6 For comprehensive discussions on the contexts of the Roman East see Ball, 2000; Pollard, 2000; Butcher, 2003; Alston and Lieu, 2007 and Kaizer, 2022.
5
Its socio-geographical distance from the mainland and the social and cultural amalgamation that it accommodates make the region a charismatic arena to discuss such themes as identity, encounter, assimilation, acculturation, and ‘Romanization’ in social and architectural spheres.7
The rich archaeological heritage offered by the immense area that covered the lands of the eastern Roman Empire, and is commonly defined as the ‘Near East’ today, became a focus of interest as early as the 17th c. AD. The earliest visitors of the ruins in the region were mainly the European and American travelers and antiquarians, who pursued a passion to discover the ‘exotic’ Orient. Scientific surveys and systematic excavations started in the 20th century and the first expeditions mainly concentrated on discovering the monumental public buildings, some of which were, at least partially, still-standing or visible on the ground. The earliest discoveries of domestic architecture, on the other hand, were mostly the result of random or chance exposures and the unearthed remains were either neglected or superficially examined. It is only in the last few decades that domestic architecture has become a focus of attention thanks to the increasing interest in the archaeology of daily life and social contexts. This interest has led to the excavation of more domestic structures and promoted more comprehensive, focused, and long-term studies on the subject.
The archaeological work conducted in modern Turkey and Syria in the last hundred years has produced significant data on the urban and architectural contexts of the Roman East. Excavations carried out in both countries that covered the lands of the eastern Roman province of Syria have yielded a considerable number of houses that are dated to the Roman period. Though still meager compared to the ample evidence from the western contexts, the sample at hand constitutes an important and auspicious
7 Jas Elsner describes ‘Romanization’ as a process whereby Gauls, Libyans and Levantines might share the ideals of a single culture despite all their manifest differences (1998: 118). This thesis approaches the phenomenon of ‘Romanization’ in a similar way and discusses to what extent the residents of the eastern cities under consideration perceived themselves as ‘Roman’ and how they, consciously or subconsciously, expressed this identity in personal and communal self-representations as well as in art and architecture, especially in the domestic contexts. The phenomena of ‘Romanization’ and ‘Roman identity’ is discussed in the following chapters.
6
data-set for the study of Roman domestic architecture as represented in its eastern reach. The thesis studies this sample, coming from the eastern Roman province of Syria, in the context of five cities: Antioch on the Orontes, Apamea on the Orontes, Zeugma, Palmyra, and Dura-Europos. These cities provide a relatively remarkable and promising archaeological content, in terms of both quantity and quality.8
The currently available archaeological data that constitutes the main source of this study has some shortcomings and drawbacks, mainly due to their state of preservation, the extent and method of excavations, and the related archaeological records. The unearthed sample of domestic contexts under consideration here has not found adequate space in the current discussions on Roman domestic architecture, most likely because of these limitations. First of all, as mentioned above, the residential buildings were not the main focus of attention during the early excavations and were mostly discovered by chance exposures. The excavated sample was either neglected or was searched further for spotting attractive finds to move to the museum collections, such as the mosaic pavements that were taken from Antioch on the Orontes and parchments and papyri from Dura-Europos. In such cases, the architecture of the buildings themselves was often superficially mentioned in the excavation reports, sometimes with only sketchy and/or partly hypothetical plans. Secondly, except for the relatively later excavations in Zeugma and Apamea on the Orontes, the archaeological finds were not recorded or studied in context. The loose finds were frequently decontextualized for preparing object catalogs whereas the remaining empty structures were treated as only physical remains appropriate to investigate the domestic behavior, as Allison also pointed out for other excavations in Roman sites (2001: 185). Another important challenge for conducting research and doing scholarly publications on the subject is the existing political and cultural borders and the language of publications, both of which have limited international collaboration and communication between scholars.
8 Though not studied in detail, some well-preserved rural residences found at the Limestone Massif in Syria (Tchalenko, 1953; Sodini and Tate, 1984; Tate, 1992), the Hellenistic insulae excavated at Seleucia on the Tigris (Hopkins, 1972; Karampekos, 2020) and a partially excavated late Roman villa excavated at modern Urfa-Haleplibahçe/ancient Edessa (Karabulut, Önal and Dervişoğlu, 2011) are also mentioned in the thesis.
7
Syrian archaeology has been published primarily in French until recently, whereas in Turkey the publications are mostly in English and French. Some local missions in the region on the other hand, published the archaeological reports only in Turkish, Arabic, or Hebrew, and that has prevented the academic studies from crossing the national borders. Reaching some excavation reports and related academic sources was one of the main challenges during this study. Another important restriction for the thesis was the ongoing civil war and political instability in modern Syria that prohibited a site survey in the ancient cities of Apamea on the Orontes, Palmyra, and Dura-Europos. Most of the unearthed remains on these sites have been destroyed and looted during this conflict. In Turkey too, some excavated buildings are no longer visible in-situ. In Antioch on the Orontes, the exposed buildings were mostly recovered with earth, after the mosaic panels were transferred to the museums. In Zeugma, too, many excavated buildings are no longer visible as they have been flooded by the water of Birecik Dam.
Taking into consideration the current state of the data, the excavation reports and the published plans constituted the main source of reference for the thesis. The related epigraphic sources and the literature by the ancient authors, if available, were also consulted. Sightseeing in modern Antakya and Gaziantep made it possible to perceive the geographical contexts of Antioch on the Orontes and Zeugma. The ruins of Zeugma that have remained in-situ and the exhibited museum collections in the Hatay Archaeological Museum and the Zeugma Mosaic Museum in Gaziantep constituted an important source for the study of domestic contexts in these two ancient cities.
To sum up, this study dwells on the Roman period houses excavated from the five ancient cities of the eastern Roman province of Syria. It gathers the unearthed houses as a sample, some of which rarely appeared in publications, other than the excavation reports. Each house is first examined in detail, focusing on the archaeological and epigraphic data, and then put into a larger chronological and geographic framework.
In the first chapter, the social, political, and economic contexts of each city under investigation are overviewed. Though the large territory called the Roman East shared
8
a common and parallel historical evolution pattern, to a certain degree, the extent and means of Roman control and influence, as well as the local traditions, differed within. The diverse contexts present in the region had a visible impact not only on the social and cultural sphere of the settlements but also on their built environments. In addition, the ethnic, religious, and social identities of the citizens determined their lifestyle and were effective on the social and architectural formation of their environment, mainly of the domestic spaces. Thus, in the same chapter, the complex nature of the social and cultural formations in each city and the sense of identity of the inhabitants are interpreted by looking at the mediums of personal and communal self-representations among the contemporaries. In that respect, some light is thrown on the language(s) used in different contexts, the onomastic in written sources, religious rituals and monuments, and some other material culture.
The next five chapters are devoted to the five cities under investigation. The order of the cities is determined according to their urban magnitude and importance, and the first chapter starts with Antioch on the Orontes, followed by Apamea on the Orontes, Zeugma, Palmyra and Dura-Europos. The order more or less coincides with the geographic location, and moves from west towards east. Each chapter starts with the study of the urban layout and development of the settlement that was effective, to a certain extent, in the determination of the location, orientation, layout, and even size of the residential buildings. In the next section of the same chapter, the history of excavations and the quantity and the quality of the available archaeological data, and the related literature is summarized to predetermine the limits and the pros and cons of the data that constitute the main reference. Then, in the light of this evidence, the plan layouts and the architectural and decorative features of the unearthed domestic buildings are scrutinized one by one in each city. The sample in each case is studied in a chronological order, according to date of construction and/or to size and splendor. The last sections of the five chapters are devoted to an overall and critical evaluation of the domestic contexts of each city.
9
In the last part of the thesis, the sample examined in the earlier chapters and the emerging outcomes and comments are put into a wider perspective and comparative framework. This chapter, in this respect, starts with an overview of the social and political dynamics of the region and the characteristics of the population structure. Then it concentrates on the domestic practices, spaces, and structures of the eastern province of Syria and evaluates them under two main headings. First, the domestic buildings are overviewed, concerning the urban contexts, and then their architectural layout and vocabulary are reconsidered in a chronological framework. As such, the continuities, changes, and transformations in the domestic context of the Roman east from the Hellenistic period until the end of antiquity are illustrated.
10
CHAPTER 2
THE ROMAN EAST:
GEOGRAPHICAL, POLITICAL AND SOCIAL CONTEXTS
2.1 Antioch on the Orontes: The Provincial Capital of Syria
The remains of Antioch, one of the major and famous Greco-Roman cities, lie beneath the modern town of Antakya, in the south-central coastal region of Turkey. The city was located on the east bank of the Orontes River and was demarcated, in the east, by the Mountain Silpius (modern name is Habib Neccar) which runs roughly parallel to the river and rises to 506 m from the sea level. 9 km south of the city, on a higher level than Antioch and overlooking the Orontes, was the plateau of Daphne (modern Harbiye). Daphne was a fertile area and had a beautiful landscape with numerous natural springs providing an ample supply of water both for its close territory and also for Antioch. The close vicinity of Antioch, including the lower Orontes valley and the Amuq plain was unusually fertile, like today, and provided a generous and valuable agricultural land (De Giorgi and Eger, 2021: 6).
This spectacular geography and its abundant water supply were probably among the reasons for Seleucus’ choice of the site for his new city. Strabo, in his work Geography, described the city as such:
Seleucis is not only the best of the above-mentioned portions of Syria, but also is called, and is, a Tetrapolis, owing to the outstanding cities in it, for it has several. But the largest are four: Antiocheia near Daphne, Seleuceia in Pieria, and also Apameia and Laodiceia; and these cities, all founded by Seleucus Nicator, used to be called sisters, because of their concord with one another. Now the largest of these cities was named after his father and the one most strongly fortified by nature after himself, and one of the other two, Apameia, after his wife Apama, and the other, Laodiceia, after his mother..… Antiocheia is likewise a Tetrapolis, since it consisted of four parts; and each of the four
11
settlements is fortified both by a common wall and by a wall of its own. Now Nicator founded the first of the settlements, transferring thither the settlers from Antigonia, which had been built near it a short time before by Antigonus; the second was founded by the multitude of settlers; the third by Seleucus Callinicus; and the fourth by Antiochus Epiphanes (16.2.4).9
Antioch was founded with three other ‘sister cities’ -Seleucia Pieria, Apamea on the Orontes, and Laodicea- as Strabo designates, in 300 BC by Seleucus I Nicator, the first Seleucid king and one of the commanders of Alexander the Great, as a part of Seleucid colonization program initiated for military purposes. The four cities which were inhabited by the Macedonians and Greeks played a vital role in north-western Syria to assure the domination of the Macedonian power in the newly conquered territory (Downey, 1961: 54). During the reign of Antiochus I, after the death of Seleucus I in 281 BC, Antioch replaced Seleucia Pieria, as the new capital. Antiochus I was an active patron of scholars and scientists and the city acquired fame as a seat of philosophers, astronomers, and historians during his reign (Najbjerg and Moss, 2014: 26).
In 246 BC Antioch was occupied by Ptolemy Eurgetes of Egypt for a few years until Seleucus II reclaimed the city and made it the principal seat of the Seleucid power. In the first half of the 2nd c. BC, while the Seleucid power was in decline, following the defeat of Antiochus the Great by the Romans in 190 BC the city saw remarkable prosperity in the reign of Antiochus IV Epiphanes who ruled between 175 BC and 164 BC. Implementing an active construction program, he brought his capital to a state of prosperity and significance that placed it among the notable cities of antiquity (Downey, 1961: 95). After the death of Antiochus IV in 164 BC, however, the city saw a steady decline in parallel to the growing weakness and dissolution of the empire caused by the struggles for the throne, Parthian invasions, and earthquakes. Finally, in 64 BC Pompey the Great brought an end to the Seleucid dynasty, annexed Syria to the Roman Empire as a new province and made Antioch its capital (De Giorgi and Eger, 2021: 71).
9 Translations of Strabo used in this thesis are by H. L. Jones (1917).
12
Under Roman rule, Antioch gradually regained and even surpassed its former grandeur. When Caesar defeated Pompey in 48 BC, the city supported Caesar who, in return, commenced a building program to endow the city with monumental Roman buildings (De Giorgi and Eger, 2021: 78-80, 116). In the following years, nevertheless, the city witnessed another Parthian invasion that lasted until the Pax Augusta (De Giorgi and Eger, 2021: 78).
A period of peaceful growth for the Roman world, as well as for Antioch, began from Augustus onwards, during whose reign (27 BC-19 AD) came the end of the civil wars. Augustus elevated the status of Syria from a senatorial province to an imperial one that was to be governed by a legate, legatus Augusti, and a procurator in charge of financial matters, both of whom were appointed by the emperor (Downey, 1961: 164). The post of the legate of Syria became one of the most important in the Empire and on occasion, the legate acted as the supreme commander in the Roman East. Both the legate and the procurator of Syria had their headquarters at Antioch.
Under Roman power and protection, commercial activities flourished in Syria, the thoroughfare of some of the most important trade routes in the Roman Empire. Thanks to its location on a strategic point along this thoroughfare and its position as the military and political headquarters, Antioch witnessed a boom in the economic activities that took place in the city.
In parallel to the political and economic welfare, the deeds of imperial euergetism too had a considerable effect on the life and urban development of the city. One of the major events in the history of Antioch initiated under the reign of Augustus was the foundation of the local games, called the Olympic Games of Antioch, an event that became one of the most famous festivals of the Roman world. Augustus was also a generous benefactor of the city and showed a keen interest in building, repairing, and improving the public buildings. Thus, under the reign of Augustus and his successors, mainly in the course of the first two centuries AD, the city was endowed with several
13
impressive public structures including colonnaded streets, bath complexes, temples, and aqueducts (Najbjerg and Moss, 2014: 28).
Antioch as such, transformed from being the capital city of the strategically important and wealthy province of Syria to a metropolis and hence to one of the most important cities of the Roman Empire, together with Rome and Alexandria.
During this period the city had suffered from a series of catastrophic events as well. A disastrous fire in 23/24 AD, three destructive earthquakes circa 38 AD, 40s AD and 115 AD, a famine in 40s AD, and several anti-Jewish revolts caused considerable damage and destruction in the city and also a decline in its population (Downey, 1961). Nonetheless, Antioch seems to have recovered, at least after some of the catastrophes, by the financial support of the emperors themselves.10
The economic and political instabilities within the empire in the late 2nd and 3rd centuries AD, including constant struggles for power and several invasions affected Antioch negatively. One of the consequences of this period of turmoil was seen in the reign of Marcus Aurelius (161-180 AD) in 175 AD, when Avidius Cassius, the governor of Syria proclaimed himself emperor and Antioch supported him. In return, the city was punished by the emperor and the games and spectacles of Antioch, including the Olympic Games, were abolished circa 175 AD; emperor Commodus would restart the games in 181 AD (Downey, 1961: 231). In the reign of Septimius Severus (193-211 AD) Antioch was punished once again, for supporting Pescennius, the governor of Syria. This time the city was deprived of its title of both the metropolis and the capital of Syria and the Olympic Games were removed from the city (Najbjerg and Moss, 2014: 32). The emperor Caracalla, son of Septimius Severus, would restart the traditional games, and stay in the city several times to conduct his military campaigns against the Parthians.
10 In the severe earthquake of 115 AD, for instance, the emperor Trajan was in the city and he managed to escape with a few minor injuries. He then financed the repair of some of the public buildings in the city (Downey, 1961: 213-218)
14
The rise of the Sassanid Empire in the East in the first half of the 3rd c. AD constituted an increasing threat for the Roman Empire. Antioch was sacked by the Persian troops in 256 and 260 AD which gave destructive damage to the city. In 266 AD the city was captured by the Queen Zenobia of Palmyra and stayed under the Palmyrene reign for six years (Najbjerg and Moss, 2014: 32).
In the reign of Diocletian (284-305 AD) when the economy and political situation of the empire were stabilized and the military was strengthened, Antioch managed to recover to a certain extent. The emperor himself visited the city several times and endowed several public buildings, the most important of which was a new imperial palace, which demonstrates the resumed importance of Antioch in this period, as a military and administrative headquarters (De Giorgi and Eger, 2021: 137-141).
The reign of Constantine (306-337 AD) marked a turning point not only in the history of the Roman Empire but also in the destiny of Antioch. The city became a prominent center of Christianity after the conversion of Constantine. Before the Edict of Milan issuing freedom of religion in 313 AD, a small organized Christian community was already living in Antioch since the first half of the 1st c. AD and the city was an important center for the missionary activities of St. Paul. Similarly, St. Matthew, one of the four gospel writers, may have written his account of Christ’s life in Antioch in the mid-1st c. AD. In the 7th c. AD Antioch became one of the five patriarchates into which the Christian world was divided, the others being Rome, Alexandria, Jerusalem, and Constantinople.
The Late Antique emperors, too, paid attention to the improvement of the city and made donations, especially after the devastating events. The city suffered a series of more catastrophic disasters, including several earthquakes, an outbreak of bubonic plague, big fires, several revolts, and invasions in the 4th, 5th, and 6th centuries AD. The earthquake of 528 AD was especially severe and destroyed almost the entire city, killing 250,000-300,000 inhabitants. In 540 and 573 AD the Sasanians invaded Antioch, looted the city, and set it on fire while returning to Persia with lots of
15
inhabitants as their captives. As a consequence, the size of population and of the urban area decreased in time. Finally, in 638 AD the city was occupied by the Arabs who ended the era of the Greco-Roman rule in the city which had lasted nearly a millennium (Najbjerg and Moss, 2014: 34).
The Muslim control of Antioch lasted more than three centuries, during which it was taken back by the Byzantines in 968 AD. The city subsequently saw the control of Seljuks, the crusader princes and regents, Memluqs, and the Ottomans. In 1939 it became a Turkish province called Hatay.
2.1.1 The Population and Community of Antioch on the Orontes
Antioch was essentially founded as a Greek city, colonized primarily with Greek-speaking people. Many of the first inhabitants were retired Macedonian soldiers from Seleucus’ army, including Jews, while others were Athenian citizens, including the inhabitants of Antigonia (Downey, 1961: 77-82; Najbjerg and Moss, 2014: 22). The population, in addition, seems to have included a group of indigenous Syrians who were either settled in a separate walled area adjoining the principal foundation area or settled in an area outside the city, which was originally open and later enclosed by a wall that joined the wall of the original foundation (Downey, 1961: 80; Müller, 1839; 28-29). With reference to the ancient sources Downey makes an estimation between 17,000 and 25,000, excluding slaves, for the size of the population of the Hellenistic city.
Under Roman control, as the locus of power at the eastern crossroads, Antioch blossomed as a metropolis. The city, with a diverse blend of population, including Greeks, Jews, Syrians, Christians, and Latins, developed into a splendid cosmopolitan urban center. Since it was settled at a junction of major trade routes and was an important trade center, merchants and tradesmen from various places also frequented
16
the city. The estimations of the total population of Antioch in its peak varied between 250,000 and 800,000 (Downey, 1958; Kloeg, 2013).11
The inhabitants of the urban center of Antioch were predominantly Greek in language, religion, and culture. Many other languages including Latin, Hebrew, Aramaic, Syriac, Coptic, and Persian were also spoken on the streets by the citizens. On the other hand, people in the outlying countryside spoke Syriac, a Semitic dialect (Maas, 2000: 16). Roman influence was much less imposed in the countryside than in the urban center, as was mostly the case in the eastern provinces.
2.2 Apamea on the Orontes: The Provincial Capital of Syria Secunda
The ancient city of Apamea on the Orontes is located on the right bank of the Orontes River, about 55 km north-west of the modern city of Hama in Syria. The city was laid out on a rich agricultural plateau that was enclosed in a loop of the river which, with the lake and marshes, gave it a peninsular form. The site is bordered by the mountains of Gebel Zawiye in the north and extends through the steppe plateau of northern Syria, commonly known as the Limestone Massif, in the east. It overlooks the Ghab valley in the west (Dewez, 1969a: 23).
The earliest habitation in the site goes back to prehistoric times when the ancient settlements had developed mainly around the imposing hill rising a hundred meters from the Ghab valley, on the western edge of the city. Excavations revealed that the site was occupied from the middle Paleolithic period onwards and was an important settlement in the Neolithic and Bronze Ages (Dewez, 1969b: 53-60). A reused Hittite hieroglyph stele discovered in 1937 indicates that in the middle of the 9th c. BC
11 For comparison, the population of Rome at its peak is estimated as 1,000,000 by Stambaugh (1988: 90) and Robinson (1992:8) whereas it is estimated as 440,000 by Storey (1997). For Ephesus O’Connor (2008: 130) makes an estimation of 138,000-172,000, whereas Hanson (2011: 252-258) suggested a much lesser number in between 33,600 and 56,000. For Alexandria it is estimated as 500,000-600,000 by Delia (1988), for Pompeii as 8,000-12,000 by Wallace-Hadrill (1991: 199-200) and for Ostia as 22,000 by Storey (1997: 975) and 27,000 by Packer (1971: 70).
17
Apamea was a part of the Hittite King Hamath’s territories. Later, the city passed under the domination of the Assyrians and then the Achaemenid Persians by whom it was called Pharnake (Balty and Balty, 1969: 30).
After the conquest of the region by Alexander the Great in the late 4th c. BC, the small Persian town of Pharnake received a Macedonian garrison and was renamed as Pella, the name of the native city of Alexander in Macedonia. Circa 300 BC Seleucus I Nicator established a colonial foundation and changed the name of the settlement to Apamea, after the name of his wife. The city became an important military base and saw significant urban development in the Hellenistic period, under Seleucus I Nicator and his successors (Cohen, 2006: 94-95). Strabo described Apamea as one of the four cities of Seleucid Tetrapolis (Geography: 16.2.4). He pointed out the strategic importance of the site and the fertility of the surrounding region. He also emphasized the military importance of the city where, according to him, most of the Seleucid army, elephant contingents, and horse breeding facilities were housed and the trainers of heavy armor fights who were paid to teach the arts of war were employed (Geography: 16.2.10).
In 64 BC Pompey destroyed the Seleucid citadel and integrated Apamea on the Orontes into Roman territory. The importance and vitality of the city, with a lessened military focus, continued for many centuries under Roman sovereignty thanks to its rich pasturelands, agricultural wealth, and its location within the Eastern trade routes. In the 1st c. AD Apamea received the particular attention of the emperor Claudius and was named as ‘Claudia Apamea’ (Balty, J. Ch., 2000: 459-481). The city carried his epithet until the mid-3rd c. AD. In the early 3rd c.AD, the association of the greatest imperial dynasty, the Severans, with Apamea is likely to have contributed to the flourishing of the city among the others (Ball, 2000: 161; Balty, 1988: 97). Like Antioch, Apamea accommodated regional games and festivals and hosted many athletes and entertainers. It stimulated commercial exchange on local and regional scales and attracted many enterprising merchants. Intellectual life also developed at Apamea and the city became an important center of knowledge in the Roman East. As
18
such Apamea became the second-largest city, after Antioch, and one of the most opulent and ostentatious settlements of the eastern empire (Balty, 1988:91-96).
In the 3rd c. AD Apamea regained its military importance with the resumption of the Persian wars at the Roman eastern border. Due to its strategic geographical location, the city functioned as a base for the operations during the several campaigns organized against the Parthians and Sasanians. The camp of Apamea served as the winter headquarters of a special legion, the Legio II Parthica, which accompanied the emperors on their expeditions, in the first half of the 3rd c. AD (Balty, 1988: 99-103). Moreover, one of the four legions assigned for the defense of the province of Syria and of the eastern limes, the Legio III Gallica was stationed at Raphanee, 50 km away from Apamea (Balty, J. Ch., 1991: 22).
In 252 AD the Persian King Shapur invaded Apamea and occupied the city for a short period of time. He massacred several inhabitants and deported many to Persia. The city, however, managed to recover shortly and regained its vitality and prosperity. The 4th c. AD was a particularly splendid period for Apamea, as it was for Antioch. Apamea became the capital of the province of Syria Secunda in the early 5th c. AD. In the course of late antiquity the city became an important center of an archbishopric and a center of philosophy and theology. Its intellectual environment paved the way for the emergence of Eastern Christian sects (Finlayson, 2012: 284).
Apamea was heavily damaged after the earthquakes of 526 and 528 AD but managed to recover, especially thanks to the personal intervention of Justinian, and continued to prosper to some extent in the 6th c. as well. The city was ravaged and burned by the Persians in 573 AD and further damaged by an earthquake in 588. After a short Persian occupation between 613 and 628 AD, it was retaken by Heraclius. After the Battle of Yarmuk in 636 AD the city passed under the Arab rule and was renamed Famieh or Famia. In the Crusades period, it was occupied by Tancred and remained within the principality of Antioch until 1149 when Nur-ad Din captured the city. In the 17th c. the settlement was given the name of Qal’at al Mudiq.
19
2.2.1 The Population and Community of Apamea on the Orontes
In the 6th c. AD the census of Q. Aemilius Secundus, the population of the city of Apamea on the Orontes was counted as 117,000 hominum civium. What this number exactly implies, however, is unclear. Cumont and some other scholars supposed that the term ‘hominum civium’ stands for only the adult males and they suggested a number of 400,000-500,000 for the total population (Cumont, 1934: 189; Balty, 1988: 96; Millar, 1993: 250). Butcher and Kennedy, on the other hand, found this number unsustainable for Apamea and argued that the number should include all free citizens excluding only slaves, resident foreigners, and nomads and suggest a total population of approximately 125,000-130,000 (Butcher, 2003: 106; Kennedy, 2006). Yet, there is a consensus that this count included both the town, which covers an area of 250 ha within the walls and its territory. In the 2nd c. AD, in the era of peace and prosperity, the population of Apamea must have increased considerably.
According to the funerary monuments, the population of Apamea was dominated by Semitic and Greek citizens. Latin names appear very rarely. The nomenclatures of the soldiers and inscriptions refer to many Thracian names that were common in Greek-speaking countries and show that most of the recruits were enrolled in the Balkan area (Balty, 1988: 102).
In the 2nd c. AD, in parallel to the population increase in Apamea, new social classes emerged in the region. The foundation of large estates in the province, in especially the villages of the north-Syrian limestone plateau, and the building of large mausolea and hypogea indicate the emergence of a new class of landowners, who had previously been state officials or had made a career in the army (Balty, 1988: 96; Tchalenko, 1953: 381-382).
20
2.3 Zeugma: A Bridge on the Euphrates
Zeugma is located near Belkıs village, within the city borders of Gaziantep in the south-eastern region of Turkey. The ancient city was laid out on the bank of the Euphrates River, within a territory of distinctive geographical characteristics. The Hobap plain on the left bank offers extensive and rich soils for cultivation while the steep topography and the hills on the right bank provide regional security. Several streams and springs, as well as the Euphrates River, supply a perennial water source for this region (Aylward, 2013a: 26).
Besides its geographical significance, the Euphrates River was important in this territory for economic and political reasons. Zeugma was one of the two places where the Euphrates could be crossed most easily in the past, the other was Samosata. Thus, the river not only acted as a water route for fluvial transport but also assumed special importance in that spot, as the crossing point for the land routes. From the Classical period onwards, the river also acted as a border between different political powers reigning on its eastern and western banks (Edwell, 2008: 7).
Due to this geopolitical significance, the vicinity of Zeugma had always been an attractive spot for settlement. Surveys and excavations have indicated that the history of settlement in the area at the Middle Euphrates, goes back to the prehistoric periods. Archaeological finds and cemeteries in the eastern and western banks of the Euphrates River, in the close vicinity of Zeugma, including Tilmusa, Tilebür, and Horum Höyüks, prove that this area had been settled since the Chalcolithic and Bronze Ages (Sertok and Ergeç, 1999a; 1999b: 86-89; Marro, Tibet and Bulgan, 2000).
In the Classical period, however, this area saw a flourishing settlement context and the total occupied area doubled the one of the Iron Age and the focus of occupation within the region shifted from Charchemish to the north of Birecik (Algaze et al., 1991: 206). It has been emphasized by some scholars that, the significant growth on the banks of the Euphrates is related to the imperial policies of the Seleucids and the Romans in
21
which the river played the role of both being a bridge and a border with the Parthians and Sasanians in the east (Algaze et al., 1991: 206). Thus starting from the early 3rd c. BC several urban foundations, including Zeugma, were established along the Euphrates River to protect this border.
Zeugma was founded, probably as a military colony, circa 300 BC by Seleucus I Nicator. Zeugma was originally planned as a twin town that faced each other across the Euphrates. On the western bank of the river was Seleucia, which was named after its founder and on the eastern bank was the city of Apamea, which was named after the Persian wife of Seleucus I Nicator. Both cities were located on the ancient caravan routes and the Silk Road and were connected by a bridge that provided the link between the Mediterranean Coast, Anatolia, and Mesopotamia. Pliny provided one of the earliest testimonies on the foundation of the twin towns and the bridge, in his Natural History dated to 1st c. AD:
The towns washed by the river are Epiphaneia and the Antioch called “on-the-Euphrates”, and also Zeugma, 72 MP from Samosata, famous as a place where the Euphrates can be crossed, Apamea on the opposite bank being joined to by a bridge, constructed by Seleucus, the founder of both towns (NH.5.21).12
In the 2nd c. BC, especially after the defeat of Antiochus the Great by the Romans at the Battle of Magnesia in 190 BC, the grip of the Seleucid Empire on the East was weakened. In 162 BC the kingdom of Commagene arose, as an offshoot of the Seleucid Empire, establishing its capital at Samosata. In the south, on the other hand, appeared independent, powerful and predatory Arab chiefs. In 88 BC the Parthians took control of Zeugma for a short period of time.
In 64 BC when the Roman general Pompey invaded Syria, Antiochus, the king of the Commagene Kingdom and his ally was rewarded with the city of Seleucia. In 59 BC he was given official recognition from the Roman senate as an ally to Rome. Shortly
12 Translations of Pliny used in this thesis are by H. Rackham (1942).
22
after, Seleucia became one of the four largest cities of the Commagene kingdom. The Greek historian and geographer Strabo mentions Seleucia, as a city of Commagene, in his famous work Geographica as such:
… Commagene is rather a small country; and it has a city fortified by nature, Samosata, where the royal residence used to be; but it has now become a province; and the city is surrounded by an exceedingly fertile, though small, territory. Here is now the bridge of the Euphrates; and near the bridge is situated Seleuceia, a fortress of Mesopotamia, which was included within the boundaries of Commagenê by Pompey; and it was here that Tigranes slew Selenê, surnamed Cleopatra, after imprisoning her for a time, when she had been banished from Syria (16.2.3).
The Hellenistic civic and cult centers found at Zeugma date mostly to the time of Antiochus I, the late Hellenistic king of Commagene. Archaeological evidence also shows that the Commagenian ruler cult was dominant in Zeugma, in at least the 1st c. BC. The surveys and excavations at Zeugma so far revealed several steles with dexiosis reliefs and inscriptions that mention the ruler cult (Rose, 2013; Crowther, 2003; Crowther and Facella, 2003; Wagner, 1976: 117-123). Görkay suggests that Zeugma may have played a key role in Antiochus’ efforts to link the Greek and Persian cultural spheres, in accordance with the kingdom’s key position between these two worlds on the Euphrates (2011: 39).
In 31 BC, with the reign of the Roman Emperor Augustus, the kingdom of Commagene was made a Roman client state. The dynasty of Antiochus continued under his grandsons until it was deposed by Vespasian in 72 AD when Commagene became part of the Roman province of Syria. After Seleucia became a Roman city, it took the name Zeugma which literally means ‘bridge’ or ‘crossing’ in ancient Greek (Görkay, 2011: 37).
From the time of Pompey’s establishment of the province of Syria in 65 BC, Euphrates came to symbolize a boundary between the Roman and Parthian worlds in the Near East (Millar, 1993: 29-30). On a number of occasions over the next 130 years, meetings between senior Roman officials or members of the imperial family and
23
Parthian representatives were held on the Euphrates, confirming its status as a boundary during this time (Edwell, 2008: 7).
Under Roman rule and especially with the reign of Vespasian (69-79 AD), the river’s role as a symbolic boundary had given way to a more practical one. The Legio X Fretensis was most probably based in Zeugma in 18 AD and a military camp was built nearby in 49 AD. Vespasian, reorganized the Near Eastern territory and established four legions on or near the Euphrates. One of these was the Legio IV Scythica which replaced the Legio X Fretensis in 66 AD. As such, Zeugma became one of the largest and most important eastern frontier cities of the empire, in both military and economic terms (Başgelen, 2000: 11; Görkay, 2011: 39).
In the second half of the 2nd and first half of the 3rd c. AD the internal political developments in both the Roman and Persian empires changed the nature of the conflict on Rome’s eastern frontier. Annexation of North Mesopotamia under Lucius Verus and reorganization of the frontier by Septimius Severus in the late 2nd c. AD brought a period of peace in the eastern part of the empire. Accordingly, many troops from the Legio IV Sychitica who used to station at Zeugma moved away. This must have had a considerable impact on not only the demography of the city but also its economy.
The overthrow of the Parthians by the Sasanians in the 220s AD marked the beginning of a period of instability in the Roman Empire. Sasanian invasions began to threaten Roman power and influence in the provincial territories of Syria and Cappadocia. In the 250s AD the Sasanians under the command of King Shapur passed the Euphrates, invaded Syria and sacked many of the Roman cities on the Euphrates, including Zeugma (Edwell, 2008: 184-190).
Archaeological evidence demonstrates serious destruction and fire at Zeugma, in the mid-3rd c. AD. The destruction deposits that have been recovered in the excavations, especially in the excavated residential structures, suggest that the inhabitants of
24
Zeugma who anticipated the Sasanian invasions into northern Syria managed to escape before the sack. The lack of human and animal skeletons suggests that the city was not inhabited during the sack (Aylward, 2013a: 30).
Until the Late Byzantine period, Zeugma remained as the principal settlement in the region but it seems that it never reached the prosperity of the earlier centuries. For the Late Antique and Early Byzantine periods, a substantial peak has been pointed out in the occupational history of the lower Euphrates region within the borders of Turkey, both in the number of sites and the total occupied area (Algaze et al., 1991: 207; Algaze, Breuninger and Knudstad, 1994: 22-23). After this period, however, there was a sharp depopulation in the same region where the total occupied area declined to less than 5% of the Classical Age.
Zeugma was resettled after the Sasanian destruction. While a funeral inscription published by Wagner suggests a military existence in the city in the 4th c. AD, the evidence for habitation in the 4th c. AD is minor (1976: 262, plate 52). By the 5th c. AD, on the other hand, some houses of rather high quality reappeared which indicates at least a partial revival in that period.
In the 7th c. AD Zeugma was probably seized by the Arabs and became a part of the new east-west frontier zone between Islamic Syria and Byzantine Anatolia. Excavations revealed some Abbasid material remains at Zeugma, but the Arab occupation seems to have covered a small area. There are several references that give information about the Christian bishops who had settled in Zeugma until the beginning of the 11th c. (Ergeç, 2005: 19-22).
Birecik continues to be mentioned in the sources after the 11th c., but there is no reference to Zeugma until the 18th c. It is likely that after 11th or 12th c. the settlement reverted to a succession of small villages as is the case today. In the 17th c., the village of Belkıs was founded on the site by Turkish settlers.
25
2.3.1 The Population and Community of Zeugma
For a group of Hellenistic cities covering an approximate area of 65-80 ha, including Zeugma, Grainger suggested a population of around 25,000 (1990: 91-92). In the Roman period, on the other hand, particularly in the 2nd and 3rd centuries AD when the city of Zeugma was at its peak in extension and prosperity, the twin towns together have covered an approximate area of 190 ha with an estimated population of 50,000-75,000 (Kennedy, 1998a: 33). Like many other Greco-Roman cities in the East, the population of Zeugma was composed of a cosmopolitan society with diverse ethnic and religious identities.
Two main groups of archaeological data from Zeugma provide information on the ethnic and religious origins of the inhabitants. The first group includes the inscriptions which area revealed from different contexts and present different languages and nomenclatures, and the second includes a variety of burial traditions as well as with some epigraphic and visual data inscribed on the steles that came from the necropolis.
Excavations at the sacral area of Belkıs Tepe revealed the remains of a pre-Hellenistic structure in 2008. One of the finds from the area is a fragment of a basalt block, most probably belonging to a statue. The block inscribed with Old Aramaic letters and probably dating to the Early Iron Age indicates the existence of a local Semitic population in the city in the pre-Hellenistic period (Görkay, 2011: 46).
In the Hellenistic and Roman periods, the common language used at Zeugma was Greek. The inscriptions of the civil population are overwhelmingly Greek, while the names were often Semitic. Even though an Aramaic dialect was probably spoken too, in both the city and the countryside, there is almost no written evidence for native languages (Kennedy, 1998c: 240). The inscriptions on the burial steles were also dominated by Greek and Hellenized Semitic names both written in Greek script. There are also a considerable number of Latin inscriptions, almost all of which belonged to the soldiers and came from the western necropolis (Kennedy and Graf, 1998: 107-108;
26
Yaman, 2013: 267). In addition to these, one inscription with a Jewish name, and four proto-Syriac inscriptions were also found in the necropolis of Zeugma and Apamea (Barbet, 2005: 293; Abadie-Reynal and Ergeç, 1998: 403-406).
The necropoleis of Zeugma which were in use between the 3rd c. BC and 3rd c. AD present a variety of burial traditions, from the simple tombs of the Greco-Macedonian inhabitants carved into the bedrock to tumuli type of burials and tombs with loculi dating to the late Hellenistic and early imperial period (Görkay, 2012a: 294). Both Parthian and Greek dresses were noted on the portraiture of the tombstones, while the frontality in representing the human body has been associated with the North Syrian tradition (Kennedy, 1998c: 241).
2.4 Palmyra: A Commercial Oasis in the Desert
Palmyra is located in south-central Syria, approximately 230 km north-east of Damascus. At the center of the Syrian Desert, the ancient city was bordered by Palmyrene mountain ranges in the north and south-west and was exposed to the desert on its south and east. A small valley, Wadi al-Qubur, crosses the city, from west to south-east. Palmyra is located in an oasis surrounded by palm trees in this geography. The site and its territory had the advantage of exploiting fertile lands and several springs. In the 1st c. AD Pliny the Elder described the city as:
Palmyra is a city famous for the beauty of its site, the riches of its soil, and the delicious quality and abundance of its water. Its fields are surrounded by sands on every side, and are thus separated, as it were, by nature from the rest of the world (NH. 5.21).
The earliest habitation in the territory of Palmyra goes back to the Neolithic period, the traces of which were found around the natural spring of Efqa to the south of the city. Bronze and Iron Age occupation layers were exposed in the ancient tell of the city that was buried underneath the temenos of the Temple of Bel (Al-Maqdissi, 2000). From the beginning of the 2nd millennium BC Palmyra was mentioned in the Near Eastern historical documents, including the Kültepe and Mari Archives, with its pre-
27
Semitic name Tadmor, which, according to Starcky and Gawlikowski, presumably meant ‘guard-post’ (1985: 33-34). The city of Tadmor was ruled by the Assyrians and Persians respectively until it was incorporated into the realm of the Seleucids.
Palmyra remained a relatively small Hellenistic settlement until the middle of the 1st century BC (Seyrig, 1950: 1; Bryce, 2014: 277). The disintegration of the Seleucid Kingdom and the establishment of the Roman province of Syria had little impact on the governance of the city and it was initially left independent (Bryce, 2014:278). Palmyra became a tributary of the Roman Empire under Tiberius, probably circa 19 AD when the statues of Tiberius, his son Drusus and his nephew Germanicus were set up in the Temple of Bel (Gawlikowski, 2015: 252).
Epigraphic evidence demonstrates that Palmyra’s administrative structure was changed in the course of the second half of the 1st c AD. In the Hellenistic period, the city used to be governed by an assembly composed of the representatives of the four main tribes. By the reign of Vespasian, this organization was transformed into the structure of a Greek city with the institution of an assembly and a council, called by the Greek terms demos and boule (Edwell, 2008: 48; Millar, 1993: 324). After the elevation of Palmyra to a colonia around 213–216 AD, the city received Roman provincial governors and ceased to pay taxes. It incorporated Roman institutions into its system while keeping many of the earlier Greek ones (Millar, 1993: 74-83).
In 129 AD Hadrian visited the city and declared Palmyra a ‘free city’, allowing it to set and collect its own taxes. After his visit, the city began to bore the epithet of Hadriana. In the early 3rd c. AD, under the reign of Septimius Severus or Caracalla, the city became a Roman colony, in the newly founded province of Phoenice, and its inhabitants gained equal rights with those of Rome and the other cities of the empire (Millar, 1993: 326).
In the later years of the Hellenistic period, Palmyra began to prosper and in the first two centuries of the Roman imperial period, the city experienced its greatest economic,
28
military, and urban development. Located in the borderland of two great empires and the halfway between the Euphrates and the Mediterranean, Palmyra had the opportunity to act as an intermediary of the commercial network between East and West. The city was an important station for the caravans traveling between the Mediterranean, Mesopotamia, and Arabia. It was also an important link on the old Silk Route that went from China and India to Europe. Moreover, the Palmyrenes developed a reputation of protecting the caravans passing through the oasis with units of mounted archers commanded by strategoi (Edwell, 2008: 32)13. This ability to control the flow of the caravan traffic in the region gave Palmyra the opportunity to levy tariffs on products that passed through the oasis which indeed became the main source of the considerable wealth of the Palmyrenes.
In the 2nd c. AD, with the extension of Roman power along the Euphrates and in Mesopotamia, the territory of Palmyra became even more important for Roman military interests. Initially, the Palmyrene auxiliaries were used in other parts of the empire. Then a permanent military garrison was established in the city and it became larger in the early 3rd c. AD. As such, the Roman influence and power at Palmyra became more direct and formalized in the course of the 2nd c., especially after the reigns of Trajan and Hadrian, while it had been expressed mainly in economic terms until then (Edwell, 2008: 31-61).
In the course of the 3rd c. AD however, the internal power struggles within the Roman Empire, the rise of the Sasanians in Persia and their rising aggression weakened Roman control over the region. The political instability between Rome and Persia and the declining security in the region affected the Palmyrene trade and economy. In such a period of disorder, a member of one of the leading families of Palmyra, Odenathus, who achieved senatorial rank in the early 250s, came into prominence. He assumed a leading political and military role in the community and was identified as the leader or
13 Strategos, plural strategoi, meant military general in Greek. The term was also used to describe a military governor in the Hellenistic world (Rostovtzeff, 1932: 6).
29
chief of Tadmor in 252 AD (Millar, 1993: 157). After the invasions of the Sasanian king Shapur in 259/260 AD and the defeat and capture of the Roman emperor Valerian at the Battle of Edessa, Odenathus formed an army. With several successful military campaigns against Shapur he managed to recover most of Mesopotamia and forced Sasanians to retreat back to the Tigris. As a reward for his military successes, he was bestowed with the title of “the Governor of the East” or “the Restorer of the East” (Bryce, 2014: 289-291). Odenathus was assassinated on his return from his latest campaign, in 267 or 268 AD (Southern, 2008: 4).
After the death of Odenathus, his second wife Zenobia took over the control, on behalf of his ten-year-old son Vaballathus. Zenobia was a wealthy and educated Roman woman. In Classical and Arabic sources, she was identified as a woman riding, hunting, and drinking with her officers, like a man, on occasion (Ball, 2000: 78). The short reign of Zenobia, which lasted for approximately 5 years, was dominated by expansionist military campaigns. In 269 AD, she initiated her first campaign to the province of Arabia against the tribal groups, especially the Tanukh confederation, which threatened Palmyra’s trading activities in the region. In this victorious campaign, she seized the provincial capital of Bostra. She then expanded her military program and took much of Roman Syria and Arabia under her control. In 270 AD, she invaded Egypt and declared herself the Queen of Egypt. At the same time, her army had advanced into Asia Minor and managed to get as far as Ancyra. Zenobia’s success, nonetheless, was closely related to the disorder within the empire in that period. Rome was so occupied with troubles in the west, like barbarian invasions and internal revolts as well as economic problems and plague that the eastern part of the empire remained relatively neglected. Thus, Zenobia was left free to pursue her military enterprises, extend her realm and consolidate her authority in the region. This situation, however, changed after 270 AD when Aurelian, a higher officer in the army, became the emperor. As soon as he had sorted out the problems in the west, Aurelian marched to the east with a large army in the late 271 AD. He managed to regain control in Egypt and Syria and captured Zenobia in Palmyra while she was trying to escape from the city. Aurelian stationed a garrison at Palmyra to keep the peace before he left the city.
30
In 273 AD, a rebellion broke out in the city and the Roman garrison of 600 archers was massacred by the locals (Bryce, 2014: 314). Aurelian returned to Palmyra, murdered a large number of the rebel citizens and destroyed the city.
Palmyra’s commercial significance and prosperity ended after these events in the late 3rd c. AD. In the 4th c. AD Diocletian initiated a substantial program that aimed to strengthen the defenses along the border which included Palmyra as well. Military structures were built and a garrison was quartered in the city. The city survived primarily as a military outpost and a small center for the caravan trade, on a much-reduced scale (Intagliata, 2018: 71-81).
In 634 Palmyra was conquered by Khālid ibn al-Walid, on behalf of the first Muslim caliph Abu-Bakr. The city was later ruled respectively by the Umayyads, Abbasids, Mamluks, Ottomans, French Mandate and the current Syria Arab Republic.
2.4.1 The Population and Community of Palmyra
The earliest known inhabitants of the territory of Palmyra were the Amorites who were settled around the natural springs in the oasis. In the course of the 2nd and 1st millennium BC Arameans and Arabs came along and predominated the population of the city throughout its history. With the Seleucid colonization, new inhabitants of Greek origin and then with the Roman domination those of Latin origin, although fewer in number, came to settle in the city.
The oasis and the city of Palmyra had always been important focal points of interaction between settled and nomadic people. The settled Palmyrenes had close links with the nomadic tribes in the territory and had a certain level of influence on them (Gawlikowski, 1987: 49-55; Smith, 2013: 5). The epigraphic data dated to the first two centuries of our era from Palmyra is remarkable with the use of kinship terminology to denote both physical and social relations. In ascending order, they mention the individual, the family, the clan, and the tribe (Dirven, 1999: 22-23, Yon, 2002: 57-97).
31
A total of 16 tribes are attested in the inscriptions, four of which played a prominent role in the administration of the city at least until the 2nd c. AD. After Palmyra was transformed into a polis and adorned with the institutions of a Greek city, the tribal borders gradually blurred and lost importance. The notables from the leading families, having civic functions or not, however, continued to be influential within the political and civic sphere of the city (Smith, 2013: 15; Yon, 2002: 131-164).
In the course of the 2nd c. AD Palmyra, as a significant cultic center and as a locus for economic opportunity and an administrative center, attracted many more people (Smith, 2013: 81). Crouch estimated the population of Palmyra around 150,000-200,000 at its peak, excluding the inhabitants of the countryside (1972: 241-250). Savino, on the other hand, makes a more modest calculation and suggests a population of 40,000 to 60,000 in the city and an additional 250,000 living in the wider hinterland (1999: 69-75).
The Palmyrenes spoke a local version of Aramaic with some Arabic terms and expressions (Bryce, 2014: 279). This is the most frequently used language in the epigraphic sources of Palmyra. From the first half of the 1st c. AD on Greek appeared too, in the Palmyrene epigraphy. In a considerable number of inscriptions, Aramaic and Greek were used together. These bilingual inscriptions were preferred especially in the public or civic contexts whereas Aramaic was used alone mostly in funerary and religious contexts. The use of Latin, on the other hand, was rare in the city and the Latin documents found came mostly from the military context. A specific feature of the Palmyrene epigraphy is the trilingual inscriptions which contain Aramaic, Greek, and Latin together. Although they are very few in number, the use of Latin in these trilingual texts in the first years of the polis is considered to represent a will for integration to the Roman sphere (Yon, 2008: 196).
Bilingualism also existed in the names of some of the Palmyrenes who bore double names. Especially from the 2nd c. AD onwards, they adopted Greco-Roman names either alone or in addition to a second Semitic one. As such, the inhabitants of the city
32
of Palmyra, especially the notables, presumably wanted to emphasize their dual belonging both to their Aramaic roots and the Greco-Roman culture or they wanted to appear better integrated to the Empire (Yon, 2002: 57-97). In addition to the Aramaic and Greco-Roman names, a considerable number of Arabic and Persian names attest to the existence of these communities within the population of the city as well.
The cultural symbiosis of the Palmyrene community can be observed in the religious and funerary contexts too. More than sixty Palmyrene gods, most of which were the ancestral gods of the families or clans, are known, originating from a variety of traditions mainly from Semitic, Mesopotamian and Arab pantheons (Dirven, 1999: 28). Several temples dedicated to these gods were endowed with Greco-Roman architectural and decorative features, while the interior arrangement of the temples, as well as the cult reliefs or statues and the ceremonials, were designed and organized according to the local traditions (Gates, 2011: 401-403; Bryce, 2014: 279).
The multiculturalism in Palmyra is also reflected in the funerary art and architecture. Cultural tensions were manifested in the choice of artistic style, clothing and hairstyle of the people or gods represented and the origins and the iconography of the myths displayed (Andrade, 2013: 184). There are funerary reliefs, in which some members of the same family reclining together were dressed as Romans while others as Parthians, as well as there are other reliefs in which the same individual was dressed in a different attire while engaged in separate social roles (Smith, 2013: 17). Many funerary reliefs display deceased men and women in Greek dress, but used Aramaic inscriptions to commemorate them. The architecture of the burial structures also reflects the same diversity. While the exterior of the tower tombs could have Greek or Roman refinements, their interiors reflect the traditional burial organizations according to the extended kinship networks (Smith, 2013: 88).
The syncretism in the language, religion, art and architecture of Palmyra reflects how the population of the city emulated and manifested classical trends while retaining their tribal roots and maintaining local traditions.
33
2.5 Dura-Europos: A Fortress on the Euphrates
The ruins of Dura-Europos stand near the village of as-Salihiyeh, 92 km south-east of the city of Dayr-az-Zawr, within the borders of modern Syria. The ancient city was settled 40 m. above the river plain on the edge of a rocky plateau flanked by two deep ravines on the right bank of the Middle Euphrates. It had a naturally defensive position commanding over the river and overlooking the edge of Mesopotamia.
The evidence for the pre-Hellenistic habitation in or around Dura-Europos comes from Assyrian or Neo-Assyrian sherds and a cuneiform tablet dating to circa 1900 BC that were found during the excavations in the city (Leriche and Mahmoud, 1988: 278-279; Stephens, 1937: 183-189).
The Greek geographer Isidore of Charax, in his Parthian Stations which he wrote in the 1st c. AD, identified the city as such: ‘the city of Dura Nicanoris, founded by the Macedonians, also called by the Greeks Europus’. Dura is a Semitic word meaning fortress, and perhaps was the name of a native village on or near which the Hellenistic colony was founded (Cohen, 2006: 157). Nicanor, the nephew and general of Seleucus I Nicator, was the governor of Mesopotamia and a well-known figure in Greek colonization. According to Pliny, he founded many cities in the name of his uncle. Dura-Europos was one of these settlements which are likely to be founded in the late 4th c. BC as a military phrourion (fort) on the road between Antioch on the Orontes and Seleucia on the Tigris, the two capitals of the Seleucid Kingdom (NH. 6.117). Throughout the city’s history, even in the Roman period, Seleucus I Nicator appears to have been honored as the founding patron of Dura-Europos (Edwell, 2008: 97).
Archaeological evidence indicates that the city remained a relatively modest military occupation with a small population for almost one and a half centuries after its foundation (Edwell, 2008: 98; Leriche, 1997: 191-210). In the middle of the 2nd c. BC, however, the colony was elevated to the rank of a city witnessing a growth of population and urban development. Thanks to its strategic location on the Euphrates,
34
as Rostovtzeff indicated, Dura-Europos became an important center of control for military and commercial traffic between upper and lower Mesopotamia (1941:484). Moreover, since it was surrounded by fertile territory agriculture constituted an important source of wealth for the region. The importance of grape production and wine supply in the vicinity of Dura-Europos is indicated in a number of documents from the Parthian and Roman periods, which probably had been the case in the Hellenistic period too (Edwell, 2008: 98, 237). Thus, from the mid-2nd c. BC onwards Dura-Europos gradually flourished as an important commercial center besides its military importance.
Dura-Europos was occupied by the Parthians towards the end of the 2nd c. BC, most probably in 113 BC, as widely accepted (Welles, 1956: 469; Perkins, 1973: 5; Bellinger 1949: 200-201). Except for a brief Roman occupation under Trajan between 115 and 117 AD, the city remained under the Parthian control until 165 AD. As a seat of the local Parthian governor, Dura enjoyed relative prosperity under Parthian control, but there is only little evidence of the Parthian rule from the site. The documents demonstrate that the institutions of the Greek city continued to be used since Parthians exercised control using the existing hierarchies and civic structures. Greek continued to be used as the main language of administration and law, as well as the language of daily life. Under such a distant and loose control during the Parthian period, Edwell states, Dura-Europos flourished as a Hellenistic city in a Semitic milieu like many other cities of the Near East under Roman control (Edwell, 2008: 113).
The war between Rome and Parthia, caused mainly by the dispute over Armenia in 161 AD, had important consequences on the region including Palmyra, Osrhoene, the middle Euphrates, and Mesopotamia. As a result of the Roman campaigns against the Parthians under Lucius Verus, the kingdom of Osrhoene became a client kingdom of Rome and Rome’s power was extended down the Euphrates from the Khabur to Dura-Europos and beyond. It was in 165 AD that Dura-Europos came under Roman control and was attached to the province of Syria.
35
The textual and archaeological evidence indicates that there was not much change in the order and functioning of the city in its early years of Roman dominance (Downey, 2000: 172). In general, institutions and the pattern of life seem to have remained much as they had been earlier (Welles, Fink and Gilliam 1959: 22). The military garrison was still relatively small and at least part of it was made up of Palmyrene archers as was the case in the Parthian period.
In the late 2nd and 3rd c. AD however the territorial and military organization of Septimius Severus and his successors caused a marked change in the city’s role in the Middle Euphrates. The important outcomes of Severus’ campaigns in the 190s AD included the formation of the province of Mesopotamia, the establishment of the province of Osrhoene, and the creation of the dependent kingdom of Edessa. Important also was the division of Syria into two provinces, that of Coele Syria, of which Dura-Europos became a part, and Syria Phoenice in 194 AD. It was vitally important, for Rome, to control this recently acquired territory and the newly founded provinces and Dura-Europos undertook an important military role in the protection of this region. From the 190s onwards there was a greater and permanent military presence in the city. Millar emphasizes that the presence of this substantial Roman force, with other detachments stationed up and down the river was the first and the most obvious of the circumstances which made Dura ‘Roman’ (Millar, 1993: 467).
With the succession of the Sasanians over the Parthians in 224 AD, the eastern Roman world confronted constant pressure throughout the region. Thus, the region needed to be supported by additional troops. In the last few years of its history Dura, too, housed a large number of soldiers. However, they could not succeed to stop the enemy and the city was captured by the Sasanians circa 256/257 AD, after which date it seems to have been widely abandoned (Rostovtzeff, 1932: 156, Leriche and Al Mahmoud 1994: 417-419). In his Rerum Gestarum, Ammianus Marcellinus reported that Dura was a “deserted town” in 363 AD when the invading Roman force of Emperor Julian, and himself, passed by the site (XXIII.5.8).
36
2.5.1 The Population and Community of Dura-Europos
The total population of Dura-Europos is estimated to have been approximately 5,000-6,000 at its peak, including 1,000 to 2,000 men capable of bearing arms (Hoepfner and Schwandner, 1986: 258; Will, 1988: 315-321). The two most prominent groups within this population were Greco-Macedonians, the descendants of the original colonists, and Semitic Mesopotamians, the indigenous population of the area (Welles, 1951). The terms Europaios/Europaioi were used to describe the citizens of Dura-Europos from at least 190 BC. It is thought to have been an epithet indicating the citizenship of Dura-Europos, and prior to 180 AD, it was only used by individuals with Greco-Macedonian names (Welles, 1951: 270; Edwell, 2008: 113).
There is also a considerable amount of evidence that suggests the presence of a Palmyrene community in the city. Under Roman control, the military power at Dura was exercised by the Palmyrene auxiliaries. In the 3rd c. AD Cohors XX Palmyrenorum, which was composed of mainly Palmyrene soldiers, was probably the most significant component of the Roman garrison at the city.
The epigraphic and archaeological data from the ancient city reveal the diversity of the ethnic, cultural, and religious identities within the city’s population. The use of several languages and onomastics in the city is an important and practical indicator of ethnic plurality. Latin, Greek and Semitic languages such as Aramaic Syriac and Palmyrene were in use in Dura-Europos in both formal documents such as inscriptions and papyri and informal texts such as graffiti. Latin was mainly the language of the military context. In public use, on the other hand, Greek was common while Palmyrene was also used. Within the civilian population, both Greek and Semitic languages were used together.
By studying a group of parchments and papyri found in Dura-Europos and a corpus of documents from the Middle Euphrates, Sommer provided a quantitative approach for the distribution of onomastics in the linguistic groups (2004: 168-183). According to
37
this study, a relative majority of individuals (49%) mentioned in the Dura papyri under consideration bear Greek names while those with indigenous names come next (38%) and the ones with Latin names constitute the smallest group (13%). Sommer also examined those names in relation to the official positions they held, the results of which are worth underlining, although this study covers a larger area in the Middle Euphrates (2004: 170-172). He initially distinguished documents dating before and after 212 when all free inhabitants of the empire had become Roman citizens by the Constitutio Antoniniana.14 The study shows that all males with Latin names and a considerable number of males bearing Greek names (48%), either currently or previously held official duties, whether as political officials, soldiers or veterans, whereas a huge majority (89%) of people bearing indigenous names held no duty at all. After 212 AD, on the other hand, while the official functions were still mostly associated with the people with Greek and Latin names, the males with indigenous names with official duties increased in proportion (26%).
Besides the language use and onomastics, the excavated cult centers constitute important evidence for the multi-religious sphere in the ancient city. In addition to the Seleucid temples dedicated to the Greek gods, two of which are the temple of Artemis and the Temple of Zeus Megistos, a number of temples dedicated to deities of local Syrian origin were constructed in the Parthian period at Dura-Europos, such as the Temple of Atargatis, the Temple of Azzananthkona and the Temple of Aphlad. Most of these Seleucid and Parthian temples were also in use in the Roman period.
There is considerable evidence indicating a civil Palmyrene population in Dura-Europos, the most prominent of which comes from the Temple of Bel, also called the Temple of the Palmyrene Gods, belonging to the 1st C. AD. The remains of another smaller temple were revealed outside the city walls which was dedicated by members
14 The Constitutio Antoniniana, also called the Edict of Caracalla or the Antonine Constitution, was an edict issued on 11 July in 212 AD, by the Roman emperor Caracalla. It declared that all free men in the Roman Empire were to be given full Roman citizenship and that all free women in the empire were to be given the same rights as Roman women, with the exception of the dediticii, people who had become subject to Rome through surrender in war, and freed slaves (P. Giss. 40) (Meyer, 1910).
38
of two Palmyrene tribes in the 1st c. BC (Rostovtzeff et al. 1939: 310). The temple underwent several enlargements and renovations in the 1st and 2nd centuries AD. According to Downey, its location outside the city walls of Dura indicates that it was designed for use by the Palmyrene traders who camped outside the city and its long association with at least one of its founders may indicate the presence of a Palmyrene trading dynasty at Dura (1988b: 98).
Two other significant remains from Dura-Europos are the Synagogue and the House Church, both of which were converted from earlier houses in the 3rd c. AD. The synagogue was adorned with a series of frescoes depicting biblical scenes. Both Aramaic and Greek were used in the inscriptions and some graffiti in the building suggest that the Jewish community in the city, both as a group and as individuals, was bilingual (Millar, 1993: 471). The house church on the other hand is important as it is one of the earliest Christian buildings of the Roman world unearthed so far. The epigraphic evidence from this structure reveals that in the Christian context as in others, 3rd century Dura was the meeting point of Greek and Syriac languages (Millar, 1993: 471).
It is obvious that the Roman population in Dura-Europos was dominated by the members of the Roman army. Although the army as an institution was separate from the rest of the population both in conceptual terms and also physically within the urban layout, the epigraphic and archaeological evidence suggest that the army was not thoroughly isolated from the civil population. On the contrary, the soldiers and civilians were in contact with each other in various occasions in the course of daily life in Dura-Europos as it was the case in many cities in the eastern part of the Roman Empire (Millar, 1993: 130-131, 133; Pollard, 1996: 211-228.).
39
CHAPTER 3
ANTIOCH ON THE ORONTES:
LEISURE AND LUXURY IN THE EASTERN CAPITAL
3.1 Urban Layout and Development of a Provincial Capital
Antioch on the Orontes, one of the major cities of antiquity in terms of urban magnitude and importance, is very little known today in terms of its built environment. Modern Antakya lies over the ancient city and has been continuously expanding at the expense of its predecessor (fig. 3). The Hellenistic levels are buried several meters deep under the modern city. The duration and extent of the excavations conducted so far in the city have been limited and there is a shortage of archaeological data. The eight archaeological campaigns directed by Princeton University between 1932 and 1939, in the ancient city and its territory, had identified the remains of only a few public buildings and several residential structures that were mostly partially excavated.15 The public structures included the main colonnaded street, six baths, two theatres -one at Antioch, the other at Daphne-two hippodromes, two major churches and some tombs. More recently, parts of a bath complex, a villa and a row of shops,
15 In 1930, the High Commission of France and Syria gave the rights of excavation in Antioch to Princeton University for a six-year term which began in 1931. At the end of the same year a Committee composed of Worcester Art Museum, the Baltimore Museum of Art, the Musées Nationaux de France and Princeton University and responsible for the excavation of Antioch on the Orontes was formed. While all the members of the committee had contributed to the funding of the excavations, the directorship and publication rights of the excavations were given to Princeton University. The fieldwork began in 1932 and five campaigns were carried out both in Antioch and its suburb Daphne. In 1936 the permission of the excavations was extended for 6 more years upon the request of the Committee for the Excavation of Antioch. The new concession also gave the Committee the right to excavate in the port city Seleucia Pieria. In 1939, when Hatay was united to Turkey, the supervision of the excavations switched to the authorities of the Turkish Republic with the validity of all of the terms of the original concession. The excavations had to be finalized in 1940, with the outbreak of World War II (Kenfield, 2014: 37-77).
40
all dating to the late antique period, have been revealed during the rescue excavations conducted by Mustafa Kemal University and the Hatay Archaeology Museum.16
As opposed to the scarce archaeological data, there exists a considerable amount of literary evidence on Antioch, especially from late antiquity, which constitutes the basis of our knowledge on the ancient city. The accounts of ancient authors, especially those of the rhetorician Libanius, who lived between 314 and 392 AD, offer unique vistas of Antioch and its actors, as well as its urban layout and major monuments.
An unusual archaeological source for the urban life and topography of Antioch and Daphne is a mosaic pavement exposed in a villa in the Yakto section of Daphne. The mosaic which is dated to the mid-5th c. AD is framed by a remarkable border depicting scenes of everyday life and many public and private buildings with the name of their owners in the background (figs. 4, 5). According to some scholars the border of the mosaic depicted an itinerary in which the spectator, following the scenes in order, would make an imaginary journey through Daphne and then Antioch (Lassus, 1934: 114-156; Downey, 1961: 31).17
According to Malalas, who lived between 491 and 578 AD, Seleucus I Nicator founded the city of Antioch on the site of the earlier village of Bottia (8.200). The existence of a pre-Macedonian settlement at the site is also confirmed by pottery finds and the
16 After a long break after 1939, the systematic archaeological research in Antioch restarted in 1995 with three projects. The first project conducted archaeological surface surveys in Antioch, the Amuq Valley, Seleucia Pieria and the Qusayr Plateau. Called the Amuq Valley Regional Projects, this survey has been carried out by K. Aslıhan Yener from the University of Chicago, and aimed “to understand the urban topography of Antioch and Seleucia Pieria, to document and analyze the remains and to follow the rural settlement pattern and distribution in the area” (Pamir, 2014: 84). The other two projects conducted the rescue excavations and also the excavation of the hippodrome area, which have been conducted under the directorship of Hatice Pamir from Mustafa Kemal University, in collaboration with the Hatay Archaeological Museum. The aim of the hippodrome excavations is stated by Pamir as: “to unearth the hippodrome and the buildings around it, as well as the buildings that were discovered in previous excavations but are currently buried and in order to implement landscaping and restoration projects” (Pamir, 2014: 114).
17 Levi on the other hand, argued that the buildings shown in the border are not orderly depicted and there is no certainty that the mosaic depicted buildings outside Daphne (1947: 323-345).
41
discovery of an earlier beaten earth road beneath the Hellenistic levels of the main street (Lassus, 1972: 140). Lassus claimed that this road was part of a major commercial route of antiquity through which Asia was connected to the Mediterranean (1972: 140). The Seleucid settlement was initially laid out on the flat land between this road and the Orontes River and covered an estimated area of about 100 to 110 ha that was enclosed by walls (fig. 6) (Cabouret, 1999: 134). The city was designed in the Hippodamian grid plan with the insulae measuring about 112.00 m x 58.00 m and oriented according to the north-east-south-west axis of the main road in the east.
When Seleucus II (246-226 BC) made Antioch the principal seat of Seleucid power he started the construction of a new urban quarter on the crescent-shaped island on the Orontes River, in order to accommodate the increasing population (Najbjerg and Moss, 2014: 26; Downey, 1961: 99). The construction was completed by his son Antiochus III (223-187 BC). The new quarter was laid out on a grid plan that was oriented to the cardinal points, and enclosed by a wall. It was connected to the city center with several bridges (Najbjerg and Moss, 2014: 26).
The urban layout of the city changed further in the next century, in the reign of Antiochus IV Epiphanes (175-163 BC) who, as a great benefactor, elevated Antioch to one of the most magnificent cities of antiquity (Downey, 1961: 95 -107). He added a new quarter to the east of the city, on the lower slopes of Mount Silpius which was named Epiphania after him. The new section of the city included a new agora, a bouleuterion, and an imposing temple of Jupiter. With the extension of the city towards the east, the early road that had marked the eastern limit of the earlier settlement became a main axis within the city and gained a more monumental character. In the second half of the 2nd c. BC, the width of the road was 7.20 m on average while the sidewalks were 4.30 m wide (Lassus, 1934: 142). The shops that were aligned along the sidewalks on both sides had a depth of 6.80 m (Cabauret, 1999: 128).
After the annexation by Pompey, as the capital city of the province of Syria, Antioch surpassed its former grandeur and reputation. With the ambitious building programs
42
of the Roman emperors and the leading citizens, the city expanded, changed in appearance, and became one of the most glittering cities of the empire.
The silhouette of Roman Antioch was defined by a circuit of walls and towers which ascended towards the slopes of Mount Silpius (fig. 7). New fortifications were added to the Seleucid walls in order to include the newly expanded urban areas on the slopes of Silpius to the east of the Seleucid city. There were at least four gates leading to the major avenues of the city by the 6th century AD (Downey, 1961: 612 -620) (fig. 8).
One of the most important urban features of Roman Antioch was the monumental colonnaded street that gave the city its grandeur. The main street of the Seleucid city saw several modifications and rebuildings from the very early years of Roman rule. In the first half of the 2nd c. AD, after the devastating earthquake of 115 AD, the street became a majestic assembly with the addition of colonnades and high-quality polygonal stone pavement. The total width of the road reached 41 m, including 8 to 9 m wide vast porticos on both sides. The columns were of red and gray granite with a diameter ranging from 0.58 m to 0.65 m and rising to 6 m in height (Lassus, 1972: 146). Approximately 3 km long, the road crossed the heart of Antioch and connected the Aleppo gate in the north and the Daphne Gate in the south. Large open plazas were situated at the major intersections with secondary streets. Libanius described the streets, porticos, and the stores of Antioch as such:
The stoas have the appearance of rivers which flow for the greatest distance through the city, while the side streets seem like canals drawn from them. Some of the side streets, which face toward the mountain, lead to the charms of the slopes; the others, which face the other way, lead to another street, which is un-roofed but built upon on both sides, just like canals made for the cross passage from river to river. The former section ends in many places among charming gardens; the latter streets come to an end at the bank of the River Orontes. About in the middle of the right hand stoa of those which, as I have said, are stretched out from the east to the west and extend over as much length as would be sufficient for three cities, apses facing in all directions, with one roof of stone, form, as the beginning for other stoas which run toward the north as far as the river, the shrine of the Nymphs, which stands about the stoas, high as heaven and turning every eye with the dazzling light of its stones and the color of its columns and the gleam
43
of its pictures and the wealth of its flowing waters. Side streets begin from these stoas just as from those [of the main street] which I first described. (Or. XI.201-202).18
The monumental colonnaded street of Antioch, and of the other eastern Greco-Roman cities, not only served as an armature that connected and organized the interaction of public buildings and spaces, but also provided a setting for different types and purposes of social interaction. The covered wide portico and the shops behind created a broad thoroughfare along the main avenue of Antioch. Libanius mentions a dense circulation under the porticos where there were stalls between the columns, in addition to the shops and workshops behind (Or. XI.254). The mosaic of Yakto, respectively, depicts the Antiochians wandering around, offering their products or sitting at gaming tables under the porticos (fig. 9).19 Libanius stated his opinions on the role of the porticoed streets and the privilege of dwelling near these streets as such:
What then is my purpose in this? And the lengthening of my discourse, entirely about the stoas, to what will it bring us? It seems to me that one of the most pleasing things in cities, and I might add one of the most useful, is meetings and mixings with other people. That is indeed a city, where there is much of this. Truly, it is good to speak, and to hear is better and to converse is best, and to add what is fitting to the fortunes of one's friend's, rejoicing with them in some things, sorrowing with them in others, and to have the same return from them; and in addition to these there are ten thousand things in being near to one another. People who do not have stoas standing thus before their houses are scattered by the winter; and although they can be said to live in one city, they are actually separated one from another not less than those who live in different cities, and they learn news of those who dwell near them as they would of those who are living abroad… Thus, the stoas do not contribute to pleasure any more than they do to those things which are of the greatest importance among men; and to these stoas are added the hippodrome and the theatre and the bath…. (Or. XI. 215-218).
18 Translations of Libanius used in this thesis are by Downey (1959).
19 In his study “Dans les rues d’Antioche”, Lassus (1935) matched the scenes on the mosaic with the snapshots he took from the streets of modern Antakya (fig. 9).
44
Libanius also informed us about the continuity of the crowd and vividness in the city throughout the evenings saying that “…night differs from the day only in the kind of the light …” and elaborated on this vivacity by saying that “Who, seeing the city for the first time, would not think he had come to a festival?” (Or. XI.266-267).
In addition to the majestic colonnaded street, Antioch was endowed by several other large-scale public buildings like theatres, baths, and aqueducts throughout Roman rule. A palace, which presumably replaced the earlier Seleucid royal residence, and a hippodrome were built on the island on the Orontes River in the first years of the Roman reign (Poccardi, 1994: 993). The island was rearranged in the 3rd c. AD mainly by the emperor Diocletian who after bringing political and economic stability to the empire, initiated an extensive building program. At his time, the existing palace was rebuilt and enlarged, the hippodrome was restored, several new bath buildings were constructed and the island took the name of “New City”. A second exercising ground that was named the “Byzantine Stadium” by the excavators, was built in a later period (Campbell, 1934: 33). The “New City” was connected to the mainland by five bridges that led to different streets (Poccardi, 1994) (fig. 6).
The total surface area of the urban Antioch reached approximately 500 ha by the time of Justinian, who ruled between 527 and 565 AD (Cohen, 2006: 84). The urban development under Byzantine rule was also accompanied by an expansion of its suburbs and hinterlands. For example, there had been a sharp increase in the rural settlements with the emergence of a network of lesser towns in the first two centuries AD (De Giorgi, 2007).
The neighboring town of Daphne shows an urban development chronologically parallel to that of Antioch. Just a short distance from Antioch, about 8 km to the south of the city along the river, and situated on a plateau of cool forests filled with falls and springs, Daphne provided the wealthy Antiochenes an attractive summer retreat from the heat of Antioch. It began to develop as a luxurious suburb of the city, in the reign of Seleucus but became a regular town in the Hellenistic or Early Roman period. In
45
the Roman and late Roman periods, the city continued to prosper with monumental public buildings and lavish residences. The only major public building excavated so far at Daphne, however, is the theatre (Wilber, 1938: 57-94).
The freedom of religion issue by Constantine marked a significant threshold both for civic life and the built environment of Antioch. According to the literary sources Constantine ordered the construction of the octagonal Great Church on the island on the Orontes which has not yet been spotted in the excavations.20 Only two major churches have been excavated in Antioch so far but it is likely that as a prominent center of Christianity, the urban fabric of the city in late antiquity was dominated by several religious structures.
The urban prosperity of Antioch continued well until the 6th c. AD when the city suffered a series of devastating catastrophes. The tremendous earthquakes in 526 and 528 destroyed the city almost entirely. Despite the efforts of Justinian to rebuild the city, Antioch never recovered (Najbjerg and Moss, 2014: 34).
3.2 Domestic Architecture in Antioch on the Orontes
The archaeological excavations conducted between 1932 and 1939 by Princeton University covered several areas in Antioch on the Orontes and its vicinity, including the suburb of Daphne, the port city of Seleucia Pieria, and a few other isolated sites. During these expeditions, evidence was found for some eighty buildings, most of which are identified as private residences. Among the spectacular finds are nearly three hundred impressive mosaics pavements (Kenfield, 2014: 76).
Most of the residential structures unearthed so far in the city are either poorly preserved or only partially excavated. The successive earthquakes in the region and the landslides
20 For detailed information and interpretation of the literary sources on the construction of the Great Church see Downey, 1961: 34-350.
46
of the mountain in antiquity led to subsequent reconstructions and superimpositions which destroyed the earlier levels of the structures. The walls of the standing structures have often been torn off to their foundations to provide building material for new constructions. On the other hand, the time pressure and other restrictions prevented the excavators to unearth the structures completely. In Daphne, especially, where there were extensive orchards and olive groves, many mosaics, some very close to the surface, were spotted by the owners of the modern houses or by the excavators who did soundings.21 In some cases, it was necessary to buy the land to clear the ground and cut down the trees but the peasants hardly wanted to sell their property (fig. 10) (Lassus: 1984: 361). Due to such local dynamics only a few houses were dug in their entirety. In the majority of cases, only the nucleus of the houses including the principal rooms with mosaic pavements were uncovered. The published plans of the houses in Antioch, thus, are mostly incomplete and the circulation patterns have been poorly defined and sometimes even remained hypothetical (Lassus, 1984: 361).
In addition to the archaeological data, the literary depictions of Antioch on the Orontes in late antiquity have contributed, often indirectly, to our knowledge of the domestic life and private space in the city. Libanius, a representative of the last epoch of paganism and affiliated to the Classical thought that exalted public life, mentioned and appraised private dwellings in relation to the urban space and city life. The Christian preachers who adapted asceticism and charity as a recurring theme, on the other hand, took an opposite position and criticized the ostentation and luxury in both the public and domestic spheres (Alpi, 2007: 45-48).
In his oration praising Antioch, Libanius portrayed the imperial palace in the “New City”, for which no archaeological data has been found in the excavations:
21 Mosaics that were discovered by chance by the local citizens were so many in number that the excavators recruited a special mosaic crew in 1934. This team was responsible for excavating, recording and in some cases raising the well-preserved portions of the mosaics. Soon, the excavations in Antioch became the excavation of mosaics. The mosaics and finds brought to light during the excavations between 1932 and 1939 were shared between the local authorities and the excavating and funding institutions. Today, they are preserved and exhibited in more than 20 collections in different countries (Barsanti, 2012).
47
This palace occupies so much of the island that it constitutes a fourth part of the whole. It reaches to the middle of the island, which we have called an omphalos, and extends to the outer branch of the river, so that where the wall has columns instead of battlements, there is a view worthy of the emperor, with the river flowing below and the suburbs feasting the eyes on all sides. A person who wished to describe this part carefully would have to make it the subject of a discourse, but it cannot be a part of a discourse on another subject. Nevertheless, one should say at least that to the other palaces which exist in every part of the world, some of which are praised for their size and others for their beauty, it is in no way inferior; but it is far superior to many, nowhere surpassed in point of beauty, and in size surpassing all others, divided into so many chambers and stoas and halls that even those who are well accustomed to it become lost as they go from door to door. I believe that, if this palace stood by itself in some insignificant city, such as are numerous in Thrace, where a few huts form the cities, it would give the one that possessed it good reason to claim a proud position in the catalogue of cities (Or. XI. 206-207).
Libanius informs us about the distribution of private residences within the urban fabric too. He talks about houses mingled with public structures in the heart of the city, some just behind the colonnaded streets, and others located on the lower slopes of the hills that surrounded the settlement:
As you go through these stoas, private houses are numerous, but everywhere public buildings find a place among private ones, both temples and baths, at such a distance from each other that each section of the city has them near at hand for use, and all of them have their entrances on the stoas (Or. XI. 212).
The mountain rises up, stretched out besides the city like a shield raised high in defense, and the last dwellers upon the lower slopes of the mountain have nothing to fear from the heights, but they have the sources of every happiness, springs, plants, gardens, breezes, flowers, the songs of birds, and the enjoyment of spring earlier than the others have it (Or. XI. 200).
So far, no domestic structure in the close vicinity of the colonnaded street of Antioch has been uncovered. On the other hand, the large number of houses that are spotted and excavated on the slopes of Silpius and Staurin mountains confirm Libanius’ description. The excavated residential structures were scattered and not a single insula of houses has been uncovered as a whole yet. Thus, the existence of a grid system in the higher and peripheral neighborhoods of the city has not been attested.
48
Most of the houses and mosaics revealed in the vicinity of Antioch came from Daphne, the favored suburb of the city, whose beautiful houses surrounded by trees and gardens were also praised by Libanius:
As soon as you pass through the gates, on the left are varied gardens and charming inns and an abundance of springs and houses hidden in trees and chambers which rise above the groves and luxurious baths, a place worthy of Aphrodite and her son the archer. As you go on you see on both sides of the road a wealth of vineyards and beautiful houses and rose gardens and plants of all sorts and streams; one thing draws you to it and another draws you away again, and it is through such pleasures that you come to the supremely beautiful Daphne (Or. XI. 234).
The insula system is detected only in a portion of the unearthed area in Daphne, where there are traces of a grid, about 30 m x 60, laid out in the north-east south-west direction (fig. 11). The orientation of later buildings, however, did not adhere to that pattern. Thus, Lassus argued that the precedence of the winding streets and the large areas devoted to gardens or orchards prevented the implementation of a regular plan at Daphne (1984: 57).
The houses excavated so far at Antioch and its vicinity are dated to a period between the early 2nd and the 6th centuries AD. Many of them were rebuilt and/or redecorated in the course of their occupancy, sometimes without paying attention to the earlier layout. In the following part of this section, the relatively well preserved and/or more extensively uncovered houses of Antioch are examined in chronological order.
3.2.1 The Atrium House (The Roman Villa)
The earliest dated residential building excavated at Antioch is the so-called Atrium House, located on the island, on the Orontes River (fig. 12). It was partly revealed in 1932, during the first season of the Princeton excavations, and was initially called the “Roman Villa” (Fisher, 1934: 13-18). In later publications, it has been inaccurately
49
named as the “Atrium House”, although it contains no space that can be identified as a Roman atrium that is exemplified in Pompeian houses.
The construction of this residential building is dated to the Augustan period and three main occupational phases have been identified until its abandonment (Levi, 1947: 15; Fisher, 1934: 8-18). There is very little evidence of the initial structure since it was highly destroyed by later modifications. It was a vast assembly of at least 1,500 m2 that was laid out around a colonnaded courtyard. The limits of the house and the entrance could not be spotted during the excavations. The first phase of occupation continued until the early 2nd c. AD, at the latest, when the structure was partly demolished, probably by the earthquakes of 94 AD and/or 115 AD (Fisher, 1934: 18).
In the second phase, the dwelling was modified and evolved into a more monumental layout. The courtyard was reduced in its surface area with the addition of new rooms, including room 86, and was reorganized. The columns were reused and replaced, with respect to the new layout of the surrounding spaces. Remains of five columns on the north side and four on the east side, with different spans, were found in-situ. The southern and western limits of the courtyard were destroyed by later constructions but the foundation wall below the ground level on the west side indicates the existence of a colonnade on this side too (Fisher, 1934: 15). A number of coarse pieces of red-painted stucco from the columns were found on the floor, on the east side of the central courtyard. These were smaller in diameter than the columns of the peristyle and may have belonged to an upper level (Fisher, 1934: 15). The northern portico of the courtyard was deeper than the eastern one, 2.60 m against 1.60 m, indicating the importance of the units behind. The portico seems to have served as a lavish vestibule for the reception rooms behind.
The northern wing of the house had three larger rooms and presumably comprised the ceremonial section of the dwelling. These rooms were differentiated not only by their size and layout but also by the good quality masonry of their walls. Room 130 was probably the main reception room of the house in the first phase of occupation and, according to Fisher it may have been centered at the axis of the courtyard in the earlier
50
layout (1934: 16). It was paved with a red stucco floor. During the reconstruction works in the 2nd c. AD, this room was narrowed to provide a spatial expansion to the adjacent room 82. In the new layout, room 130 opened to the courtyard via a wide door and communicated with another large room, room 85, to its east.
In the 2nd c. AD room 82 was reshaped and enlarged without respect to its former layout. Its floor was raised and received a sumptuous mosaic pavement. With these modifications, room 82 became the largest and finest unit of the house with a surface area of about 50 m2. The entrance of the room was highlighted by the corresponding two columns on the western end of the northern portico, which offered a wider opening than the others, and by the figural scenes of the mosaic pavement on the floor. It was entered via a step since its floor was approximately 0.30 m higher than the floor of the courtyard. The mosaic pavement on the floor, measuring about 7.20 m by 4.80 m, had a fine quality workmanship. It contained five figural panels arranged in T shape and surrounded by geometric motifs, indicating the use of the room as a triclinium (fig. 13). The three panels on the crossbar of the T, depicting the drinking contest between Herakles and Dionysus in the center and depicting a dancing satyr and a maenad on either side, faced the courtyard and addressed the entering guests, while the remaining two placed at the foot of the T and depicting Judgement of Paris and Aphrodite and Adonis, were oriented towards the back of the room for the view of the dining guests. Room 82 of the Atrium House is noteworthy as one of the earliest triclinia that were unearthed in the eastern territories of the Roman Empire.
Remains of a stone mouth to join pipes at the south-western corner of the courtyard, roughly aligned to the south wall of rooms 101 and 106, indicated the existence of a water feature in this spot. This was probably a decorative fountain placed on the visual axis of the triclinium since the availability of water supply was already assured by a well in the same courtyard. The coexistence of triclinium, portico, and an ornamental fountain, though not necessarily located on the same axis but often was the case, has been attested in most of the domestic structures revealed in Antioch so far.
51
On the northern limit of the excavated zone, which was probably the northern limit of the house itself too, was a large open space. This space was connected to the main courtyard by a narrow corridor (room 93) and presumably communicated with the reception rooms. It may have served either as a ceremonial courtyard or may have been used for domestic activities together with the row of rooms partially revealed on the eastern part of the northern wing.
The eastern end of the north portico of the colonnaded courtyard gave way to a series of rooms that were only partially unearthed. To the east of the courtyard was room 90, an elongated rectangular room that recalls the elongated apsidal rooms that were formed by blocking the colonnade of one of the porticos in the peristyle courtyards of late Roman period houses in Apamea on the Orontes. The function of these rooms could not be identified.
The second phase of occupation in the Atrium House continued until the late 3rd c. AD. After this period, the structure lost its grandeur and unity and was subdivided into at least three smaller apartments (Ellis, 2004b: 130). Larger spaces like the triclinium, northern courtyard, and the porticos of the colonnaded courtyard were subdivided into smaller units by poor-quality walls with reused material. Some doors were blocked and some spaces were totally abandoned. In a later phase, the Atrium House remained beneath a bath building, named Bath B, and a series of small houses.
3.2.2 The House of the Calendar and the House of Cilicia
The House of the Calendar and the House of Cilicia are two partially excavated domestic structures of which only a few rooms were revealed. These rooms are insufficient to understand the overall architectural layout of the houses but they are noteworthy as they present the earliest examples of the common layout that links the triclinium, portico, and nymphaeum, found in most of the excavated houses in Antioch.
52
The House of the Calendar was located on the higher slopes of Mount Silpius, adjacent to another poorly preserved house, the House of Drunken Dionysus, which has a similar plan with a different orientation. The houses are dated to the 2nd c. AD, a little after the earthquake of 115 AD (fig. 14) (Levi, 1947: 36-40; Stillwell, 1961: 49). The House of the Calendar had a large reception room, apparently a triclinium as indicated by the T-shape mosaic pavement. This room opened to a corridor-like portico to the south-west through a triple opening. Across the portico, a rectangular nymphaeum with apsidal niches on its back was placed behind a screen of three columns (fig. 15). The length and axis of the nymphaeum matched more or less those of the room itself. Beyond the nymphaeum was an unexplored open space which might have been a paved courtyard or a planted garden.
The T-shaped mosaic pavement of the triclinium was composed of two main panels, a square and a rectangular one that were surrounded by a geometric border with lozenge and diamond-shaped panels. The square panel at the back of the room comprised a circle with twelve radiating segments each of which represented allegories of the months of the year. The rectangular one, depicting Oceanus and Thetis enthroned on rocks and surrounded by marine creatures, and facing the entrance greeted the entrants. The mosaic pavement of the corridor was divided into three panels. The central panel, between two geometric ones, figured a black macrophallic fisherman moving with baskets on his shoulder (fig. 16). This mosaic faced toward the triclinium for the view of the diners.
The House of Cilicia, found in the port city of Seleucia Pieria, presents a very similar triclinium, portico, and nymphaeum layout (figs. 17, 18). The triclinium of the house, about 9.50 m X 9.50 m in size, is one of the largest rooms excavated so far in the city. In accordance with the square shape of the room, there was a square panel on the mosaic floor, instead of a T-shaped scheme that was used in the House of Atrium and the House of Calendar. The main panel of the poorly preserved mosaic decoration depicted personifications of two provinces, Cilicia and probably Mesopotamia. This panel faced the diners. On the four corners of the main panel smaller square panels
53
depicting the busts of the river gods were placed radially looking outwards, and between them were three rectangular panels with geometric motifs. On the floor of the room, a little west of the center of the main panel was a small, shallow water basin that was sunk in the floor.
The triclinium of the House of Cilicia opened to a portico that was longer than the room and extended more towards the north. On the other side of the portico, a small court with a nymphaeum at its back was located non-axially with the triclinium. The portico and the court were paved with geometric mosaics whose patterns were dated to the 3rd c. AD (Stillwell, 1941: 5). At a later date, a pool with three niches at its back was superimposed on the earlier one. The portico gave way to other unexcavated rooms to the north and a possible courtyard to the west behind the nymphaeum (Stillwell, 1961: 49).
3.2.3 The House of the Drinking Contest
The House of the Drinking Contest, also found at Seleucia Pieria is dated to the Severan period, 193-235 AD, based on the style of the mosaic pavements (Stillwell, 1961: 50). With several rooms unearthed, it displayed a more complete architectural context especially for the layout of the triclinium and its dependencies (figs. 19, 20).
The central unit of the House of the Drinking Contest was composed of a narrow courtyard with a nymphaeum and two colonnaded porticos that ran along its east and north sides. The rooms were organized around these units which also served as the main medium of circulation between the different parts.
There were two entrances to the house, one from the west and the other from the north. The western one was probably a service entrance and gave way to two possible service rooms and to the north portico. The latter one opened to the northern end of the eastern portico through a small space like a vestibule.
54
At the eastern end of the house was situated the triclinium, the largest unit of the structure, measuring 7.10 m in depth and 8.20 m in width. It was decorated with a luxurious and richly colored mosaic pavement depicting the mythological scene of the drinking contest between Heracles and Dionysus in an architectural setting with a three-dimensional effect (fig. 21).22 The figurative panel was bordered by a continuous geometric pattern in a U-shaped arrangement and faced towards the back of the room where the most honored guests should have been reclined (Dunbabin, 1999: 161; Malmberg, 2005). There was no panel facing the entrance on the mosaic, most probably because of the proportions of the room whose width is broader than its depth.
The triclinium opened to the eastern portico through a large triple opening. The portico itself was paved with a mosaic too, on which a figurative panel was laid out between two geometric panels. The central panel, representing Eros and Psyche, was oriented to the triclinium (fig. 22). Beyond the portico, on the sightline of the triclinium was the courtyard, at the back of which was placed a small decorative nymphaeum that had already replaced an earlier one. The courtyard was decorated with a mosaic carpet that had a triple arrangement. The central panel figured a marine scene in which erotes were fishing on the backs of dolphins and were surrounded by fish. The representation had no single direction and could be viewed from all sides as one walks in the courtyard. Since the southern wall of the courtyard had been completely robbed out, it is not possible to know whether the courtyard communicated with the outside in any means. Dobbins recalled the possibility of the existence of a low terrace wall or a wall with large windows and shutters for the south wall of the courtyard which would have provided a view of the landscape both for the triclinium and the northern rooms (fig. 23) (2000: 54).
The northern portico was aligned with three rooms, two of which were decorated with geometric mosaic pavements. The portico itself had a mosaic floor with square panels,
22 Dunbabain stated that the use of dramatic color contrasts for exaggerated light and shadow effects and a taste for three-dimensional effects and architectural decoration were common in the ornamental representations of the late 2nd and early 3rd centuries, throughout the Roman world (1999: 163).
55
the four easternmost of which were preserved and had figural representations of seasons in the personality of erotes. The figures were oriented towards the northern rooms. The four columns on the northern portico were not evenly spaced, probably not to obstruct the view from the rooms towards the courtyard.
To the west of the courtyard were four rooms without mosaic pavements some of which must have been used as service spaces. To the north of the triclinium, there were other rooms but they remain unexcavated.
3.2.4 The House of Dionysus and Ariadne
The House of Dionysus and Ariadne was located in a residential zone on the slopes of Musa Dağ at Seleucia Pieria. There is very little information on the chronology of the building. The mosaic pavements of the house are dated to the Severan period based on their style, just like the one in the House of Drinking Contest located at about 1.00 km east (Stillwell, 1961: 51). According to the archaeological records the quarter was abandoned in the late 2nd or 3rd c. AD, probably after the completion of the Titus tunnel which prevented the mountain torrents and enabled the residents to move down to the harbor area (Stillwell, 1941: 4; Levi, 1947: 141).
Built on the slopes of the mountain, the plan of the House of Dionysus and Ariadne was adapted to the steep topography and organized at different levels (fig. 24). It was bordered by narrow paved roads on its east and north that descended from the slopes of the hill with steps. The southern part of the structure had totally disappeared due to landslides. In the excavation reports and the published plan, a third street was suggested on the southern limit of the house (Stillwell, 1941: 4; Levi, 1947; fig. 56). Nevertheless, a portico opening to a street is not likely, and relying on the repeated pattern in the region, this part of the structure may have comprised a courtyard that commanded the panoramic view of the harbor (fig. 25).
56
The relatively modest house presented a compact layout covering an area of about 200 m2. Two entrances were identified on the eastern street in the restored plan. Stillwell, who described the house in the excavation reports, later argued that these entrances were improbable either in number or in position and the plan should be slightly corrected (1941: 4; 1961: 51). The placement and layout of steps in the northern vestibule do not seem plausible either.
Assuming that the circulation patterns displayed on the published plan are correct, the south-eastern vestibule of the house gave way to a large space that might have been a service room or a kitchen. Through this space and an adjacent corridor, the main rooms of the house were accessed. The second entrance door on the north-east corner gave way to a row of three rooms at the northern limit of the house, through which the reception rooms were reached. The very indirect and complicated path to the reception rooms, passing through all other rooms of the house, seems an unlikely situation and hence there might have been an additional entrance door on the destroyed southern part of the house, or else there might have been an unspotted passage between the south-eastern vestibule and the adjacent portico.
Two interconnected rooms and a large portico, all with sumptuous mosaic pavements seem to have constituted the reception suite of the house. Both rooms were identified as triclinia by Stillwell, while Levi identified only the larger one, despite the fact that none of them presented the common T/U or U-shaped mosaic pattern (Stillwell, 1941: 4; 1961: 51; Levi, 1947: 141). The pavement of the larger room had a mosaic panel with figural scenes oriented towards the back of the room where geometric patterns were used (fig. 26). The mosaic panel after which the house was named represented an architectural setting rendered in perspective in which Dionysus awakening the sleeping Ariadne accompanied by Eros, Silenos and a maenad were depicted. This room was annexed by a smaller room with a more damaged mosaic panel displaying another architectural composition that shows a temple façade. Within the temple a seated female figure holding a woolen distaff was represented. In front of these two rooms extended a spacious colonnaded portico with a series of figurative mosaic
57
panels (fig. 27). The panels of the portico were oriented towards the rooms. The entrance of the larger room from this portico was accentuated by two columns.
At the northern part of the house, at a considerably higher level than the southern part, were a row of three interconnected rooms. Besides the entrance door on that part of the house, these rooms were accessed through the service corridor mentioned above and from the reception suite (fig. 28). The larger reception room was connected to the space on its north via a wide opening and several steps.
Levi argued that the House of Dionysus and Ariadne provided one of the most complete plans of the excavated houses in the territory of Antioch (1947: 141). Nevertheless, considering the existence of very sumptuous mosaic pavements and the very modest dimensions of the house as well as the unconvincing circulation patterns, it is not unlikely that the house might have extended further or more probably had an upper story.
3.2.5 The House of the Buffet Supper and The House with Floral Pavements (Building B)
The House of the Buffet Supper, a 3rd c. AD house in Antioch’s luxurious suburb of Daphne, was built in an insula within a quadrangular plan which had been laid out in the Hellenistic or early Roman period (Stillwell, 1941: 28). The north-west and south-west limits of the insula were determined by two perpendicular streets; a wide one of about 10.00 m to the north-west and a narrow one of about 2.50 m to the south-west. The location of the entrance(s) of the house could not be specified during the excavations.
The architectural remains brought to light in this building plot were concentrated on three separate points around a large open space (fig.29). There is no conclusive evidence to specify whether these remains belonged to a single large residence of at least 2,500 m2 or to different structures. They had slightly different chronologies.
58
Morvillez interpreted the complex as reminiscent of a system of pavilions associated with green areas and water features (2004: 275). The vast space separating, and perhaps also connecting, three sets of architectural units have not yielded any trace of construction. Therefore, it has been assumed that it was a large and either planted or paved open space (Stillwell, 1941: 29; Levi, 1947: 127).
At the south-east end of the excavated area of the insula were three adjacent rooms that opened to a portico (fig. 30). The central room C2 was the largest one, measuring 6.50 m by 5.90 m, with an apse at its end. The semi-circular pattern of the mosaic floor followed the apsidal form of the room and depicted a very large sigma table, set with refined dishes which gave the house its name (fig. 31). In the center of the table was a circular medallion depicting Ganymede and an eagle. There was also a rectangular panel with representations of erotes, birds, and a crater, in front of the entrance. All the figurative panels were oriented towards the back of the room to be seen by the diners. This room, which had functioned most probably as a dining room, was enlarged and became rectangular with the omission of the apse, probably at the end of the 3rd c. AD (Morvillez, 2004: 277). A new rectangular geometric mosaic panel was added to the floor of the enlarged section of the room. The restoration in the dining room is a noteworthy one, as it had been done in a period when sigma tables, stibadium, and the apsidal rooms became common throughout the empire.23
The dining room was flanked by and communicated with two smaller square rooms which were almost identical in size and layout. Their mosaic floors had similar arrangements with a figurative panel placed in front of the entrance and a geometric panel placed at the back of the room. The mosaic panel of room C3 represented Narcissus and Echo while that of room C1 was totally destroyed (fig. 32). These rooms were interpreted either as secondary reception or dining rooms (Stillwell, 1961: 51); or more intimate rooms such as bedrooms where the geometric panels would have
23 For the replacement of the triconch with stibadium and sigma tables, and the design of the apsidal dining rooms in the late Roman period see Bek, 1983, Dunbabain, 1991, Ellis, 1997, Malmberg, 2012.
59
indicated the place of beds (Lassus, 1984: 366; Morvillez, 2004: 275). Considering the layout of the mosaic panels which were different from the more traditional T pattern, and the depicted scenes that were not related to the consumption of wine or food, it is likely that these spaces did not specifically function as dining rooms.
The central dining room and the two annex rooms opened to a portico via narrow doors of approximately 1.00 m wide that could not provide a grand view to the portico and the garden behind. The portico received a mosaic floor in the same period of the restoration of the apsidal room, in the late 3rd c. AD. The mosaic had geometric panels alternating with figurative panels, representing a couple of characters, placed on the axis of the three rooms (fig. 33).24 These panels were not oriented towards the rooms but towards the courtyard. At the back of the portico, on the axis of the dining room was a two-sided nymphaeum that was decorated with niches. On the garden side of the nymphaeum was a basin to collect water.
A separate set of rooms, called C300 by the excavators, was uncovered in the eastern part of the same insula. There is no data to link these rooms with the House of the Buffet Supper or the third complex called C100. Archaeological material and the style of the mosaic pavements suggest a later date in the 3rd c. AD, corresponding to the restoration of the apsidal room and the laying of the mosaic of the portico (Stillwell, 1941: 29).
The excavated and relatively well-preserved spaces of the C300 complex seem to constitute a smaller scale reception suite (fig. 34). An open courtyard, room 21, provided light and air to the spaces behind. The courtyard was paved with a geometric mosaic and had a basin at its south-east side, which is supposed to be a part of a decorative fountain (Martz, 2011: 60). The courtyard was followed by a deep portico
24 Doro Levi interpreted the figures on the mosaic panels of the portico as the banqueters and the heterae. Considering this theme and the number of dining rooms in this part of the house, he and Stillwell suggested the possibility that this building could have been used for non-domestic purposes (Levi, 1947: 217; Stillwell, 1961: 51).
60
with two columns (room 19). On the mosaic floor of the portico, a geometric pattern framed a central rectangular panel which was lost. Behind the portico was placed the square room 17, 12 m2 of which had a good quality mosaic pavement. On the pavement, a geometric panel was framed by a figurative border depicting birds. Room 17 opened to the smaller room 16 via a wide opening. This room was higher than the previous one and a column base found in-situ against the south wall indicates a monumental entrance. It had a lavish geometric mosaic representing cubes in perspective. A small pool was also noted to the east of these rooms (Stillwell, 1941: 29).
While the courtyard, portico, and the reception room sequence of the C300 complex repeats the common pattern seen in many of the Antiochene houses it differs in some aspects. First of all, considering the room layout and the pattern of the mosaic floor, the main room behind the portico was apparently not a triclinium. It opened fully to an elevated and lavishly decorated alcove which was resembled to an iwan by Stillwell and Martz (Stillwell, 1941: 29; Martz, 2011: 70, 77). The lack of figurative scenes depicting Greco-Roman themes in this part of the building is also striking and, taken together with the iwan-like space, indicate an eastern influence. In this layout, rooms 16 and 17, probably with the other unexcavated annex rooms, seem to have constituted a reception suite used for some purpose other than dining. Room 16 could have been the place where the owner of the house enthroned and received his guests who were waiting in room 17.
The third set of rooms with mosaic pavements occupied the northern part of the excavated plot. Among these poorly preserved rooms, one looks prominent because of its size and decoration. It was a large room with a good quality figurative mosaic panel showing two standing figures in an architectural setting. The size, design, and good workmanship of this fragmentary mosaic floor indicate the importance of the room. To the east of this room was a large basin of at least 5.00 m wide and 1.00 m deep.
61
The architectural layout and the elaborate decoration of the uncovered spaces as well as the several water elements indicate a wealthy occupation in this insula, with either a single residence or a complex of separate buildings, which remained in use in the 3rd and 4th centuries AD. Then, after a period of abandonment in the early 5th c. AD, the boundaries of the insula were redetermined and a large part of it was occupied by a new dwelling named Building B or House with Floral Pavements, after its mosaic floor patterns.
Probably around 420-430 AD the principal street measuring 10.00 m wide, to the north-west of the insula, was reduced to 5.00 m and the remaining 5.00 m was incorporated to the adjacent property (Morvillez, 2007: 75). In almost the same period a new house with a totally different layout and orientation was constructed in the insula (fig. 35). This house was provided with a new access from a small street established to its north.25
The walls of the new structure were not superimposed to the precedent walls, except the ones on the south-west façade, but were constructed next to them. Similarly, new water features were established instead of recovering the existing ones, which would have reduced the building cost and was often done in late antiquity.
House with Floral Pavements had a fairly large vestibule (fig. 38). The vestibule (room 19) had a wide arched opening on its west wall, giving way to the unearthed parts of the house. To the east of the vestibule was a second smaller space, room 15, reminiscent of an inner vestibule that led both to the principal room of the house and to the northern portico. It was flanked with two small rooms that were probably the service spaces on its north.
25 Narrow streets that give way to the entrance doors of the houses are common also in modern Antioch (fig. 36).
62
The central element of the house was a large colonnaded courtyard, measuring about 14.60 m by 18.50 m. The courtyard was surrounded by porticos on three sides and ended with a nymphaeum on the fourth side, which rested against the south-east façade of the house. The nymphaeum had a monumental design comprising a majestic basin of approximately 10 m long and it was connected to a second fountain located at the center of the courtyard by a pipe.
In the same axis with the nymphaeum, on the opposite side of the courtyard, was centered the largest room of the house (room 12).26 Measuring about 11.50 m by 15.00 m, this room was most probably the reception room and was entered both through the north-eastern portico of the courtyard and also directly through the inner vestibule 15. The main space of the room opened to an alcove (room 13) at its end via a wide arch. This arrangement recalls the layout of rooms 15 and 16 of the earlier complex C300, though the latter one was much more monumental in scale. The floor of the main space was paved with marble and the floor of the alcove was paved with four panels of opus sectile while the front portico had a mosaic floor with a combination of geometric and floral motifs (fig. 36).
The south-east portico of the courtyard constituted the limit of the house in this direction. The floor of this portico and those of the two small rooms on its both ends (rooms 1 and 4) were covered with geometric mosaics. Adjacent to room 4, at the north corner of the house was another smaller room, which could have been used as a service space.
The north-western portico was organized as a linear space with an apse at its end (room 7-7a). This room measuring about 12.80 m by 4.30 m had a mosaic floor decorated with geometric patterns in the rectangular section and with seedling flowers in the
26 The dimensions of the room led some scholars to question the possibility that this space may have been open to the sky (Stillwell, 1941: 29; Morvillez, 2004: 278). Nevertheless, the existence of rooms with similar dimensions and layouts in the same city, like the one in the Constantinian Villa, and the several late antique houses at Apamea on the Orontes makes it more probable that it was a roofed space.
63
apsidal part. The function of this apsidal room could not be defined but its proximity to the vestibule should be highlighted. In one corner of the north-western portico was a square space, room 16 that was placed symmetrical to room 1 and probably functioned as a transition space between the vestibule and the inner courtyard. In the other end, behind the apse were the latrines. At the north-west limit of the house, behind the apsidal room were two small projecting rooms at both ends which were suggested to be towers (Stillwell, 1941: 30). No traces of construction were found further north-west and hence this area was suggested to have been a garden (Stillwell, 1941: 30).
The mosaic fragments found fallen from upper levels in room 16 provided evidence for an upper story, although no staircase was revealed. Considering the layout of the ground floor and reception spaces, the daily rooms of the family would likely to have been planned on the upper level.
The coins found from the site indicated an occupation in the House of the Floral Pavement up to the mid-7th c. AD, after which there is no evidence for habitation until modern times. In the latest phase of occupation, some spaces were subdivided and new rooms were added beyond the façade wall of the house in its south-eastern part (fig. 38).
3.2.6 The House of the Boat of Psyches
The mosaics of the House of the Boat of Psyches were discovered almost at surface level at Daphne and the house was partially excavated in 1934-35. The walls of the structure had been totally destroyed down to the foundations, thus the limits of some spaces and the means of communication between them remained undetermined. For this reason, the published restored plans of the building are hypothetical and they show different circulation patterns (figs. 39, 40, 41). Furthermore, Doro Levi argued that the indication of the north in the excavation reports was wrong since the north arrow actually pointed to the west (1947: 167). Almost all the later studies followed the north
64
direction shown on the plan in the excavation reports while this study uses the corrected orientation shown on Levi’s plan.
The excavated part of the House of the Boat of Psyches contained eight different spaces, each with sumptuous mosaic pavements (fig. 41). The house was dated to the 3rd c. AD based on the style of the mosaics (Levi, 1947: 167-191). The core of the excavated area of the house had a row of three rooms. In the center was the largest room 1, measuring about 5.00 m by 5.00 m. The mosaic panels of the room present the common T pattern of a triclinium, with a different organization (fig. 42). The mosaic floor was divided into nine panels with a central square one depicting Europa and the bull that faced the back of the room. The three panels in the front depicted busts of mythological characters, with Oceanus and Thetis in the center, to be seen by the incomers. The other five panels that would receive the banquet beds at the back of the room had geometric motifs. The triclinium opened to a portico via a wide entrance and had a view of the nymphaeum. Stillwell and Levi, in their plan, proposed three more doors for this room through which it would communicate with the flanking rooms and the corridor at the back (figs. 38, 39). Considering the need for a service entrance and the common presence of annex rooms that communicate with the main reception rooms in many Roman period houses, the existence of lateral doors in triclinium 1 is very likely.
To the west of room 1 was the small room 2 with a figurative panel, representing Pegasus and two nymphs that faced to the back of the room where there was a rectangular geometric panel. The function of this room has been undetermined. It could be a smaller reception room or a more private chamber. The access to this room was treated differently in three different plans. Levi suggested a single lateral door opening to triclinium 1 while Dobbins, following the plan of Kondoleon, proposed an entrance from room 8 to the north. Stillwell, on the other hand, assumed that this room opened both to triclinium 1 and to portico 4 just like the adjacent rooms. Considering the usual orientation of the main panels to the opposite side of the entrance doors and the figural
65
panel on the floor of portico 4 that faced room 2, it seems more reasonable to place the door on the south wall opening to the portico.
Room 3, on the east side of room 1, functioned as a triclinium too, as suggested by the T-shaped arrangement of the mosaic floor. What is unusual in this mosaic arrangement is that the main figural panel representing the boat of psyches led by Eros, after which the house was named, was not oriented towards the back of the room, but to the front, in the same manner with the other three panels depicting Dionysiac scenes (fig. 43).
Portico 4 had a colonnade that ran along the three rooms described above and brought light and air to probably all three of them. Some blocks of the stylobate were found in-situ during the excavations while the columns were already lost. Thus, the number of columns and their spans shown in the plans are hypothetical. The mosaic floor of the portico had three figurative panels on the axis of the three rooms that were laid between geometric panels with varying widths. The figural panels in front of the two triclinia depicted erotic scenes with one or more characters in symplegma (fig. 44).27 The third panel in front of room 2 represented an ithyphallic dwarf.28 All three panels were oriented towards the rooms.
To the south of the portico was an elongated court that is supposed to have been open to the sky. At the back of this court was a nymphaeum that was decorated with five semi-circular niches and a mosaic representing erotes riding dolphins and different species of fish. In front of the niches was a basin of 0.60 m wide to collect water. Since the southern area behind the nymphaeum was not excavated, it is not certain that this courtyard with fountain constituted the limit of the house in this direction.
27 In modern art history and archaeology, the term symplegma refers to explicit depictions of sexual intercourse, in particular, to more complex and unusual sexual positions or unusual pairings. Doro Levi pointed out that this theme was probably derived from the Hellenistic painted models and was a very rare theme in Roman mosaics (1947: 183).
28 According to Stillwell, the erotic themes in the portico and the co-existence of three triclinia within the House of the Boat of Psyches may suggest a meeting place for dining and diversion rather than a purely domestic establishment, as he also proposed for the House of the Buffet Supper (1961: 52).
66
In the northern end of the excavated area were two large spaces and a corridor in-between whose northern limits remain unexcavated.29 The organization of the mosaic pavement of room 8 was identical to that of room 2, and has a figurative panel in front and a rectangular geometric one at the back. This arrangement is not the T pattern commonly seen in triclinia but its iconography which depicts Opora and Argos reposed on a banquet bed and being served by Oinos in the form of a satyr, connotes a banquet room. Room 8 opened to a narrow corridor with a geometric mosaic pavement and through it to room 6. Levi proposes a second lateral entrance on the north wall, because of the orientation of the peacock figure in the mosaic floor to this wall. Dobbins, on the other hand, connects room 8 and room 2 via a door.
The northern part of room 6 was destroyed by a later construction and little is known about this space. It was paved with a large rectangular panel representing the bust of Thetis surrounded by different species of fishes and an erotes riding a dolphin (fig. 45). Room 6 has been identified as an open courtyard because of the marine iconography that was usually associated with courtyards and the two niches at the eastern wall that were probably the remains of a nymphaeum.
At least one later phase of construction and occupation was attested in the House of the Boat of Psyches. A geometric mosaic was laid out on the earlier mosaic on the north-east end of the excavated area and the colonnade of the portico 4 was replaced by a wall in which some parts of the earlier columns were reused. The chronology of the reconstruction remains undetermined.
3.2.7 The House of Menander
The House of Menander in Daphne is located about 100 m south of the House of the Boat of Psyches. It was laid out in a densely occupied insula that was shared by at least
29 Stillwell argued that the three unearthed rooms to the north may have belonged to a separate property (1961: 51).
67
five contiguous houses (fig. 46). The insula was bordered by a colonnaded street in the south while its limits on the other three sides were not reached during the excavations that were carried out in 1939.
The archaeological data indicates a continuous occupation in the House of Menander between the 2nd and 6th c. AD. The published restored plan of the house belongs to the main phase of reconstruction that probably took place between the mid-3rd c. and the beginning of the 4th c. AD (Morvillez, 2007: 55; Martz, 2011: 55). With the restoration works the initial structure of the 2nd c. was extended and, according to Stillwell, at least two other houses were consolidated (1961: 52). At its maximum extension, the House of Menander was a vast dwelling covering an area of more than 1,000 m2.
The excavations revealed some well-preserved in-situ architectural elements and many mosaic pavements by which the boundaries of spaces were well-defined. Nevertheless, like many other houses excavated in Antioch, its outer limits and circulation patterns are poorly known. So far, two entrances, from the colonnaded street to the south, were identified. The one at the south-east corner of the house was probably the principal entrance. It opened to a large rectangular vestibule (room 21) that indirectly led to different groups of rooms within the house. The other entrance, near the service spaces in the south-west corner, was likely to be a secondary or service entrance.
The exposed rooms of the House of Menander were organized around four inner courtyards (fig. 47). At the south-east corner of the house, near the entrance, were the first group of spaces including two probable reception rooms and a service room that opened to courtyard 1 (fig. 48). Courtyard 1 had a mosaic pavement with a central panel representing three fishing erotes and a geometric frame. It was also endowed with a fountain, decorated with niches and ornamental columns. Rested on the south wall of the court, the fountain was not centered to the axis of the court but to room 2 behind. Likewise, the figural panel of the mosaic floor was oriented to that room. The entrance of this privileged room was emphasized by two columns that created a sort of portico. Below the floor level of the northern part of the courtyard a mosaic
68
pavement and a basin that belonged to an earlier phase are discovered (fig. 49). The mosaic panel, which depicted birds and flowers, faced room 2. The orientation of the earlier pavement and the basin indicate the unchanged functions of rooms 1 and 2 as well as the means of communication between them in two subsequent phases of occupation, despite a change in layout.
Room 2 was a triclinium as understood by the organization of its mosaic pavement. The poorly preserved pavement had nine panels; four of which were figurative and were laid to form the T shape of the triclinium while the rest were geometric and surrounded the figurative ones forming a U-shape. The panel at the entrance of the room faced towards the court to be seen by the arriving visitors. The central panel representing Narcissus, was also oriented towards the court, instead of the back of the room as would be expected. Room 2 was laterally entered from a narrow corridor, room 10, that provided direct access from vestibule 21. In the excavation reports a hypothetical door opening to portico 12 on the north wall was proposed too.
Room 3 was a smaller room with a mosaic pavement in U shaped arrangement, which led it to be identified as a triclinium too. Its highly destroyed figural panel faced the back of the room. This room opened to the south-eastern corner of court 1 and had only a partial view of the court and the nympheum. Nevertheless, the asymmetrical placement and design, with a circular niche only in its western edge, of the nymphaeum may indicate a concern for the sightlines of that room.
The layout of room 9 suggests a kind of transition space between corridor 10 and courtyard 1. There was no trace of a pavement and probably it was used as a service space. Dobbins suggested an additional door between courtyard 1 and space 4, which provided the connection between that part of the house and the latrine to the west of room 4.
To the north of the first group of rooms mentioned above was another suite of spaces that were highly destroyed by later constructions. Two rooms and a portico were well
69
identified in this area that was organized around a courtyard. The courtyard, itself, was totally demolished by the addition of a later shallow pool (fig. 50).
Room 11 opening to portico 12 was reminiscent of room 3 in its dimensions and layout of its mosaic pavement. It had a U-shaped arrangement with a figural central panel, representing Menander and Glycera on a kline, after which the house was named. The theme of the mosaic and the U-shaped arrangement of the pavement suggest a triclinium function but Dobbins states that the dimension of the U is too narrow to contain three beds (2000: 66).
Room 19, to the east of the courtyard, was poorly preserved but its mosaic floor helps to identify its dimensions and function. The figural panel of the pavement, within a U-shaped organization, depicted a banquet scene oriented to the back of the room. This arrangement not only specifies the use of the room as a triclinium but also indicates that it most probably had an opening on its west wall that gave access to the courtyard.
To the south of the courtyard was colonnaded portico 12 that lied along the north wall of rooms 2 and 11. The portico was paved with a mosaic panel of which only the eastern part was preserved. This part of the panel represented a man and a woman on a chariot in front of a building. This scene was intended to be seen from room 11. Portico 12 was perpendicularly aligned to room 13 and a hypothetical door to this room was suggested in the excavation reports.
The largest spaces uncovered within the complex are a courtyard and a triclinium opening to it that constituted the third reception suite at the westernmost part of the house. The court had a large decorative nymphaeum non-axially attached to the north wall. A basin was added against the west wall of the court in a later period. In front of the nymphaeum, the court had a poorly preserved mosaic floor pavement with an aquatic iconography.
70
Triclinium 13 opened to the fountain court through a wide opening with two supports. The mosaic floor of the room which has a T/U layout shows no concern for symmetry or axiality. The bar of the T was composed of four figural panels each oriented towards the back of the room. They are not of the same dimension or style. The main panel as well was not placed on the axis of the room. It was divided into four by diagonals and represented four different scenes that were oriented towards the walls of the room. A rectangular geometric panel was placed next to the central panel to the south of the room, from where the best view of the nymphaeum could be captured.
The courtyard and room 13 were entered from a large rectangular corridor-like space through a narrow door to the south that may well have been a part of this reception suite. Fragments of geometric mosaic panels are discovered on the floor of room 7. The main access to that suite had not been clearly identified. It may have been reached through the secondary entrance mentioned above. The rooms located in the close vicinity of this entrance were not paved with mosaics except room 5 and were described as probable service spaces.
In the northernmost zone of the excavated area of the insula, remaining between two neighboring dwellings, was a group of small rooms that were organized around the fourth courtyard of the House of Menander. The remote location of this section and the small size of the rooms led it to be identified as a more private or familial suite (Dobbins, 2000: 59; Martz, 2011: 49).
Courtyard 17 was surrounded by a colonnade with three columns on the south and east sides and at least with a single column on the north side. It had a fountain with three niches rested against the center of the south wall (fig. 51). The basin on the north wall was a later addition. The mosaic panel on the floor which depicts the busts of Oceanus and Thetis is oriented towards the room to the east. Room 18 was the largest and the most privileged room in terms of its location across the fountain. Unlike the pavements of other similarly placed rooms of the house, the rectangular mosaic pavement of room
71
18 did not present the common pattern of a triclinium but had a geometric decoration. Behind this room, two smaller rooms were partly excavated.
To the south of court 17, rooms 14 and 15 with simple geometric mosaic pavements opened to portico 16. The portico itself had a mosaic pavement with a figural panel depicting Apollo and Daphne, situated between two geometric panels. Considering the orientation of the figural panel towards the courtyard and the location of the columns with no respect to the openings of rooms 14 and 15, it can be said that the visual perception of courtyard 17 was given more importance than either of the rooms.
To the north of the courtyard, at the northern limit of the excavated area, was another portico (portico 20). The fragmentary mosaic floor of this portico was composed of a geometric panel and two figural panels facing north which may indicate the existence of at least one room at this side.
The House of Menander provides a good sample for the layout of the Roman period houses of Antioch with the considerable number of spaces it yielded. Nevertheless, the limited state of excavations and the lack of sufficient information on the circulation patterns make it difficult to identify the limits of the house and its neighbors. It seems possible that some rooms, mentioned above, especially the triclinium 13 and its dependencies may well have belonged to a separate property. Likewise, the shortage of archaeological data and the lack of archaeological finds, prevent suggesting precisely the function of the building, as was the case in most of the excavated structures in Antioch. Dobbins, for instance, called attention to the presence of several dining suits within the building and asked for the possibility that the House of Menander, at least partly, might have functioned as a dining club (2000: 59).
3.2.8 The Constantinian Villa and The House of Ge and the Seasons
The Constantinian Villa and the House of Ge and the Seasons were two partially excavated houses in Daphne. Few spaces were unearthed in these houses, some of
72
which were poorly preserved, but they constitute a good comparative sample and are worth examining.
The Constantinian Villa is dated to the second quarter of the 4th c. AD by a coin of Constantin the Great after which the house was named. The excavated part of the house consisted of a large room, two corridors or porticos, and a small square room (figs. 52, 53). The elaborate room 1, measuring about 8.50 m by 12.50 m, was probably the principal room of the house. The rich and complex mosaic floor of this room was divided into two panels, one square and one rectangular, reminiscent of the arrangement of some triclinia found in other houses (fig. 54). Nevertheless, it differs in its organization. The square panel placed at the front of the room was divided into four triangles with figural scenes that were separated by four standing figures representing four seasons at the diagonals. It was surrounded by a series of small trapezoidal figural panels with hunting scenes. In the center of this portion of the mosaic floor was placed an octagonal fountain. The rectangular panel at the back of the room had alternating geometric motifs and medallions with busts or figures. Stillwell described this room as an oecus or a great tablinum instead of a triclinium. Morvillez, on the other hand, resembled its two-partite organization to the large reception rooms with alcoves found in the late antique houses of Apamea on the Orontes (2007; 76).
A corridor, or portico, (room 2) with a geometric mosaic pavement ran along the south wall of room 1. Beyond this long corridor, a small square room 3 with a geometric mosaic floor pavement was exposed. Because of the slight irregularity in its alignment and the later date of the pattern in the mosaic, Levi argued that this room belonged to another building (1947; 226). According to Stillwell, however, this room was part of the Constantinian Villa. Considering the layout of this section of the house, the portico does not likely constitute the border of the property and there should have been more rooms behind it.
73
To the east of room 1 was a wider portico with a damaged mosaic floor. The length of the portico exceeded the limits of room 1 towards south and north, which may indicate the existence of a northern corridor that was symmetric to the southern one. Stillwell argued that the unexcavated area to the east of portico 4 could have been a courtyard (1961: 53).
The House of Ge and the Seasons was dated to the late 5th c. AD. During the excavations in 1935 a large room, a portico, and two small rooms of the house were revealed (fig. 55). The main room measuring about 8.50 m by 10.50 m opened to a wide portico and to the courtyard beyond. The mosaic pavement of the room did not present the common triclinium pattern. Instead, it had five medallions, with Ge in the center and the four seasons around, arranged within a geometric pattern. The central motif was laid to be viewed from the entrance while the others were oriented to the center. The wide portico to the west of the room had a mosaic pavement with geometric motifs. The length of the portico exceeded that of room 1 in both directions, like portico 4 in the Constantinian villa, which recalls the existence of annex rooms to the north and south of room 1. To the south of the courtyard, two small rectangular rooms with geometric mosaic floors and a rectangular space between them were exposed.
Below the 5th c. floor level of the House of Ge and the Seasons, two earlier phases of construction were detected. In the second phase of occupation, an earlier room was reorganized to provide a more monumental layout. This room of the first phase was extended, paved with a geometric mosaic and received a rectangular basin on one side and an apse on the other side (fig. 56). During the restoration of the 5th c. AD the room was reenlarged one more time and reorganized to receive a totally different layout.
3.2.9 The Yakto Complex (The Villa of Ardabur)
A chance discovery of a sumptuous mosaic with a hunting scene and a topographical border, the one mentioned above, in a garden in the Yakto section of Daphne in 1932 led to the excavation of a large residential complex. According to Lassus, who
74
excavated the area, this large building was probably the residence of Ardabur, the military commander of the Orontes region around 450 AD, since one of the buildings on the topographical border of the mosaic was designated as his private bath (1984: 362).
Dated to the second half of the 5th c. AD, the Yakto complex replaced a 3rd c. building, or adjacent buildings, whose structure was partially reused with some transformations and additions (fig. 57). According to the coin finds it was occupied until the beginning of the 6th c. AD. The building differed from the other houses excavated in Antioch so far, in terms of its architectural layout and decoration. It had mainly three parts, with varying size, structure, and appearance that had a very limited and indirect communication between each other (fig. 58).
The complex rested against the slope of a hill on its south-east and was open to the panorama on its north-west. Its entrance(s) was probably placed on the north-east façade that could not be reached during the excavations or on the north-west façade that was totally destroyed (Lassus, 1938: 99). A long corridor, room 5, that was arranged in the manner of a street formed the main axis of the house and gave way to different sets of rooms. The corridor was 39.00 m long and 3.15 m wide. It was paved with a simple geometric mosaic of rough workmanship (fig. 59).
To the north-west of the corridor was the first suite of rooms of the complex that were arranged around a large courtyard measuring approximately 11.00 m by 15.50 m The courtyard had colonnades on its three sides and was bordered by a wall on its north side. It was paved with a rich mosaic that was replaced with an earlier one in the 5th c. AD. The latter pavement was highly destroyed but a preserved fragment of it shows a hunting scene. The porticos, on the other hand, had marble pavements that were composed of neatly cut marble fragments with different colors displaying a geometric pattern.
75
Two exedrae, rooms 8 and 47, opened to the south-eastern and south-western porticos of courtyard 2. Exedra 8 originally belonged to the first building phase and continued to be used in the 5th c. AD. It measured about 5.50 m by 3.70 m. Its pavement was not preserved but the layer of hard cement on its floor indicated the presence of a mosaic pavement (Lassus, 1938: 107). In addition, the small mosaic fragments with a gold background, composed exclusively of cobalt glass and whitish fragile cement, different from the ones in the floor pavements indicated that the walls and/or the semi-dome of the exedrae were also covered with a sumptuous mosaic. Exedra 47, did not belong to the disposition of the first building but was added to the structure in the 5th c. AD, to the detriment of some earlier spaces. Its northern wall was totally constructed over an earlier fishpond. This later exedra was larger than the first one and was approximately 6.00 m wide and 4.30 m long. Its floor level was raised from the portico it opened. Several fragments of a mosaic decoration similar to those found in exedra 8 were found in this room too. Since the north-western edge of courtyard 2 was totally destroyed, the existence of a third exedra on this side of the courtyard could not be specified.
In the south-eastern side of the courtyard, near or behind exedra 8, were a series of spaces. Rooms 7, 9, and 21 had mosaic pavements with geometric patterns, while room 15 did not yield any pavement. The communication patterns of this part of the house could not be totally determined except the openings between rooms 7 and 15, 9 and 21 and 21, and the portico.30 Room 9 was provided with a basin carved on the floor and a marble plaque with a hole leading to the sewer. The irregular forms of the spaces, except room 7, and the existence of running water may indicate a service function for at least some of the spaces in that part of the house.
Two adjacent spaces on the southern corner of courtyard 2, rooms 48 and 52, were differentiated by their architectural layout and structure. They were contemporary with exedra 8 and belonged to the reconstruction period of the building in the 5th c. AD.
30 The other doors drawn on the plan are hypothetical.
76
Room 52 was a cruciform room inscribed on an almost square plan measuring 8.00 m by 7.41 m. The four arms of the cross had a common width of approximately 2.90 m and length of 1.50 m in average. The superstructure of the room should have been composed of four vaults above the arms of the cross and a dome in the center (Lassus, 1938: 114). The pavement of the floor was also differentiated between the arms of the cross, which were paved with small marble panels with black and white checkerboard, and the central space, which was paved with the same material in eight concentric circles (fig. 60). The walls were, at least partly, revetted by white marble, the fragments of which were found in-situ. Room 48 communicated with the portico 6 and room 52 and seems to have served as an antechamber for the latter one. On its north-western edge, facing the door of room 52 it had an arched or vaulted part as attested by the reinforcements in two corners. The floor of room 48 was paved with marble too but the pavement of the vaulted part which had a hexagonal motif was differentiated from the rest of the room with a simple grid. The symmetrical counterpart of room 48 was the highly destroyed room 46 that was smaller in size and had a geometric mosaic pavement.
The second suite of rooms of the Yakto complex was placed to the south of the first one and separated from it by a narrow corridor (room 14). This suite was more unified, organized and enclosed in itself and had only a very limited and indirect communication with the rest of the house. A principal door gave way to it from the main corridor 5 and another, probably a service door, provided communication between a rear room (room 13) and corridor 14. This part of the complex was originally built in the 3rd or 4th c. AD and continued to be used in the 5th c AD. The principal room of the suite, room 10, was composed of a cruciform space inscribed in an almost square plan by the construction of four corner rooms. The main room measured approximately 12.70 m by 13.90 m and the corner rooms had slightly different dimensions varying between 2.50 m and 2.95 m for each side. The north-eastern wall of room 10 that totally disappeared, was composed of a portico with three arcades (Lassus, 1938: 119). On the same axis with the portico, on the south-western wall of room 10 was a large nymphaeum that was decorated with niches, small columns, and
77
marble revetments (fig. 61). The floor of room 10 was paved with opus-sectile which replaced a mosaic pavement in the late 5th c. AD. The four corner rooms, on the other hand, preserved their original geometric mosaics. Fragments of glass mosaics with white cement and remains of painted plaster found in the debris indicated that the walls of the room were also highly decorated. According to Lassus some fragments of a stone mosaic found in the debris probably fell down from an upper gallery (Lassus, 1938: 120).
Three interconnected rooms with mosaic pavements were annexed to room 10 on its south-west. The first one, room 13, was organized like a passage room with at least three doors that linked room 10, corridor 14, and room 19. Room 20 was connected to rooms 18 and 19. Room 19 had a basin while room 20 was provided with a drain hole. To the south of this second suite, at a 0.60 m higher level, was a large and elongated space, probably an open court that had a geometric mosaic floor and a basin. This space was annexed by three rooms to its south which were highly destroyed.
The remaining part of the Yakto complex was situated to the north-east of corridor 5. This part was highly destroyed and most of the walls are disappeared. Rooms in that area were organized around a vast and irregular court in trapezoidal form, measuring about 190 m2. The courtyard had porticos with colonnades on the north-western and probably on the north-eastern side. To the south-east, it gave way to a number of rooms through a corridor that ran along corridor 5. There were two large fishponds in the courtyard that belonged to the earlier phase of occupation; these were filled and covered with mosaic pavements in the 5th c. AD (Lassus, 1938: 132).
The north-western portico (room 1) extended to the wall of courtyard 2 within the first suite of rooms of the complex and was connected to it via a door with two steps. Behind the portico was a row of four rooms. The entrance of one of these rooms, that of room A, was highlighted by a triple opening arranged between two columns. This room had the above-mentioned sumptuous mosaic that represented the hunting scene and was framed by the topographical border (fig. 62). In the center of the pavement was a
78
medallion showing the bust of Megalopsychia that faced the entrance. Room A opened to the annex room B that had a rich mosaic pavement too. At the center of the mosaic of room B was a representation of Thetis that was oriented towards room A; the fishing erotes that were represented in the same area were oriented towards the walls of the room (fig. 63). The plan layout, emphasized entrance and the sumptuous decoration of this suite of two rooms indicate that this was an important spatial unit within the complex, the composition of the mosaic pavements, on the other hand, makes it difficult to suggest a function or determine the layout of the furnishing.
At the south-eastern end of this part of the house were two further rooms (room 31 and 32), one of which was paved with a very roughly applied mosaic. Since this part of the house underwent multiple rearrangements, it presents an incoherent situation (Lassus, 1938: 139). To the east of the complex, separated from it by a free space on one side and by a thick wall on the other side, was discovered a poorly preserved small bath. Nevertheless, the split location of the bath led the excavator to propose that it did not belong to the Yakto complex.
3.3 Overview of Domestic Architecture in Antioch on the Orontes
The duration and the extent of excavations conducted so far in the ancient city of Antioch on the Orontes and its vicinity, as well as the poorly preserved condition of the unearthed remains, have provided a limited amount of data on the domestic context of the ancient capital. The current archaeological information is mainly composed of incomplete and partly hypothetical plans which makes it difficult to precisely define the chronology, extent, architectural and structural layouts, and in some cases, even the function of the excavated buildings. Most probably for this reason the houses of Antioch have hitherto been excluded in many studies that focused on Roman domestic architecture. Nevertheless, the recurring patterns in the architectural and decorative schemes, the presence or absence of some specific architectural elements, and the continuities and changes in subsequent occupational phases provide some valuable information about the houses of Antioch.
79
First of all, the excavated houses in Antioch and its suburbs vary in size and layout. The limits of many excavated houses could not be established but the uncovered cores indicate that except a few vast residences like the Atrium House or the House of Menander, they were generally modest and compact dwellings with a certain level of luxury. The plans showed no regularity or axiality in the overall design. In the houses, in which at least one street entrance was spotted, like the House of Menander or the House of the Drinking Contest, no axial route could be defined from the entrance towards the core of the house. The symmetry, on the other hand, had been sought for in few instances, specifically in the triclinium-courtyard-nymphaeum layouts, as in the House of the Boat of Psyches. The smaller size of the dwellings and the irregularity of their plans may have been, at least partly, due to the topographic conditions of the sloped areas and the availability of the buildable area within the dense settlement.
The spatial focus of the Antiochene houses was the triclinium that usually came into prominence with its size, layout, and decoration. The triclinia were identified by the T/U or U-shaped arrangement of the mosaic pavements, indicating a Roman form of space destined for reclined banqueting.31 The number of triclinia varied and some houses had multiple triclinia. The existence of multiple dining rooms in the eastern and western contexts has already been attested by the ancient literature and the archaeological evidence. In the Satyricon of Petronius (1st c. AD) Trimalchio invites his guests to take a hot bath before changing triclinium (Pet. Sat. 73-5). Similarly, in his letter to his friend Cninius Rufus, Pliny asks for his grand dining halls and smaller ones that were for family and intimate friends (Ep.I-3.1).
The triclinia were mostly annexed by smaller rooms like those seen in the House of the Buffet Supper, the House of Dionysus and Ariadne, and the House of the Boat of Psyches. The mosaics of the annex rooms were not necessarily organized in the manner of a triclinium but they were of high quality too, in their iconography and
31 For social and architectural contexts of Roman banqueting and triclinia see Dunbabin 1991 and 2003; Nielsen 1998. For a discussion on triclinia in Northern Syria see Martz 2015.
80
workmanship. They might have been used either as secondary dining rooms for smaller and more private banquets and/or as private chambers. The triclinia, and in most cases the annex rooms, opened to a colonnaded portico beyond which was an open courtyard. In some cases, the entrance of the main room was emphasized by columns.
The courtyards of the Antiochene houses differed, in function and layout, from the ones excavated in many other Roman cities including Pompeii, Apamea on the Orontes, or Palmyra. Except for the courtyard of the Atrium House, the earliest dated house excavated in Antioch, most of the courtyards were not conspicuously central and distributive. They were usually smaller and sometimes were multiplied in number. There was a privileged link between the triclinium and the courtyard and the latter functioned as if it was an annex or extension of the former. The westernmost courtyard of the House of Menander, for instance, served only to room 13. The courtyards were not planted and designed as gardens in the Roman fashion but were paved following the Greek or eastern traditions.32 Peristyles were not arranged either, except in the Atrium House and the Yakto Complex.33 Instead, porticos that are reminiscent of pastas or prostas in Greek houses were used.34
Since Antioch and especially Daphne had abundant water supplies, running water was an important feature in the houses of Antioch. Fountains were revealed in almost all
32 For gardens found in the Roman houses in Italy see Jashemski et al., 2018, von Stackelberg, 2017.
33 Adopted from Greek public architecture and the Hellenistic palaces, peristyle courtyards became common in the elite houses from the 3rd c. BC on (Wallace-Hadrill, 1994: 20-21, Zimmer, 2010: 158, Rumscheid, 2010: 122). With the Roman conquest of the East in the 2nd century BC the peristyle courtyards became popular in the Roman domestic context too (Clarke, 1991: 12, Zarmakoupi, 2010).
34 For the Greek houses with pastas plan in Olynthos see Robinson and Graham 1938, for those with prostas plan in Priene see Akurgal 1993; Schede 1964 and Rumscheid 2010. For houses in Pergamon see Wulf-Rheidt 2011 and 2014. For recent discussions on the architectural and social contexs of the Greek and Hellenistic houses see Ault and Nevett, 2005; Winter, 2006 (157-182), Nevett, 1999 and 2010. For the Hellenistic houses excavated in Seleucia on the Tigris, see Hopkins, 1972 and Karampekos, 2020.
81
of the unearthed courtyards, confirming Libanius who praises the waters of his city as such:
One can judge the wealth of our waters by the number of the houses, since there are as many fountains as there are houses, or rather there are many fountains in each house, and indeed the majority of the shops are also adorned in this way (Or. XI.246).
Water features were not used only for sanitary needs or for climatization but they were utilized as an important decorative element constituting a medium for displaying comfort and luxury and reflecting the house owner’s high status.35 Almost every courtyard was provided with a monumental nymphaeum that rested on one of the walls. The nymphaea were decorated with multiple niches, colonettes, paintings, mosaics, and in some cases with statues. As such, while the courtyards magnified the spatial value of the triclinia, the nymphaea provided a theatrical effect for the diners. Lassus recalled traces of paintings representing fake gardens on fountains which created an illusion of perspective (1984: 370). Thus, the existence of nymphaea in the courtyards, and even in small porticos like the one in the House of the Boat of Psyches, may indicate a demand to compensate the absence of gardens for which space was lacking.
The ostentation in the excavated houses of Antioch comes much from their mosaic pavements. All the excavated houses, even the modest ones, had sumptuous mosaic floors in their reception rooms, courtyards, and probably in the living units of the family. The technique and iconography of the mosaic panels sustained the Hellenistic tradition and represented scenes from Greek mythology or literature. In triclinia, the themes related to the consumption of food or drink were common while the aquatic themes were preferred for the courtyards.36
35 Hobson, 2009: 117-118; Wilson, 2012; Uytterhoeven, 2013: 15 and 2020: 336; Vanesse, 2014.
36 For a comprehensive study on the mosaics of the Greco-Roman world see Dunbabin, 1999. See also Kondoleon, 1991.
82
The mosaic pavements within the houses did not only enhance the visual quality of the rooms but they were also utilized, as an architectural tool, to link the spaces by sightlines and accentuate the circulation patterns. The figural panels depicted in the horizontal bar of the T-shaped mosaic in the triclinia, were oriented to outside to welcome the entering guests while the main ones were facing the back of the room, where the furniture was placed on geometric panels.37 The smaller figural panels placed in front of the entrance doors on the porticos were also oriented towards the rooms, for the view of the diners, probably with the intention to magnify their viewpoint and attract their attention towards the courtyard and the nymphaea.
There is a lack of information on the service spaces in the houses of Antioch since the rooms with mosaics, or even the mosaics themselves, were given priority during the excavations, to the detriment of the unpaved areas. Similarly, there is no information about the upper levels. No staircases have been revealed in the domestic structures so far but few archaeological remains in the debris, like the column plasters in the Atrium House or the mosaic fragments in the House of Floral pavements and the Yakto Complex, indicated the existence of upper levels. The written sources also confirm the presence of two and three-story dwellings in Antioch. Libanius, for instance, praised the wind Zephyr “as it flows over not only the three-story dwellings of the rich but also over the humble houses of the poor” (Or. XI.225). He also appreciated Daphne as there were pavilions that were higher than the trees (Or. XI.234).
The archaeological evidence and the written sources indicate that the opulence of the dwellings in Antioch and its suburbs continued well throughout late antiquity, with some changes and transformations in architectural and decorative schemes. In this period, the reconstructions were done either in the form of modification, as in the Constantinian Villa, or of superimposition, as in the House of the Floral Pavements and partly in the Yakto Complex. The first practice, in which the existing buildings
37 The main figural mosaic panels of room 3 in the House of the Boat of Psyches and room 2 in the House of Menander are exceptional since they were oriented towards the entrance door.
83
were modified according to the newly emerged needs and tastes of the period, was common in late antiquity. The latter one, in which new walls were constructed over an earlier building without paying attention to the original layout, however, was neither a frequent application nor a feasible one for economic terms. For Antioch, nevertheless, this could have become inevitable because of the frequent earthquakes that devastated the region and the shortage of empty building plots. Libanius, who informed about the dense building activity in late antique Antioch, also pointed out the reuse of the existing buildings as such:
Thus, it is not without reason that the city is always filled with building activity; some buildings are being torn down, some are half completed, and for still others the foundations have just been laid or are being excavated. Everywhere are the cries of those urging on the workmen, and ground that last year was planted with vegetables is built upon this year. Men know, indeed that here they have, while still alive, all the things which the poets promise to the righteous dead (Or. XI.227).
Thus, today when you excavate in order to lay a foundation, everywhere you meet some trace of former times, and many people, using what has been preserved instead of what they intended to use, add their other material and build. So, if some buildings had not been destroyed and others built on them, and if as much as is now used for rebuilding were employed instead for enlargements, many people would now be deprived of much land which is now under cultivation (Or. XI.229).
The most remarkable modification of late antiquity that has been observed in the houses of Antioch was the increased size and the changed layout of the reception rooms. The T/U or U arrangement of mosaic pavements had started to disappear from the 4th c. AD onwards. Instead, the rooms became paved with single larger panels in which the figures were radially oriented towards the four walls, as in the Constantinian Villa and the House of Ge and the Seasons.
New forms of reception spaces were also introduced in late antiquity. The most common of these was the two-partite room in which a larger, rectangular front space had widely opened to an elevated alcove that was reminiscent of an iwan, the distinctive spatial unit in Parthian domestic architecture (fig. 64). The passage between
84
the two spaces was sometimes accentuated by arches. Two rooms designed in this layout were rooms 16-17 in the C300 complex, dated to the late 3rd c. AD, and rooms 12-13 in the House of the Floral Pavements, dated to the early 5th c. AD. The latter one was much larger and displayed a more monumental layout. Both rooms were accessed from an open courtyard through a portico.
The absence of apsidal rooms in the excavated houses of late antique Antioch is noteworthy since such spaces were widely used in the houses of many other cities of the later Roman world. The only example of an apsidal space is room C2 of the House of the Buffet Supper. The room, however, was turned into a rectangular space in the late 3rd c. AD. In this rearrangement the size was, presumably, a more primary consideration than form. Similarly, the reception room of the House of Ge and the Seasons received an apse together with a basin and mosaic pavement at a given time, but when the house was reorganized in the 5th c. AD the room was extended and the apse was omitted. The only apsidal room unearthed in Antioch and dated to late antiquity so far was the elongated room 7-7a in the House of the Floral Pavements, which did not apparently serve as the main reception room. The Yakto Complex had two apses around the colonnaded courtyard 2 but these spaces were organized as exedras or iwans rather than the apsidal rooms that were frequently used in late Roman period houses.
Archaeological evidence has demonstrated that the ostentation in the decoration of the domestic spaces in Antioch continued well into late antiquity. This luxury, however, was often mentioned and criticized by Christian preachers in that period. John Chrysostom, a 4th c. AD bishop from Antioch, argued, in his sermon on the Holy Martyrs that:
There are some who are unfortunate enough to enhance their homes of this world with golden ceilings, with mosaics and floral paintings, with magnificent columns and all sort of ornaments (PG 47. 360).
85
In this period new themes and designs were introduced for the mosaic panels. From the 4th c. AD on the mythological repertory, the main source of the figural compositions, was replaced by repetitive geometric or vegetal patterns, animal motifs, and hunting themes.38 Besides, marble which was used more often in the public sphere, became a common material of decoration in the domestic context. Many rooms within the houses were paved with marble, mostly in the form of opus sectile.39
38 For a detailed and comparative study on the mosaics with hunting scenes at Antioch see Lavin, 1963.
39 The transformation of the architectural and decorative schemes in the domestic contexts of late antiquity will be discussed further in the next chapter since the sample excavated at Apamea on the Orontes provides a relatively more sufficient archaeological data.
86
CHAPTER 4
APAMEA ON THE ORONTES:
DOMESTIC SPLENDOR IN LATE ANTIQUITY
4.1 Urban Layout and Development of a Later Provincial Capital
The earliest habitation on the site of Apamea on the Orontes was concentrated on the hill dominating the Orontes River and the Ghab valley. The Seleucids made this promontory hill the acropolis, the defensive nucleus of their settlement and the lower city spread towards the east (figs. 65, 66). Hellenistic and Roman Apamea covered an area of approximately 250 ha that was encompassed by ramparts at least from the 2nd c. BC onwards (fig. 67). However, since the urban area is thought to be rather large for a colonial settlement, it is assumed that the space within the walls was not fully built in its early history (Balty, J. Ch., 1991: 17). The city was entered through three main gates on the north, west, and south, the latter of which had a concave outline recalling the majestic circular entrance of the city of Perge, in ancient Pamphylia. Within the walled enclosure, the main colonnaded street-oriented towards the north-south axis, while two other main axes in the east-west direction intersecting each other organized the distribution of urban functions and introduced an urban hierarchy. The streets intersecting at regular intervals formed an orthogonal network with insulae of approximately 110.00 m by 55.00 m with the ratio of 2:1. Noticing the similarities in the basic principles of the urban layout, especially the orientation and dimension of the street network, in Apamea, and Antioch on the Orontes and Laodicea, Balty raised the possibility that the four Seleucid ‘sister cities’ -Antioch, Apamea, Laodicea, and Seleucia Pieria- were planned more or less simultaneously by the same urban planner (Balty, J. Ch., 1991: 16; Balty and Balty: 1969: 34).
87
The main principles of the Hellenistic town planning were maintained through the Roman period with some spectacular touch-ups. The city highly benefited from the largesse of emperors and witnessed a remarkable urban development under Roman sovereignty. The devastating earthquake of 115 AD which destroyed much of the city was followed by a large-scale urban renovation undertaken by both imperial and local patronage. The renovation and urban flourishing took place during the reigns of Trajan (98-117 AD), Hadrian (117-138 AD), Antonius Pius (138-161 AD), and Marcus Aurelius (161-169 AD), and hence continued for more than fifty years. Influential local families and municipal elites contributed to the renovation and embellishment of the city following the earthquake. In return, they received public honors and had their statues erected at the visible spots of the urban terrain to which they contributed (Andrade, 2013: 157). A series of consoles on the columns of the main streets carried the statues of these local benefactors and the emperors (Balty, J. Ch., 1991: 16).
One of the most striking reconstructions after the earthquake was the enlargement of the main streets that formed the basis of the urban layout of the city. The Grand Colonnaded Street, Cardo Maximus, which was 30.00 m wide in origin, and comprised porticos, was enlarged to 37.00 m, and the intersecting Decumanus which had a width of 16.00 m in origin was enlarged to 22.00 m (figs. 68, 69). The Doric orders built in the Hellenistic period, on the other hand, were replaced with Corinthian orders in many sections of the porticos. The columns were 0.90 m in diameter and 9.00 m high. The Cardo Maximus of Apamea that stretched approximately 1850 m from the North to the South Gate, became one of the largest and most impressive colonnaded streets of the eastern empire where such streets were one of the main characteristic features of the Greco-Roman urban landscape.40
40 For colonnaded streets in the eastern cities see MacDonald, 1982: 33-51; Segal, 1997:5-53; Ball, 2000: 261-272 and Burns, 2017.
88
Another important building activity in the course of the 2nd c. AD was the renovation and the expansion of the theatre which was built in the early Seleucid period. This Greek-style theatre was enlarged and its stage and access points were reorganized to convert it into a Roman type in the very first years of the empire (Finlayson, 2012:308). After the earthquake of 115 AD, the destroyed theatre was repaired and received a water catchment system and cement waterproofing for the orchestra floor. After these additions and expansions, the theater of Apamea, with its estimated cavea diameter of 139 m and an estimated seating capacity for at least 20,000 people became the largest edifice of its kind in the Roman East and one of the largest in the whole empire. Moreover, the locations of the catch basins and cisterns, as well as other traces of hydraulic infrastructure beneath the orchestral floor and stage indicate the possibility that the Great Roman theater of Apamea had the potential to host both theatrical and festival water displays (Finlayson, 2012: 292-293).
The change of the water supply system of Apamea in the Roman period also had visible consequences on the urban landscape. While water was largely supplied from the cisterns dug in the ground of the plateau in the Hellenistic era, the city was provided with running water by means of the aqueducts built in the Roman period. Subsequently, the city was adorned with several monumental nymphaea, sometimes in distinct urban locations just like the one at the north gate and the one at the intersection of the two main streets. The city also received many bath buildings in the Roman period (Balty, J. Ch., 2000: 477).
As such, in the course of the 2nd century, Apamea became one of the most ostentatious urban centers in the Near East with several Roman buildings superimposed on the Hellenistic layout. The city continued to prosper in the course of late antiquity, despite several catastrophes such as the invasions and earthquakes that hit the city in the 3rd c. AD. In the early 5th c. AD, Apamea became the capital of the province of Syria Secunda, and its political importance led the city to sustain imperial attention and maintain its rich urban topography.
89
A vast impressive mosaic floor pavement exposed under the portico of Cardo Maximus had an inscription with the date of 469 AD. The mosaic extended along the length of an insula for about 100.00 m and covered all the entire width of the 6.50 m wide portico. Considering the existence of floor mosaics in some parts of the Grand Colonnade too, it can be assumed that a large part of the porticos of the colonnaded street, if not all, was ornamented with mosaics in the second half of the 5th c. AD (Dulière, 1974: 9).
In the course of the 5th and 6th centuries AD, besides its political significance, Apamea became a prominent center for Christianity. With the Christianization of the city, several churches were inserted into the urban topography, often replacing the earlier religious structures. The Atrium Church, for instance, was built in the 5th c. over a 4th c. synagogue and the Eastern Cathedral was built on the site of an earlier pagan temple in the 6th c. Apamea, as such, was adorned with some of the earliest important examples of Byzantine churches in the East (Balty, 1989: 84-89).
In the 6th c. AD especially after the devastating earthquakes of 526 and 528, the city saw an extraordinary building activity. In this period the emperor Justinian not only granted a suspension of taxes but also intervened in the reconstruction of several buildings personally, as was the case in Antioch, Seleucia, and Laodicea (Balty, 1989: 80). Besides the emperor himself, the municipal authority, the church, and wealthy residents participated in the restoration work in the course of the 6th c. AD.
The most remarkable projects of the 6th c. AD were those carried out in the Grand Colonnade, churches, and the private residences of the elite. The Grand Colonnade was closed to cartage traffic and sidewalks were added on both sides of the road under the porticos. As such the main street on the N-S axis of the city that used to take the vehicle transport and lead to Antioch and Chalcis to the north and Raphanea and Hama to the south, from at least the 1st c. AD on was transformed into a monumental pedestrian alley. Moreover, a rich marble pavement was built over the existing mosaic pavements of the late 5th c. AD in the porticos. The same use of marble or opus sectile floors in place of the earlier mosaic pavements is also observable in the restored
90
churches and the private residences, both of which were enlarged and elegantly transformed with respect to the preferences and trends of the period (Balty, 1989).
Following the Justinian era, both the urban character and appearance of Apamea and the lifestyle it framed began to change. The allocation of the Grand Colonnade for pedestrians brought a denser commercial activity under the porticos. The earlier temporary commercial structures like counters and stalls were turned into permanent shops occupying the sidewalks and blocking the porticos by small walls.
In the 7th c., however, the Persian invasions and occupation of about 15 years, a destructive earthquake, and the fall of the city to the Arabs changed the urban landscape irreversibly. From this period onwards, the city staged a more rural life under Islamic rule.41 With two more devastating earthquakes in the 12th c. that caused the collapse of the Grand Colonnade and many other structures, the lower city along the valley was abandoned and the settlement shrank to the hill of the earlier acropolis (Balty, J and J.Ch., 1969: 44-51).
4.2 Domestic Architecture in Apamea on the Orontes
The site of modern Qala’at al Mudiq and the surrounding area has been identified as the ancient city of Apamea since the mid-19th c. AD. The first archaeological explorations at the site started in 1928 by the Belgian archaeologist Franz Cumont and seven campaigns took place before World War II between 1930 and 1938. These were followed by two brief missions in 1947 and 1953. In the very first seasons of the archaeological surveys, some urban residences were detected and some were even partially cleared. However, they were largely neglected and not studied in favor of more monumental public buildings (Balty, J. Ch. 1984a: 12).
41 For the economic and urban transformation of eastern cities after the rise of Islam see Kennedy, 1985; Walmsley, 2012 and Bessard, 2020.
91
From 1965 onwards, with the request of the General Directorate of Antiquities and Museums of Syria, Belgian archeologists resumed the regular missions and constituted the Belgian Centre for Archaeological Research in Apamea of Syria. Jean Charles Balty, professor at the Free University of Brussels, took charge of the excavations for many years, and with his initiation, the domestic architecture of Apamea became a focus of study in the 1970s. In this period, several rich urban mansions were brought to light and studied in detail.
In 2001 the archaeological work at the site was restarted by the Belgian Centre for Archaeological Research in Apamea of Syria, under the direction of Prof. Didier Viviers, with the support of the Free University of Brussels and in collaboration with some other research centers. The focus of the study during the later excavations was the north-eastern district of the city, the western rampart, and the Tycheion and its surroundings. In parallel to these works, a joint American and Syrian mission conducted excavations in the Great Theatre. The fieldwork at Apamea on the Orontes had to be suspended when the Syrian War began in 2011.
Many artifacts and unpublished field notes from the earlier excavations were already destroyed when the University of Brussels was bombed during World War II. The finds of the later campaigns including mosaics, sculptures, and tombstones are now exhibited in two different museums in Brussels. The remains on the site, unfortunately, have been destroyed and looted during the conflict in the region.
So far at least nine dwellings were spotted in the site of the ancient city of Apamea, four of which were largely excavated while the others were partially revealed or only slightly touched (fig.67).42 The four residential buildings that were spotted in the site and named the House of the Aqueduct, the House of the Arab Graffiti, the House with
42 The excavation reports and the related publications of Apamea are almost entirely in French and the excavated houses were denominated accordingly. The names used in this study are translated to English by the author. In the map of Apamea (fig. 67) and in the section titles of the houses, both the French and English names are used together.
92
Two Peristyle and the House of the Doric Peristyle remained unpublished. One of the buildings that were partially excavated during the first campaigns in the 1930s was initially identified as a residence and named as the House of Atrium but it turned out to be a public latrine of considerable dimensions in the later excavations. The unearthed and published houses in Apamea on the Orontes are examined in detail in the following part of this chapter.
4.2.1 The Building with Triclinos (L’Édifice “Au Triclinos”)
The discovery of a well-preserved mosaic with a hunting scene during the excavations of the Eastern Cathedral in 1935, led to the uncovering of one of the most sumptuous residential structures excavated so far in the ancient city of Apamea on the Orontes. This splendid mosaic was laid out on the floor of a large apsidal room providing an inscription that defined the room as a triclinos after which the residential building was named. In 1937, a number of spaces around this triclinos were partially exposed but the archaeological work did not continue (Mayence, 1936: 409-411; 1939: 204). It was in the 1960s that the systematic excavations resumed in the building and a large part of the dwelling was brought to light (Balty, J., 1969; Balty, J. Ch., 1969).
The Building with Triclinos was laid in an insula that touched to the principal Decumanus of the city with its north side, some 400 m east from the Cardo Maximus. It adjoined the building plot of the large public structure to the west, the so-called Eastern Cathedral. The northern limit of the house could not have been identified, since this part of the insula was completely obliterated by the encroachment of later shops and workshops.43 In the current state of excavation, the dwelling covered at least two-thirds of the insula, about 4,500 m2 (fig. 70). The initial entrance door was on the west wall near the south-western corner of the insula.
43 A large room paved with a geometric mosaic pavement belonging to a phase earlier than the Building with Triclinos was recognized in the lower levels of the shops b4, b5, b9, b10 and b12. Another sounding in b37 and b38, on the other hand, revealed no level of occupation under the Arab occupation (Jourdain, 1972: 115-116).
93
The principal rooms of the Building with Triclinos were arranged around a large and almost square courtyard with a Rhodian peristyle.44 The columns of the northern colonnade, being larger in dimension than the others, were approximately 7.00 m high while the eastern and western sides had two superimposed orders for the same height. The southern border of the peristyle courtyard had disappeared, as well as the rooms further south. A sculpted capital of a pilaster in the north colonnade in front of the entrance of room AB and a column shaft with reliefs found on the floor of the courtyard indicated the ornamental character of the peristyle (fig. 71). The northern gallery of the courtyard being wider than the other sides led to one of the principal rooms of the house.
The apsidal room AB was the largest room of the structure, measuring about 160 m2. It was a non-axially planned room and opened to the first three intercolumniations of the northern portico via a large central door and two lateral bays. It was a richly decorated room with a spectacular mosaic pavement and crustae wall decoration composed of marble floral motifs similar to the ones found in Room Q (fig. 72)45. The floor mosaic was composed of a network of double meanders forming frames for a series of alternating square and rectangular panels with iconographic depictions (figs. 73, 74). These panels represented some mythological and philosophical figures in different compositions. Based on the presence of several confrontations in style, décor, and composition this mosaic was dated, roughly, to the mid-4th c. AD (Balty, J. Ch., 1969: 110). It was largely damaged in a later period, by a violent fire that ravaged all the northern wing of the house. Room AB had six symmetrical openings, three on each of its east and west walls through which it communicated with the adjacent spaces. To
44 In his description of the Greek house Vitruvius defines Rhodian peristyle as a colonnade of which the southern portico is higher than the others (Book VI-3). A few number of public and private buildings with Rhodian peristyle courtyards are dated to the Hellenistic period; among them are House XXXIII at Priene (Rumscheid, 2010) and the House of theTrident and the House of the Masks at Delos (Trümper, 1998). Winter argues that it is possible that the ‘Rhodian peristyle’ was possibly developed by Alexandrian architects; and the idea may have originated from the courtyards with an upper storey on only one of the four sides (2006: 173).
45 Crustae is a kind of decorative wall veneer that was mainly composed of marble slabs in Roman architecture.
94
the east, it opened to a rectangular room W and communicated with the inner courtyard P via two windows. To the west, it gave way to AC and AH which formed the first two units of a suite of four rooms that were linearly arranged. This suite was composed of three rectangular rooms with geometric mosaic pavements, AC, AH and AS, and ended with an apsidal room AW that was paved with a mosaic depicting Ge holding a cornucopia. Several mosaic fragments, almost exclusively geometric, coming from upper rooms were found fallen in the floors of AC and AS, as well as the northern gallery of the courtyard which points out the existence of the second floor in this part of the house. All these mosaics, too, were dated to the 4th c. AD (Balty, J. Ch., 1969: 112).
In the northern wing of the house were two open inner courts, P and J that were paved with stone slabs and furnished with narrow and slightly elevated sidewalks. They provided light and air to the surrounding spaces which did not have direct communication with the peristyle courtyard. Court J served two distinct rooms T and B to its north and south and a smaller room K to its east. Room T and its antechamber O formed the largest and most elaborate unit in the north-east corner of the house. The room was paved with a splendid mosaic depicting two amazons hunting on horseback within a geometrical frame, while its antechamber was paved with a simpler mosaic representing seedling of roses both of which were dated to the second half of the 5th c. AD (fig. 75) (Dulière, 1968: 11; Balty, J. Ch., 1969: 112).46 These two rooms were adjoined with a row of smaller units on the north, X, Y, Z and AA, and connected with further rooms in the upper level. Room M that provided passage from the peristyle courtyard to the inner court P gave access to the upper rooms in the north-eastern corner of the house, via a monumental staircase (fig. 76). Besides the door on the western end of the south wall, this staircase opened to the peristyle by means of a screened window. Considering the relatively secluded character of the suite formed by rooms O-T, the annex rooms including the courts P and J and the staircase in M as well
46 A splendid mosaic depicting hunting Amazons was brought to light in a partially excavated Roman villa in the ancient city of Edessa, in the modern Haleplibahçe-Urfa (Özbek and Özcan, 2015).
95
as the theme of the Amazon mosaic, this unit might have been used as a private quarter, perhaps more by women and children.
In reference to such archaeological evidence as masonry techniques and the mosaics, J. Ch. Balty suggested that the design and development of the peristyle courtyard and the surrounding spaces of the Building with Triclinos can be attributed to the 4th c. AD (1969: 111-112). Although the soundings in rooms A, C, D, E, and V revealed the existence of earlier walls on the levels below, their organization was totally different than that of the later ones and hence their destination could not be identified yet. The house underwent some modifications in the 5th c. AD during which rooms O and T received new mosaics. It was in a later period, presumably in the course of the 6th c. AD that the house experienced a more comprehensive renovation which led to a shift in the spatial focus of the structure from the northern section to the eastern one and a modification in the main circulation axis accordingly.
To the north of the peristyle courtyard was apsidal room A that was located at a central position and was slightly smaller than room AB. This room yielded the sumptuous hunting mosaic mentioned above, which had an inscription on the threshold of its western door saying: that “at the time of the very magnificent Apellion, the triclinos was restored, at the month of Gorpiaios, the third indiction, in the year of 851” (figs. 77, 78). The date provided by the inscription corresponds to September 539 AD and is consistent with the coins dating to the reigns of Anastasius (491-518), Justin I (518-527), and Justinian (527-565) that were found as burnt on the surface of the rooms AB, AC, and W (Balty, J., 1969). For this reason, presumably after the violent fire that destroyed the northern section of the house in the 6th century, an extensive restoration was undertaken in the Building with Triclinos. This process led room A to come into prominence while room AB also witnessed a radical change in its layout without losing its splendor.
Room A opened to the eastern portico of the courtyard via a single central door, the east-west axis of which corresponded to the central column of the colonnade, instead of an intercolumniation as was the usual case. This axis led to the vestibule AN on the
96
opposite side of the courtyard, which probably was opened during the restoration process and gave way directly to the western street. Room A communicated with two rooms, C and D, to the south, and with narrow corridor G to the north. Rooms C and D constituted the first two units of a set of linearly placed and communicating rooms, similar to the western annexes of room AB. Corridor G, on the other hand, had direct access from the northern portico and opened to both AB and B through symmetric doors. Room B was also a distinct room as indicated by its larger dimensions and the apsidal layout and communicated with both the peristyle courtyard and the inner court J.
Room AB and its annex room AC, on the other hand, underwent several renovations since they were highly destroyed by the fire. The floor levels of these rooms were raised and new cement floors, probably covered by opus sectile pavements, were established a few centimeters above the earlier mosaic floors. Room AB was enlarged by the removal of the apse and the extension of the room towards the north. Resting against the new north wall of the room, a monumental nymphaeum embellished with three semi-circular niches and small white marble columns was inserted.
A small bath installation, too, was introduced to the back of the annex rooms AS-AW, adjoining the western façade of the house, probably in the 6th c. AD. It consisted of six intercommunicated rooms, AG-AJ-AK-AX-AY-AZ, that seem to have been designed in the Italian fashion with hypocaust system and rooms with different temperatures, to which water was supplied from the reservoirs AF and BA (Balty, 1984: 479). Room AW was transformed to function as an antechamber to this small bath complex. There were also two latrines within the house, a small one in the south-eastern corner recess of room F and a larger one adjoining the western street in BP.
The Building with Triclinos displayed an architectural and decorative richness and ostentation that surpassed its contemporaries unearthed so far in Apamea. In addition to this grandiosity, its proximity to the Eastern Cathedral and the use of the hunting theme in two mosaics, which was associated with power and virtue, led this structure
97
to be identified as the probable official residence of a high dignitary or even the ‘palace’ of the governor of Syria Secunda (Verhoogen, 1964: 14; Balty, J. Ch., 1969: 113; Balty, J., 1969: 35). However, the excavation of many other residential structures that did not differ much in size and layout weakens this possibility as Duval argued (1984). Besides, the scenes of activities associated with the lifestyle of the elite such as outdoor games and hunting already became widespread in the decoration of rich late antique houses (Ellis, 1991: 124, 127; Scott, 1997: 58). Several mansions, most dating to late antiquity, in North Africa and Asia Minor yielded mosaics representing hunting scenes. Thus, although it is apparent that the Building with Triclinos belonged to a well-to-do citizen of Apamea, it is not appropriate to associate it with the residence of a senior officer or a governor with the current state of the evidence.
After at least three centuries of splendor and luxury, the Building with Triclinos witnessed radical transformations, similar to the other residential structures studied so far in Apamea. According to the ceramic finds and coins, dated to the mid-7th c. AD and later, the whole insula was characterized by a utilitarian occupation. The northern part of the insula, along the Decumanus, was occupied by shops and workshops, at least fifty of which have been unearthed so far. The existence of several presses, millstones, cisterns, and basins attested the commercial and artisanal destination of this district. The elaborate spaces of the Building with Triclinos underwent similar transformations too. The intercolumniations of the galleries were blocked, and northern and eastern galleries themselves were subdivided by shear walls. The western gallery, on the other hand, was blocked by an apse with a central passage that should belong to an earlier phase of occupation just like the similar examples built in other structures. New reservoirs were installed in the south-west and north-east corners of the courtyard. Likewise, two stone tanks into the apse of AW, a basin and its flow into AB, two large troughs into P, and a brick furnace into the mosaic floor of room AH were established, while many openings between the rooms were blocked.
98
4.2.2 The House of the Consoles (La Maison aux Consoles)
The House of the Consoles was situated in the south-eastern part of the central district of Apamea, to the north of the Decumanus. Several residential buildings excavated or spotted in the neighboring insulae indicate that this vicinity was a residential district, close to the intersection of the Cardo Maximus and the main Decumanus of the city.
The House of the Consoles was partially excavated in 1937 by Mayence who noticed large stone blocks with molded consoles on the surface of the ground and thought that they belonged to an important edifice (1939: 205-206). The building was named after these stones with consoles but the function of the building could not be determined. It was between the years 1973-1978 that the building was almost fully excavated by J. Balty and was identified as a large residence (J. Balty, 1984).
The House of the Consoles was constructed in the 2nd c. AD and was inhabited for at least eight centuries with several phases of occupation. In the course of this period, it had undergone many functional and architectural transformations but the later rearrangements did not completely obliterate the initial structure. Thus, the House of the Consoles constitutes a good reference for the study of the development of domestic architecture in Apamea under the Roman reign.
The house was laid out on the north of an insula that was adjacent to the main Decumanus to the south. The insula that covered an area of about 6,000 m2 was shared by three dwellings, the House of the Consoles, the House of the Pilasters, and a third unexcavated house, each having a similar surface area of approximately 2,000 m2 (fig. 79). The House of the Consoles extended along the entire width of this insula from west to east, and along nearly one-third of its length from north to south. It constituted almost a rectangle except for its south-west corner where an area of 14.00 m x 7.00 m, which according to J. Balty, was attached to the neighboring house in the original plan rather than in a later modification (figs. 80, 81) (1984: 20).
99
The entrance of the House of the Consoles was located close to the north-west corner of the insula and displayed a monumental scheme with its decorative posts, tripartite lintel resting on the consoles with double scrolls and two cippus on both sides of the door (fig. 82).47 This façade organization was extensively used in the domestic structures of Apamea and also had parallels in several public buildings overlooking the main streets of the city. The monumental entrance, being 2.20 m wide and 3.50 m high, gave way to a large vestibule (Room P-Q) in the north-western corner of the house. This vestibule was subdivided into two unequal parts by two pillars and two columns which according to J. Balty helped to reduce the span of the beam supporting the upper floor (1984: 23). This division also separated the passageway into two, as BA and BB, which were likely intended for different users as mentioned below.
Located in the north-western extremity of the building, the vestibule of the House of the Consoles prevented any direct visual and physical access to the heart of the house and led to the central courtyard through intermediary space BB. The visitor, who entered from the street door had to make two ninety-degree turns to reach the central courtyard. The vestibule, on the other hand, had direct communication with several smaller spaces in its vicinity which were likely to have been destined for service functions. Just to the south of the vestibule, was room BC which had a window opening to the street. This scheme that provided easy access to the entrance door and enabled the observation of the outside indicates the use of room BC as a porters’ lodge. Located next to room BC, two communicating rooms BE and BJ presumably had utilitarian functions too. J. Balty proposed that they were used either as a storage space to keep the supplies or a kitchen (1984: 20). The vestibule also provided direct access to the upper level via room BA that sheltered a staircase and was presumably used by the inhabitants, rather than the visitors. Behind the stairs room BD, communicating with both room BA and the peristyle, provided access to the upper level from the peristyle without making a detour through the vestibule.
47 A cippus (plural: cippi) is a low, round or rectangular pedestal set up by the ancient Romans for differing purposes, such as to function as a milestone or a boundary post.
100
The central peristyle courtyard of the House of the Consoles was the largest and the principal element of the structure around which other spaces were organized in four wings with different proportions and layouts (fig. 84). This was a vast open space, nearly 25.00 m by 18.50 m, with a Rhodian peristyle of Corinthian order with six columns on its short sides and nine columns on its long sides. The height of the columns with the capitals on the east side of the colonnade was nearly 6.50 m and was higher than the ones on the remaining three sides that were nearly 5.10 m high. The eastern portico of the courtyard, measuring 4.00 m in width was also 1.00 m wider than the other three, indicating the prominence of the section behind it. The intercolumniations on this side of the portico were closed with low screens which did not prevent the view but restricted the access to the central part of the courtyard and oriented the users to the northern and southern galleries. Similarly, the northern colonnade of the courtyard was confined by a sort of grid-like design composed of three parallel rods in different heights. Remains of a mosaic pavement were discovered under the northern gallery.
The eastern wing of the House of the Consoles was by far the most elaborate part of the dwelling in terms of size, layout, and decoration. Nevertheless, this was also the most modified part of the house. Room A-A', the largest space of this wing, and also of the entire structure, was located centrally on the axis of the peristyle and opened to it via three bays towards the courtyard. The room was lit by tall windows opening to the street. Nearly one-third of this vast space (A') was separated from the rest (A) with a semicircular arch resting on columns. This division was also marked on the ground by a low step that raised the back part. As such room A' became a kind of an alcove designed at the very end of the house.
Room A-A' communicated with several side spaces to its north and south via seven doors that were placed more or less symmetrically. Two of these doors placed at the back of the room were larger in size and flanked by niches. These doors gave way to two large rooms, AE and D-D', with different orientations, which were obviously the most prominent spaces within this arrangement. Similarly, room AD-AC with its
101
central location with respect to A-A' must have been a privileged space too. At the back of AD-AC, were two small rooms, AH and AG, which were accessed through large room AE. Room AH had a side door opening to the street and presumably functioned as a service space together with room AG. At the southern end of this section, on the other hand, was an elongated space, L-G that was probably a backyard (Balty, J., 1984: 23). Attached to this open space was room E which was located in the southern wing of the house but was accessed only from the eastern wing. The small rooms AB-AF and B, as well as room C, on the other hand, probably constituted passages with their openings. Remains of mosaic pavements dated to different periods were yielded in many spaces of the eastern wing, including large rooms A-A', AD-AC, D, and the probable passage spaces B and AB-AF. They attested to the splendid decoration of this section applied within the long period of occupation.
The northern wing of the house was the narrowest one with a row of five rooms and was roughly symmetrical with the wider and more monumental southern wing. In addition to rooms BA and BD which led upstairs, this wing contained a set of three communicating rooms. Among these rooms, only the central and larger one, room BH, opened to the peristyle and side rooms BG and BK remained relatively secluded. Room BG, however, was decorated with a mosaic pavement and had a window opening to the street to the north. J. Balty recalled the possibility that this suite may have been reserved for women while the southern section was intended for men’s use (1984: 20). Nevertheless, no specific archaeological evidence was found during the excavations to support this gender-based spatial differentiation.
The southern wing of the house constituted a second prominent part in the house, after the eastern wing. It included a large room (F-H-J) which displayed a monumental character with its elongated proportions and the tripartite scheme. The central part of this room was opened to the south portico of the peristyle via a triple entrance designed by two columns. To the east of this room was a large space separated by two pillars and two columns, and to the west a small space was created by three smaller columns. The central placement of this room on the north-south axis of the peristyle court and
102
its communication with it, as well as its dimension and elaborate layout indicates the importance of this room in the routine of the house. It was likely used as a reception suite or a ceremonial dining room as J. Balty and J. Ch. Balty suggested (Balty, J., 1984: 22; Balty, J. Ch., 1997: 292).
Room BF can also be considered in relation to the southern wing of the house. Although it was situated in the western section of the house that was dominated by the vestibule and the probable service spaces, room BF differed with its larger dimensions and more ornamental layout. It had two sections separated by a wide arch carried by two columns and had direct access to the eastern portico of the peristyle. It was also in the close vicinity of the monumental reception suite which was reserved for men, according to J. Balty’s unjustified hypothesis. Thus, room BF could have served as a smaller reception room or as the office of the owner whose visitors would come from the vestibule and pass along the western portico of the peristyle courtyard, perceiving the ostentation of the house without entering into its core.
Archaeological evidence suggests the existence of an upper floor, in at least some parts of the house. Several columns found during the excavations and did not belong to the peristyle itself attested to the presence of an upper gallery above the northern, southern, and western porticos (J. Balty, 1984: 23). The existence of a staircase in room BA also indicates the existence of upper rooms at least in the northern and western sections of the house. The eastern portico and the rooms behind it, on the other hand, were higher than the remaining parts so they did not likely carry an upper floor.
The House of the Consoles maintained its original plan arrangement without major changes for at least four centuries. In the course of the 6th c., probably after the successive earthquakes of 526 and 528, the dwelling underwent a relatively large-scale restoration process which did not change its architectural layout and functional organization but enhanced its splendor. In this phase, the peristyle courtyard received a large rectangular water feature encompassing the four central columns of the western portico. While satisfying the need of water supply, the reservoir must have functioned
103
also as an element of decoration and pleasure serving both the peristyle itself and the large reception hall A-A' that was on the opposite side of the courtyard on the same axis.48 The floor and wall decorations of room A-A' were also renewed within this restoration process. The mosaic floor pavement of the room was covered with a sumptuous opus sectile pavement which presented a much richer repertoire in A' (fig. 85). The walls of the room were also decorated with elaborate opus sectile panels alternating with painted panels (fig. 86). Room A received a rectangular fountain, the remains of which were detected at the center of its floor. The mosaic pavements of rooms AD-AC and BG probably dated to the same period. Another significant modification of the 6th c. took place in the north of the structure, where a new room, AJ, was created by encroaching on the street. This room was situated above the sewer in the street and was accessed from a new door opened on the north wall of room AG. This room included a small latrine and a pipe to the east (AK) which communicated with an upstairs room.
The archaeological finds including several coins found in rooms A-A', AJ and the peristyle attest that the House of the Consoles was occupied as such at least until the end of the first quarter of the 7th c. AD (J. Balty, 1984: 36; J. Ch. Balty, 1984b: 497). From this period on, however, the house underwent radical changes and gradually transformed into a rural type of habitat. The process was probably accelerated by the Arab conquest and the damages of the earthquake of 658. J. Balty proposed a period corresponding to the 7th c. for the abandonment of the house and its rehabitation by new occupants of a different social class (1984: 37).49
In the course of this late phase of occupation, the upper floor had collapsed and went out of use. The northern portico of the peristyle courtyard was partly blocked and an apse was built in its eastern end (fig. 87). As such an elongated space with three
48 A niche revealed at the back wall of the masonry of the water feature supports the possibility of a decorative function (J. Balty, 1984: 34, 40; J. Ch. Balty, 1989: 90).
49 For the social, economic and architectural transformation of the domestic contexts in the early Islamic period see Bessard, 2020: 74-98 and Polci, 2003.
104
entrances was provided. Similarly, several intercolumniations of the southern portico were blocked and a large furnace was laid at its center. A new well was dug in the southern end of the western portico. Several rooms were subdivided by new walls and some were transformed into workshops. All the passages from the peristyle to the eastern wing and the southern wing were closed to provide separate dwelling units. The bones of a dog, several donkeys, horses, and a camel were found under the collapse of the columns of the colonnade. Thirteen Arab coins found during the excavations attest to the continuity of habitation in the house at least until the 9th c.
4.2.3 The House of the Pilasters (La Maison aux Pilastres)
The House of the Pilasters was laid out between the House of the Consoles and a third unexcavated dwelling in the same insula. The southern part of the structure remains under the modern road passing through the whole insula from east to west and has not been revealed. In the excavated part, on the other hand, the initial structure was highly demolished by later modifications which make the original plan and the date of its construction difficult to specify (figs. 98, 89). Yet, some architectural and decorative elements that still stand in-situ enabled the excavators to reconstitute the plan of the more or less original layout (fig. 90). The earliest datable element of the house was the mosaic pavement of room AI that belonged to the end of the 4th c. or the beginning of the 5th c., and provided a terminus ante quem for the original plan (Gisler and Huwiler, 1984: 88).
The House of the Pilasters was a more modest structure than the neighboring House of the Consoles in terms of size but it was not less elaborate in its spatial layout and decoration. It was also organized around a large central peristyle courtyard, but the location and the orientation of the main spaces with reference to the main axis of the peristyle differed.
The main entrance of the house was no doubt from the western street but it could not be localized with certainty. Room AC which included a staircase and an entrance door
105
and dated to a later phase of occupation, likely served as the vestibule of the house in the earlier stages as well. Nevertheless, the possibility of an entrance door and a vestibule to the south, and thus remaining under the modern road should not be excluded. Room AC communicated with the adjacent room AD in which remains of a staircase were preserved too. The room provided direct passage to the peristyle which was not very common in the Apamean houses.
The house was named after a large number of column bases and capitals of pilasters which originally belonged to the peristyle courtyard and were reused as spolia in the later walls. The eastern, northern, and western sides of the peristyle kept the large blocks of a stylobate, on which no column bases were preserved. On the southern side, on the other hand, there was no stylobate but found five column bases in-situ, each resting on a high rectangular plinth of about 1.00 m x 1.00 m. Considering the intervals between the preserved column bases and their diameters, it was presumed to have been a Rhodian type peristyle with fourteen columns in total, and with a higher order in the south (Gisler and Huwiler, 1984: 80-81). The gallery behind the more monumental southern colonnade was not wider than the other three, as was usually the case in many houses, but it was decorated with a geometric mosaic dated to the beginning of the 5th c. AD.
The elaboration of the southern section of the colonnade and the embellishments of the gallery implied the importance of the room(s) behind. Room AI was the prominent space of this section which probably extended long enough to stay under the modern road. However, it was not situated on the central axis of the peristyle, as would be expected, but was centered at the first intercolumniation of the southern portico and opened to it via a 2.60 m wide central bay and two 1.00 m wide lateral bays, the jambs of which were decorated with painted plasters. The room yielded a relatively well-preserved geometric mosaic which was dated to the end of the 4th c. AD. It was also flanked by two small rooms, AA to the west and another one to the east.
106
To the north of the peristyle courtyard a second distinct room, room A, was situated almost symmetrical to room AI. This room displayed a monumental character with its elaborate dimensions, layout, and ornamentation. It was accessed from the northern gallery of the courtyard via a triple door. The floor was elevated approximately 0.60 m at the back of the room and was accessed by four steps (fig. 91). This elevated part was also separated and emphasized by a wide arch resting on two columns placed on the second and third steps, 1.45 m away from the side walls. The southern part of the room yielded few in-situ tesserae, which indicated the presence of a mosaic pavement. Room A communicated with side rooms AN and I via two identical and symmetrically placed large bays that were flanked by pilasters. The large room I was identified as an open courtyard by Vanesse (2014: 494). It had a direct connection with the peristyle and had two windows opening to the backyard to the east. Room AN was an elongated room and communicated with the small room AP at the back. With its architectural and decorative attributes room A presents resemblances with room A-A' in the House of the Consoles but its location and orientation with respect to the peristyle courtyard were different. It was neither situated axially within the overall layout nor was located behind the more monumental section of the Rhodian peristyle. This divergence from the common practice recalls the possibility of a later modification in that part of the structure, during which an existing room in the original plan might have been enlarged and modified to create a new reception room. However, the current archaeological data is not sufficient to make such an inference with certainty.50
The eastern wing of the house was articulated around large room L, which was situated centrally on the east-west axis of the peristyle courtyard. This room provided passage from the peristyle to the large backyard via two central doors on its west and east walls and was flanked by four rooms on both sides. To the south, it overlooked two small interconnected rooms X and Y, and to the north another small room M and a larger
50 Considering the layout of room A in connection with the corresponding recess of the House of the Consoles, a later modification seems not unlikely, although Balty tends to relate this irregularity to the original plan as mentioned above (1984: 20). Both houses were presumably damaged by the earthquakes of 526 and 528 and were partly rebuilt with some modifications in the course of the 6th c. AD. Thus, it is possible that if this corner of the House of the Consoles had collapsed by the earthquakes it might have been rebuilt in favor of the neighboring house in a reconstruction phase during which the surrounding rooms underwent some modifications too.
107
room T. Room T was accessed from room L by a triple door formed by two columns and had access to the peristyle courtyard through a narrow corridor at the back of room M (fig. 92). Room M, itself was also opened to this corridor via two arched windows. As the archaeological evidence suggests, this part of the house was highly decorated. Room M and the corridor at its back, and presumably also the room T, were paved with geometric mosaic pavements. Many collapsed fragments of painted plasters were found in the floors of the rooms, X, Y, T and L. Room T was also embellished with niches on its west and north walls (fig. 93). This wing opened to a large courtyard to the east which has been only partially opened and was probably extended to the street.
The House of the Pilasters had an upper floor at least above some parts of the structure, as the remains of two staircases in rooms AC and AD and also the large amount of mosaic fragments scattered in the embankment of the rooms X and Y attest (Gisler and Huwiler, 1984: 86).
Two main phases of reconstructions and/or modifications took place in the House of the Pilasters following the initial stage of its construction which established the original plan (fig. 91). The first one, dating to the 6th c. AD, is likely to have been carried out after a partial destruction, probably caused by the earthquakes of 526 and 528 (Gisler and Huwiler, 1984: 88, 94). After this disaster, the demolished parts of the structure were rebuilt mostly with a neat bonding and with respect to the initial layout, but with some touch-ups to introduce some new concepts of ornamentation.
The peristyle was one of the most damaged parts of the house, in the 6th c. AD, with the partial, if not the total collapse of the colonnade. The original layout, however, was respected during its reconstruction, despite the use of some disparate elements. The existence of column bases with different diameters in the colonnade exemplifies such replacements. During this reconstruction, pilasters, mostly recut or resized from the earlier ones, were applied to the corners of the peristyle to give the impression of corner pilasters. A fountain with a semi-circular basin was installed to the eastern end of room I. Connected to this fountain, a reservoir (Rb) and a latrine (Rc) were installed
108
in the backyard. The connection of the fountain with this reservoir and a drain linked with the water channel system under the street attest to the presence of running water. On both sides of this fountain were benches that were about 0.50 m high from the ground and were decorated with floral mosaics. Remains of an opus sectile pavement in room T likely belonged to this phase too. Sometime after these more or less contemporary refinements of the 6th c. and before the latest phase of occupation, some other rearrangements took place within the house (Gisler and Huwiler, 1984: 89). First, a new reservoir (Ra) was adjoined the existing one (Rb) and doubled its water capacity. Secondly, an apse was built in the south-east corner of the peristyle with a neat bonding with limestones. The apse was endowed with a mosaic floor which replaced the earlier one and a non-axially placed door at its back, providing access to room T.
The latest phase of occupation in the House of the Pilasters was manifested by the degradation of the structural elements, fragmentation of spaces, and alteration of circulation patterns. This transformation seems to have started already in the early 7th c. and continued gradually, in a relatively long period of time, until the abandonment of the dwelling two centuries later. In the course of this period, the porticoes of the peristyle were closed and divided with partition walls where the architectural elements of the peristyle, like column drums, capitals and pilasters were largely reused. As such twelve new rooms were created around the courtyard. Likewise, reception room A was divided into six small rooms, and larger rooms L, T, and I were divided into two. The late occupation also encroached the street by the installation of shops adjacent to the façade. Some traces of artisanal activity were also detected in some spaces. Several wells were carved or basins were installed on the floor of many rooms like I, A, and L while the latrine became out of use and the fountain went out of water. In its latest stage, the backyard became a dump, the south-east part of the house was destroyed, many rooms were abandoned and the habitation concentrated in the small dwelling units around the courtyard. The monetary and ceramic finds attest a continuity of habitation in the house until the middle of the 9th c. AD.
109
4.2.4 The House of the Console Capitals (La Maison des Chapiteaux A` Consoles)
The House of the Console Capitals was situated to the north-west of the House of the Consoles and the House of the Pilasters. During three archaeological campaigns between 1977 and 1979, about 2,000 m2 of the building was cleared, which presumably constituted only two-fifth of the entire structure (Baratte, 1984: 107). The peristyle and the northern and eastern parts of the house were partially revealed (fig. 94). The southern facade with the entrance door and the eastern limit of the house were also specified while the northern area was hypothetically determined. As such it is assured that the house occupied a surface area of at least 4,500 m2, extending on an area occupying the three quarters of the insula. With its vast dimensions, the House of the Console Capitals is the largest residential structure excavated so far in Apamea.
The construction of the House of the Console Capitals dated back to the first decades of the 2nd c. AD and just like all the other excavated houses of Apamea, it was inhabited for several centuries (Baratte, 1984: 107). The structural and architectural modifications that were carried out in the original structure during this long period of occupation, gave rise to the formation of new dwelling units with disparate spatial organizations and functions in its later phases (fig. 95).
The peristyle courtyard constituted the central and the most impressive space of the structure with its exceptional dimensions. It extended 56.00 m x 24.00 m, covering an area of approximately 1,350 m2. Although the southern section of the courtyard remains unexcavated, the known three corners of the stylobate and the elements in-situ suggested that the peristyle had a Corinthian order with six columns on each of its short sides and eighteen columns on each of the long sides. The colonnade had collapsed and remodeled probably after an earthquake with some modifications, and the current disposition in this sense, does not belong to the initial phase. Some arrangements in the northern section, however, indicated that it was likely to be a Rhodian peristyle in origin, with the higher order on the north side (Baretta, 1984:110). The northern gallery was also wider than the others and maintained a very small
110
fragment of a mosaic pavement. Traces of notches and lock holes, on some elements of the northern and western colonnade testify the existence of a kind of a barrier, perhaps a fence or chancel, and a gate that separated the galleries, at least partially, from the central courtyard.
The water feature resting against the stylobate on three sides at the northern end of the peristyle courtyard probably belonged to the initial phase of the structure (Baretta, 1984:110-112). This was an imposing design constituted by two symmetrical polylobed basins dug on the ground, and covering a total area of about 15.20 m x 6.00 m (fig. 96). Each had a depth of 0.80 m and was not raised in elevation. The two basins were separated by a 2.80 m wide lane on the floor level but were communicated with two tunnels carved under this lane. Considering its elaborate layout, location on the main axis of the peristyle, and the large room behind, this water installation obviously had a decorative purpose besides collecting water.
The northern wing of the house looks prominent not only because of its location and layout but also of its structure. The walls were built with very large blocks, some having an exceptional size with a length of nearly 3.00 m and a height of 1.50 m. This wing was organized very symmetrically, around the major room F which was situated on the main axis of the peristyle and opened to it via a large opening of 2.00 m. Room F was narrowed to the north and formed a sort of a large alcove that was surrounded by a paved courtyard on three sides. The alcove communicated with this courtyard via three large openings with presumably double and triple doors on its east, west, and north walls (Baretta, 1984: 116). Room F was flanked by a pair of small rooms on each of its eastern and western sides (rooms B, D, H, and O). These adjacent rooms did not communicate with each other via doors, but via two windows as attested between O and H. Rooms D and H also opened to the peristyle, while room O opened to backyard P. Although the spaces in this part of the house were highly suffered from the late occupation alterations, some traces attested the presence of mosaic pavements in every room, in at least some periods.
111
Along the west side of the north wing of the house, two small rooms were partially exposed. The first room, room Q, opened to the northern gallery of the courtyard and communicated with the next room, room U, via a door. Room U ended with a semi-circular basin that bore the traces of a mosaic pavement at its bottom. The limited extent of the current excavation and the very poor state of preservation in these rooms do not permit to specify their function. However, considering the existence of the apsidal cistern and the pipes in the vicinity, Baretta and Balty recalled the possibility that rooms Q and U might have constituted the first spaces of a small bath, similar to the one found in the Building with Triclinos (Baretta, 1984: 116; J. Ch. Balty, 1984: 479).
In the eastern part of the house, only four spaces were excavated. The largest of these was room I which opened to the peristyle via a large door. Its side walls totally disappeared but it was likely to have measured 10.00-12.00 m wide and 10.00 m deep (Baretta, 1984: 113). To the north of room I was room K/L which extended along the width of the large room. It was accessed from the eastern portico of the peristyle and gave way to rooms M/N and J to the north. Room M/N differed in its layout since it had a semi-circular niche on its south wall, by the entrance door, and a small basin leaning to this wall in front of the niche. Across this installation, a column base was found in-situ which may indicate the existence of an arch that provided access between the two parts of the same room. The chronology of these installations, however, could not be specified. Adjacent room J was accessed only from M/N, at least in its later phase of occupation.
The west wing of the house has not been cleared but the examination of the back wall of the western gallery revealed a triple door, approximately in the middle of the peristyle. This monumental door placed on the east-west axis of the peristyle courtyard indicates the presence of another prominent room in this part of the house.
In the excavated parts of the House of the Console Capitals, no reliable trace for an upper floor was recorded, other than a group of capitals. The house was named after a
112
relatively large number of capitals, found as scattered or reused, with varying size and type and with one side projecting more than the others (fig. 97). The placement and function of these capitals have not been specified but a group which are smaller in size and without decoration could have belonged to an upper gallery (Baratte, 1984: 120).
The House of the Console Capitals seems to have been occupied in late antiquity without a notable change in its plan, except some radical modifications in the peristyle courtyard. Perhaps after a collapse, the colonnade of the courtyard was rearranged and the probable Rhodian peristyle was transformed into a regular Corinthian peristyle. The width of the northern gallery was reduced and two bilobed columns were installed on the corners. The intercolumniations in the eastern gallery, on the other hand, was closed by a low wall of rubble while the first drums of many columns were also replaced by large stone blocks. The mosaic pavement of room E likely belongs to the same rebuilding phase of the northern wing of the house.
In the other end of the courtyard the third, fourth, and fifth columns from the south were removed and along these two intercolumniations, a monumental cistern that was supported by the stylobates in the east and west was installed. This was a massive structure with a wall thickness of approximately 1.00 m and a height of at least 2.00 m (fig. 98). It blocked the passage and the visual perspective in the courtyard, leaving only the galleries free. The basin was endowed with a relatively small fountain and a basin of 4.40 m by 2.00 m in the north. The south face of the basin was also treated in a monumental manner with rectangular grooves or niches carved out in the back wall with double thickness. Considering the vast dimensions and the large water capacity of the reservoir, it seems that it was destined mainly for a utilitarian purpose. Yet, its location, the frontal fountain, and the niches at the back indicate a concern for monumentality and decoration.
The ceramic finds and a considerable number of Umayyad and Abbasid coins testify the existence of active life in the House of the Console Capitals in the 8th and 9th centuries AD. In the course of this period, however, the structure was totally
113
transformed to satisfy the spatial and functional needs of a different way of living. The larger spaces were compartmentalized so as to multiply smaller units, some of which received new utilitarian functions like ateliers. Existing doors were narrowed or entirely blocked. The pavement in some places of the eastern and western galleries was repaired. A small staircase was installed in the eastern gallery that was probably going up to an elevated platform, and a second one was constructed in the south-eastern corner of room F. In the northern gallery, just in front of room F a well was dug out while the large cisterns in the courtyard went out of use. Probably in a later phase, the galleries of the peristyle were more compartmentalized and the courtyard itself was divided into two parts by means of a new wall built at the alignment of the reused column drums (fig. 99). Several stone troughs and hooks pierced to large stones and used to tie animals indicate the rural character of this latest phase of occupation.
4.2.5 The House of the Deer (La Maison du Cerf)
The House of the Deer was situated just to the west of the House of the Console Capitals, from which it was separated by a north-south oriented street. It extended along a little more than half of the insula, having an estimated area of about 3,500 m2 (J. Ch. Balty, 1984b: 472). It shared the same insula with another dwelling to the north, as indicated by the presence of the piers and lintels of a second entrance door, spotted some 23.00 m away from the north-eastern corner of the insula.
The excavations conducted between 1977 and 1981 revealed an area of about 1,400 m2 of the house, including the peristyle courtyard and some of the southern rooms (fig. 100). The monumental entrance door of the house was also spotted on the eastern façade. Few uncovered spaces and archaeological finds affirmed that the House of the Deer was a large and rich peristyle house with an elaborate layout and rich decoration.
The central courtyard of the House of the Deer formed an irregular quadrilateral of about 560 m2, with a Rhodian peristyle. It had an ornamented Corinthian order with grooved columns and pilasters. The eight columns on the southern side were raised on
114
molded quadrangular bases while the others, seven on the north side and presumably nine on the eastern and western sides, rested directly on the stylobate (fig. 101). The southern gallery of the peristyle, with a width varying between 5.30 m and 5.65 m, is the largest of its type that has been brought to light at Apamea so far. The northern gallery, on the other hand, was about 3.80 m wide, and the western gallery varied between 3.00 m to 3.50 m. The floors of the north and south galleries were decorated with mosaics while the eastern and western ones were presumably paved with white stone slabs, just like the central courtyard. The distance between the central columns in the southern and northern colonnades was larger than the others and corresponded to the monumental doors behind.
The southern section of the house was organized around a principal room (room A) that opened to the peristyle courtyard through a large and central door of 3.75 m wide, and two lateral doors of 1.40 m wide. This large room of about 130 m2 was constructed by large stone blocks with a masonry thickness of 0.75 m. It displayed a strict symmetry with the openings on its four walls through which it communicated with the surrounding spaces. The floor was paved with a geometric mosaic that was enriched by figures of different birds like geese, ducks, and pigeons (fig. 102). This pavement was laid directly on an earlier mosaic, presumably in the 6th c. AD (Donnay-Rocmans and Donnay, 1984: 159). In the center of the room stood a sigma table of green marble -verde antico-, the broken fragments of which were found in-situ (fig. 103). The table was supported by a wooden base whose charred remains kept stuck to it (Donnay-Rocmans and Donnay, 1984: 159).
Room A directly gave way to three rooms: G to the south, E to the west, and C to the east. At its back, it opened to room G through a central arched door of 2.60 m wide and two side windows, each with a width of 1.40 m. Room G has not been excavated yet, but a sounding conducted on its north revealed that it was paved with a mosaic floor, too. Smaller and symmetrically located rooms E and C flanked room A on both sides. The wall masonry of these rooms differed from the adjacent room A as they were built with brick and small irregular blocks. The floor of room E was paved with
115
a geometric mosaic and its walls were covered with plasters. Room E gave way to an unexcavated room to its south.
Room C was slightly larger than room E and it commanded the small apsidal room F to its east. It also communicated with courtyard D through a single door, which was partially blocked and turned into a window at a given time. Room C had a mosaic pavement representing a deer on a floral background, after which the house was named (fig. 104). The floor mosaic of room F, on the other hand, presented a purely geometric decoration that was coherent with the apsidal form (fig. 105). This was the earliest dated mosaic brought to light so far in the house and was dated to the 4th or 5th c. AD (Donnay-Rocmans and Donnay, 1984: 162,180). The demolition layer of this room yielded the fragments of another sigma table similar to the one found in Room A (fig. 106). This second table was made up of white marble and surrounded by a molded rim and flow channel for cleaning. Its dimensions exactly corresponded to the semicircular geometric pattern of the mosaic floor. Considering the spatial aspects of the two close rooms with dining tables, it seems likely that Room F was reserved for smaller groups of diners or for more intimate guests, perhaps only for the family, while room A served for more formal and ceremonial banquets (Duval, 1984: 464; Balty, 1997: 291).
Room A had three identical windows on each of its east and west walls through which it communicated with courtyard D and with an unexcavated room respectively (fig. 107). The eastern windows had holes for fixing grills which were probably decorated with iron sconces that were found fallen down in courtyard D (Donnay-Rocmans and Donnay, 1984: 159). Courtyard D was paved with rectangular stones of diverse dimensions. In front of its south wall, found a sewer and a water channel. The courtyard opened to unexcavated rooms to the south and east via two doors, the latter of which was neatly blocked later. The demolition layer of the courtyard yielded several fragments of white limestone cornices which could have come from the coronation of the north and east walls, and numerous fragments of broken ceramic pots (Donnay-Rocmans and Donnay, 1984: 164).
116
Adjoined the apse of room F, was small square room H which opened to two unexcavated spaces on its north and south. There was a bench of 0.60 m high and 0.70 m deep along the west wall of the room, the top of which was covered with 0.40 m x 0.50 m terracotta tiles, while its vertical face was plastered. The floor of the room itself was paved with a flat brick tiling. The demolition layer of this room yielded many remains of kitchenware, including jars of various sizes, bronze utensils, glass vases, and a grindstone. Room H was obviously a kitchen or a storage room annexed to a kitchen and served the dining rooms found in the close distance but did not directly communicate with them.
Many fragments of burnt mosaics were found in almost all of the rooms in the southern section of the house. These mosaics, presenting geometric or floral patterns similar to the ones on the ground floor, probably belonged to the upper rooms that had collapsed during the fire which destroyed the house (Donnay-Rocmans and Donnay, 1984: 164). Some mosaic fragments with a semi-circular motif, found in room F may indicate, according to Donnay-Rocmans and Donnay, the existence of a second apse in an upper room. Likewise, numerous fragments of flat glass within the same deposits point out the existence of glazed windows in the upper rooms (1984: 164).
In the current state of excavation, the House of the Deer does not provide a detailed chronology for the date of its construction and the subsequent occupational phases. The represented layout of the house belongs to a later stage, presumably just before its destruction at the end of the 6th c. or the beginning of the 7th c. AD (Donnay-Rocmans and Donnay, 1984: 155). Most of the mosaic pavements in the house were dated to the 6th c. AD, by the analogy of their style, suggesting a large-scale renovation in the course of this century (Donnay-Rocmans and Donnay, 1984: 162,180). Besides the mosaic pavements, the monumental dimensions and layout of the house, the ornamentation of the peristyle courtyard with fluted columns and high molded bases, and the furniture including the two marble sigma tables and some bronze objects affirm the ostentation of this privileged residence in its latest phase of occupation, prior to its destruction.
117
The thick layer of ash in the demolition layer and traces of fire observed all over the structure, indicate that it was demolished by a violent fire. After this catastrophe, the house witnessed a period of abandonment and was reoccupied in the Arab period. The new inhabitants reused many of the existing spaces without touching the floor level of the late antiquity as was the case in room A and its eastern annexes (fig. 108). On the other hand, the peristyle courtyard was strongly disturbed by this reoccupation. While the existing walls were reused in this phase, new walls were also added to divide the larger spaces. Some walls, such as the south wall of courtyard D and the west wall of room A, were raised. In the peristyle, the columns were blocked with new walls and the north and south galleries were cut by transverse walls. Some doors and windows, including the door and window of room E and the southern door of room H, were blocked.
4.2.6 The House with the Bilobed Columns and The House with the Trilobed Columns (La Maison aux Colonnes Bilobées & La Maison aux Colonnes Trilobées)
The House with the Bilobed Columns and the House with the Trilobed Columns were situated in the north-west part of the city. They were respectively laid out on the third and second insulae to the west of the Cardo Maximus and about 500 m north of one of the main Decumanus. This urban district, in the close vicinity of a prominent crossroad, was largely occupied by sumptuous residential structures.
In 1934, a short survey in the north-west quarter of the city revealed a bilobed column resting on a geometric mosaic, after which the structure was named (fig. 109). In 1970, the archaeological work resumed in the area and the triclinium of the House with the Bilobed Columns was partly revealed. This was a large room with a T-shaped mosaic pavement, probably dating to the 3rd c. AD (Balty, J. Ch., 1972: 21-22). The geometric frame of the pavement has been relatively well preserved while the central panel with a mythological iconography was almost entirely destroyed. The bilobed column that marked the south-east corner of the room and a simple column base to its west were laid directly on the mosaic and obviously belonged to a later phase of the building.
118
The House with the Trilobed Columns, too, was spotted during the early archaeological surveys in the ancient city of Apamea and partly excavated in two brief campaigns in 1978 and 1979. It was a large, two-story and sumptuous residence of about 2,700 m2 which was organized around a peristyle courtyard (fig. 110). It also had a second floor as shown by the staircase in the north-east corner and many mosaic fragments found fallen into different rooms.
The House with the Trilobed Columns was initially entered from a monumental door of 3.25 m wide in the north-eastern corner of the insula. The entrance door gave way to a large paved vestibule, room AA. Next to the vestibule, in room AB, was a staircase resting against the north and east walls. The room with the staircase was illuminated with a window in the wall of the north façade.
The large central courtyard of the house was surrounded by a colonnade with six columns on the north and south sides and seven columns on the west and east sides. The north-west corner of the colonnade was occupied by a trilobed column with a molded base and an Ionic capital after which the house was named. This column was dated to the 2nd or early 3rd c. AD, but considering its form that was maladjusted to its place Raepsaet-Charlier argued that it might have been reused, perhaps in the same space, in a later period (1984: 184).
The courtyard was paved with large limestone slabs of varying dimensions while remains of both mosaic and opus sectile pavements were traced in the floors of the porticos. In the western portico, a mosaic fragment with animal motifs framed by a geometric pattern was preserved. The mosaic depicted a fine-tailed sheep articulated with a beaded necklace and grazing a flowering plant (fig. 111). Above the sheep was a parakeet. At the top of the panel, was inscribed “wild sheep”. This type of animal motif is frequently seen in the mosaics of the second half of the 5th c. and in the 6th c. AD (Dunbabin, 1999: 179-185, Balty, 1977: 9-10). In the eastern end of the southern portico, fragments of an opus sectile revetment composing a multicolored geometric
119
decor were preserved. Fragments of marble revetments were also preserved on the south walls of the southern portico.
The southern portico of the peristyle courtyard, being approximately 4.00 m, was a little wider than the others which were about 3.20 m wide in average. This portico gave way to the most elaborate part of the house that has been only partially exposed but was relatively well preserved. In the southern section of the house was room M, a distinct space with an ample room size of 15.00 m x15.00 m. It was located non-axially according to the courtyard to which it was opened via a large central door and two symmetrical windows. The floor of the room in the north-west part was covered with large stone tiles while the north-western corner yielded a fragment of opus sectile pavement framed by a marble band. For now, it could not have been defined whether these two types of pavements were contemporary or a replacement was done in the course of late antiquity. The walls of the room were also covered with marble veneer, as some in-situ fragments in the southern wall indicated.
Room M was flanked by a row of three rooms to its west. The partition walls of these rooms were more recent constructions than their eastern walls and the wall on the western façade indicated a later rearrangement in the layout of this section, presumably during the late antique occupation. Central room OA functioned as a monumental vestibule that gave access from the western street to large room M through a wide door of 3.20 m. This room was paved with white and black tiles forming a checkerboard pattern. Room O, also opened to large room M and room A to its north. It had a sumptuous decoration of an opus sectile floor pavement and colored marble wall panels (fig. 112). Many sculpted fragments found in the filling of the room also indicate the presence of a decorated molding around the room. Room OB, likewise, communicated with room M via a passage that was demarcated by a doubled wall comprising two resized columns, in a manner to emphasize the monumental character of the passage. Room OB received an opus sectile floor pavement and stucco wall decoration that was contemporary with the rearrangement of this section (Raepsaet-Charlier, 1984: 191). To the east of room M, was another unexcavated but seemingly
120
distinct room, whose entrance from the south-eastern corner of the portico was monumentalized by two columns.
The western wing of the dwelling was occupied by a row of rooms all of which opened to the western portico of the peristyle courtyard while some also communicated with the adjacent spaces. Room A, paved with a mosaic floor, had two doors opening to room B to the north and to room O to the south, the latter of which was blocked during the rearrangement of the southern rooms. The debris of this room included fragments of a table in green marble and remains of small columns in both limestone and colored marble. Room B was also a highly decorated room with a mosaic floor of floral pattern and mural decoration with marble panels. It had a deep basin in brick in its north-east corner and a second shallow one, probably a later one, in its north-west corner. Room N also communicated with the adjacent rooms G and PA.
The northern rooms of the house were only partially excavated. However, in-situ mosaic fragments found in the floor coinciding the passage between rooms L and F point out the decorative aspect of this part of the house too. This mosaic depicted a hunting scene in which a bear could be recognized.
The House with Trilobed Columns, just like the other Apamean residential structures, underwent a series of repairs and rearrangements in late antiquity. The major works of this period included the creation of a new entrance in the western façade and the rearrangement of the surrounding spaces, repair works on the façade walls, installation of mosaics and opus sectile floors as well as marble wall decorations, and the insertion of the basin in Room B. These works, mostly aimed to enhance the architectural and decorative quality of the structure and hence the House with the Trilobed Columns was still a sumptuous residence in its late antique phases.
The final state of its occupation in the Arab era, however, was characterized mainly by the abandonment of certain zones, variation of floor levels, and the general interference of the circulations. During this phase, the southern section of the house
121
including rooms M, O, OA and OB, was abandoned while the rest of the structure underwent radical changes. The floor levels of the late occupation in the southern portico of the peristyle was about 0.30 m above the opus sectile pavement while it was 0.30 m below the mosaic pavement in northern rooms L and F. Many new walls were built to create new rooms, in the porticos of the peristyle courtyard while the existing spaces were divided into smaller units as in the case of rooms C and G. These walls were mostly built by placing large reused blocks like lintels, columns, jambs and capitals vertically and inserting between them a block of rubble, gravel and collected elements like tiles, bricks, ceramics. Two wells were dug in the western and northern porticos of the courtyard, the latter of which was enclosed by a new small room. Remains of two staircases in the northern portico indicate the existence of an upper level in this part of the house.
4.3 Overview of Domestic Architecture in Apamea on the Orontes
The relatively well-preserved residential structures excavated so far at Apamea on the Orontes, constitute a good comparable sample for the study of the domestic context in the ancient city. The remains of these structures not only shed light on the main architectural and decorative features of the urban residences but also provide a good amount of information on the changes and transformations that took place in the private domain throughout the Roman rule, thanks to their long phase of occupation.
The urban houses of Apamea are distinguished especially with their vast dimensions covering a surface area between 2,000 m2 and 4,500 m2 and also with their elaborate architectural layout and decoration. The unique facade designs with very characteristic monumental entrance doors, the equivalents of which were yielded in some public buildings in the city, recalls the possibility that they were designed as part of a large-scale urban development plan which took place in the 2nd c. AD, after the earthquake of 115 (Balty, J. Ch., 1984:472). The entrance doors were located close to the corner of the insula and gave way to non-axially placed vestibules, thereby restraining any
122
direct access to the core of the house.51 In the close proximity of the vestibules were usually located service spaces and staircases leading to upper rooms.
The essential and central element in the architectural layout of the houses was the peristyle courtyard which regulated the articulation of the main rooms and provided light and air for the surrounding spaces. In these vast and monumental peristyle courtyards the Rhodian layout was adopted in which one side of the colonnade was accentuated by the columns that were larger in diameter and height. The gallery behind this privileged colonnade was also enlarged and elaborated in decoration. In some of the houses were secondary courtyards or backyards, without a colonnade, around which relatively more secluded rooms as well as some service spaces were located.
At the back wall of the distinguished gallery of the colonnade, a monumental triple door with a wider central opening and two lateral bays gave access to the most distinct and often axially located room of the house. This room was differentiated by its wall masonry composed of larger stone blocks and dominated the other rooms with its size, layout and decoration. The entrance of the room was also elaborated by wide triple doors, often decorated with pilasters. It was flanked by several smaller but decoration-wise spectacular rooms, sometimes symmetrically located on both sides. The larger annex rooms were also differentiated in their architectural and decorative vocabulary. The sidewalls of the entrance doors or interior walls of some annex rooms were ornated with niches. This distinct room and its annexes obviously served for the reception ceremonies including banqueting, as the remains of a dining table found in-situ in the House of the Deer indicates. As evidenced by a second table in the same house, there might have been several dining rooms used alternatively according to the season or the number, affinity, and even the gender of the diners.
51 Some houses excavated so far in Apamea seem to have had a second service door in their back-yards, although they were not spotted with certainty during excavations.
123
The direct or indirect communication of the reception rooms with secondary courtyards should also be pointed out. The aim of these courtyards could have been either to provide light, air, and a pleasant outdoor setting for the reception rooms, as was likely the case in the backyard C-G-P in the House of the Console Capitals or to function as service spaces, as was probably the case in inner court P in the Building with Triclinos and in court D in the House of the Deer.
The living rooms and the private rooms of the family in many houses were likely to have been planned at the upper levels. Many mosaic fragments and wall decorations fallen from upper levels were yielded during the excavations in many houses, indicating that these rooms, too, displayed a certain level of luxury. On the ground floor, on the other hand, the relatively secluded character of some units also suggests a more private usage. Such units composed of several rooms had limited communication with the peristyle courtyard as rooms BG-BH-BK in the House of the Consoles or had no direct communication as rooms T-O-X-Y-Z-AA, in the Building with Triclinos. In the latter case, the two inner courtyards provided access as well as light and air to this suite of rooms.
The long occupational phases of the dwellings in Apamea necessitated subsequent restorations and renovations of the structures. Until the 6th c. AD, these were relatively limited to minor touches that mostly maintained the principal layout of the house but enhanced its decorative quality. The mosaic pavements dating to different periods between the 3rd and 6th centuries AD within the same house attest a continuous renovation in this sense. In the 6th c. AD, however, probably triggered by the devastating earthquakes of 526 and 528, large-scale and remarkable refurbishment works took place in all the residential buildings in parallel to the restoration works in the urban structures including the Grand Colonnade of the Cardo Maximus. With these comprehensive works, not only the destroyed parts of the residential structures were restored but also the domestic spaces were modified according to the changing needs, preferences, and tastes of the period.
124
It is argued that, in parallel to the changing political, economic and social contexts in late antiquity, the power became more associated with the individuals in the eastern empire (Ellis, 1988, 1991; Brown, 1992; Sodini, 2003; Wickham, 2003). Accordingly, the urban residences of the wealthy aristocrats became much more elaborated in their layout and decoration as seats of individual power and prestige. In the excavated houses of Apamea, the reception spaces were more monumentalized with the addition of apses or alcoves to the end of the rooms. These back sections were sometimes raised with a few steps and separated by means of a wide arch that rested on two columns. In this new layout, the master of the house received his guests in a more ceremonial and autocratic manner at the center of an apse or alcove, sometimes in a higher position emphasizing the increased hierarchical order within the society.52
The opening of new entrance doors which provided direct and/or axial access to the large reception rooms in some houses should also be underlined. In the House with the Trilobed Columns the reception room M became directly accessible from the street through the new vestibule OA, thanks to the rearrangement of its western annexes. Similarly, in the Building with Triclinos, the new door and vestibule AN provided a new and almost axial access to the distinct room A on the opposite site of the peristyle courtyard. In both cases, the new circulation patterns between the street and the reception rooms accentuated the visual and physical prominence of these rooms. The archaeological data also indicates a change in the dining practice and its architectural setting in the latter phases of occupation, at least from the end of the 4th c. AD onwards, in Apamea on the Orontes. Several houses yielded remains of sigma tables with marble tops inserted on wooden frames which replaced the more or less fixed setting of three klinai. This change of habit also led to another significant change, to the disappearance of triclinium in Apamea in the course of the 6th c. AD.53 The only triclinium with the standard setting of a T-shaped mosaic floor was spotted in the House with the Bilobed
52 For the reception ceremonies and spaces in late antiquity see Scott, 1997; Ellis, 1991 and 2004a. For Asia Minor in particular see Çonkır, 2005 and Özgenel, 2007.
53 For the replacement of the triconch with stibadium and sigma tables and the design of the apsidal dining rooms in the late Roman period see Bek, 1983, Dunbabain, 1991, Ellis, 1997, Malmberg, 2012.
125
Columns, which too could have been changed in its layout or even function, since the two column bases and a pillar placed directly on the mosaic floor indicates a division of space into two compartments.
The installation of water features and sanitary units into the residential spaces was common in late antique Apamea. The damages in the aqueducts caused by the earthquakes presumably forced the proprietors to provide personal water supply amenities and hence to built water reservoirs, latrines, and private bath installations in the residences. Besides, water elements with a decorative function were set either in the peristyle courtyards facing the large reception rooms or in the rooms themselves (Vanesse, 2014). These water features not only indicate the need for water supply but also a desire to increase the spatial richness and luxury as well as comfort within the domestic context.
In addition to the architectural vocabulary and the spatial quality, the decoration schemes in the urban houses of Apamea were also modified in late antiquity. In this period the reception spaces were much lavishly decorated with the widespread use of new materials and themes. Marble, a common architectural and decorative material in the monumental public architecture of the early Empire, became widespread especially in the form of opus sectile floors and wall revetments.54 Likewise, the mosaic pavements of the 5th and 6th centuries depicting hunting scenes as the one in the Building with Triclinos find parallels in the houses of Antioch as discussed above and many late antique urban residences excavated in other Roman provinces.55 It is argued that the mythological, allegorical, and heroic themes in this period were used to impose
54 Several late antique houses with marble floor pavements and wall revetments in Asia Minor are excavated. For the ‘Urban Mansion’ at Sagalassos see Uytterhoeven, 2014: 428 and Uytterhoeven et al. 2014: 397; for the ‘North Temenos House’ and the ‘Bishop’s Palace’ at Aphrodisias, see Campbell, 1996: 188 and Berenfeld, 2009; for the ‘House of Ionic Capitals’ at Hierapolis see Whittow, 2001: 141 and D’Andria, 2002: 417; for the ‘Northeast House at the Lycian Acropolis’ at Xanthos see Manière- Lévêque, 2006 and 2007.
55 For the mosaic pavements in North Africa in general see Dunbabin, 1978 and Thebert, 1987; for the mosaic pavements of the Great Palace at Constantinople see Brett, 1942; fo the ‘Late Roman Villa’ at Halicarrnassos see Poulsen, 1997; for the ‘North Temenos House’ at Aphrodisias see Campbell, 1996.
126
the power and the virtue of the owner to the visitors; the ‘superiority’ of the dominus and his companions was indicated by their costumes, equipments and their activity (Thébert, 1987: 401; Ellis, 1991: 124, 127; Scott, 1997: 58; Dunbabin, 2003: 146).
Probably sometime after the modifications that took place in the 6th c. AD but before the latest state of occupation, around 7th c. AD, a noteworthy arrangement took place in the peristyle courtyards of the House of the Consoles, the House of the Pilasters, and the Building with Triclinos. Elongated apsidal rooms were created by blocking the intercolumniations and adding apses to one end in one of the galleries of the colonnades. In the first two houses the apses were oriented towards the east while in the latter one to the north. In all the apses was a door that opened to the next gallery or to the courtyard. The function of these apsidal rooms cannot be specified with certainty but Duval suggested a new form of reception room, the parallels of which he found in the unearthed Roman period houses in Cyrenaica (1984: 470).56 The existence of elongated and apsidal audience halls, often in the perimeter of the late antique domestic structures in Asia Minor, support this possibility (fig. 113)57.
From the early decades of the 7th c. AD on, the weakening of the Byzantine power and influence and the subsequent Arab conquest gave rise to radical changes in the social and economic dynamics in the region, the consequences of which had significant impact on the urban and domestic contexts. The urban residences of the wealthy elite in Apamea were gradually transformed into poor-quality habitats with a rural character. Large residences were divided into smaller units to accommodate more families.58 Courtyards continued to be the major open and regulating spaces shared by
56 For general information on Roman houses excavated in Cyrenaica see Stucchi, 1975; Ward-Perkins et al., 1986; Spinola, 1996; Zelazowski, 2012; Mikocka, 2015; Mikocka and Misiewicz, 2015.
57 Among the late antique houses with elongated apsidal halls in Asia Minor are the Villa above the Theatre at Ephesus (Ellis, 1991 and Baier, 2013), the Northeast House at the Lycian Acropolis at Xanthos (Manière-Lévêque, 2002; 2006; 2007), the Late Antique Residence at Perge (Abbasoğlu, 2001: 184) and the Late Roman Villa at Halicarnassos (Pederson, 1992).
58 The subdivision of large urban residences into smaller poor-quality habitats was a common phenomenon in late antiquity in both the estern and the western empire. In parallel to the major political and economic problems, the urban decline and the fragmentation of the residences started mainly in the
127
multiple units, however with a quite different setting and functioning. The galleries of the peristyles were blocked and divided into smaller rooms, as were the larger rooms. The circulation patterns changed radically with the blocking of many doors and windows and sometimes with the opening of new ones. Many of the spaces received artisanal and commercial functions with the installation of presses, millstones and cisterns. Wells were dug in random places while many of the earlier reservoirs and fountains went out of use. Several troughs and hooks indicate the existence of animals kept within the dwelling units. Consequently, such modest dwellings became inhabited by a different social class and witnessed a disparate way of living. Balty argued that the new inhabitants were most probably the refugees from the countryside who left their mountain villages and settled in the residences of the wealthy aristocratic landowners, who themselves retreated to the provinces of Asia Minor or to Constantinople mainly for economic and security concerns (1984: 498-501). A terminus ante-quem for the occupation in the urban residential quarter of Apamea on the Orontes must have been the early 12th c. when the region was hit by subsequent earthquakes.
5th and the 6th centuries in Italy and the western provinces like Northern Gaul whereas it occurred later, in the 6th and 7th centuries in the eastern provinces including North Africa, Asia Minor and Syria (Ellis, 1988: 566-567; Burns and Eadie, 2001: XVII-XIX; Whittow, 2001: 149; Wickham, 2003: 391-393; Uytterhoeven, 2019).
128
CHAPTER 5
ZEUGMA: TRANSFORMATION IN BETWEEN
5.1 Urban Layout and Development of the Twin Towns on the Euphrates
The two main determinants in establishing the cities of Seleucia and Apamea contemporaneously on the Euphrates should have been the defensive concerns and the control of the bridge. The two defense-based establishments, though being contemporary and contingent, displayed distinct urban characteristics mainly due to the different topographical conditions on both sides of the river (fig. 114). The twin cities also witnessed dissimilar urban developments and occupational phases after their foundation circa 300 BC. For this reason, the urban layout and development of the twin towns should be studied separately.
5.1.1 Urban Layout and Development of Apamea on the Euphrates
The town layout of Apamea was characterized by an orthogonal grid plan enclosed by fortification walls (fig. 115). The main body of the walls had a chevron pattern and was composed of polygonal rubble masonry. The north and east walls consisted of 28 towers, in total, placed in regular intervals. Four gates, each with two towers on either side of a forecourt, were detected on the north and east walls of the city. In the western wall, too, the traces of an opening which were mostly demolished by the river, an opening was identified by magnetometric surveys (Abadie-Reynal, Ergeç, and Bucak, 1999: 347; Abadie-Reynal and Bucak, 2000: 272). This opening, situated on the axis of one of the main streets of the city, was the only one spot along the river bank and was assumed as the “river gate” by the excavators (Abadie-Reynal, Bucak and Bulgan, 2000: 324; Abadie-Reynal and Bucak, 2001: 233).
129
The fortification walls of Apamea enclosed a total area of approximately 45 ha with a possible acropolis on the hill at Tilmusa. The eastern fortification wall was most probably the reference point for the orthogonal plan since the streets ran along either parallel or perpendicular to it. The streets situated on the axis of the gates were more than 10.00 m wide while the others were approximately 5.00 m wide. Longitudinal building plots that were determined by the intersection of the streets at right angles, measured approximately 107.00 m long in the north-south direction and 42.00 m wide in the east-west direction, with minor differences (Abadie-Reynal, Ergeç and Bucak, 1999: 341-343).
The current information on the urban and architectural features of Apamea on the Euphrates within the orthogonal plan is limited. Shortage of time before the city was inundated restrained long-term and comprehensive excavations. Yet, the archaeological surveys conducted so far have revealed that the city of Apamea had a remarkable unity in its urban conception (Abadie-Reynal and Ergeç, 1998: 403; Abadie-Reynal and Bucak, 2000: 272). A quite homogeneous construction technique was observed throughout the city indicating a contemporary or uninterrupted building activity. Among the two main construction phases observed in the city walls and the residential areas, the second phase mostly respected the main principles adopted at the time of the city’s foundation (Abadie-Reynal and Bucak, 2000: 272). Thus, it is likely that the city took its final form shortly after the construction of the walls.
The insulae within the city walls of Apamea were not fully utilized for building. There were empty islands, even in the periphery of the main streets, indicating a relatively short period of occupation. An urban disruption, dating to the end of the 2nd c. BC or the beginning of the 1st c. BC was observed in the region that corresponded with the abandonment of Apamea. The reason for the abandonment could not be precisely determined but according to Abadie-Reynal the continuous risk of floods and/or Parthian invasions as well as the development of Zeugma on the opposite side of the river are the most possible reasons (Abadie-Reynal, 2001b: 20; Abadie-Reynal and Ergeç, 1999: 406).
130
Traces of minor constructions and occupation after the abandonment were detected mainly along the riverbank in the south and in the north-western part of the city. Jörg Wagner recorded a wall in Tilmusa Höyük, which, according to Görkay, could have been a Roman addition to the Hellenistic wall (Wagner, 1976: 76-78; Görkay, 2012a: 276). Besides, on the terrace of the southern wall along the riverbank, remains of a number of Byzantine residences dated to the 5th or 6th c. AD were partially brought to light (Abadie-Reynal and Ergeç, 1998: 402).
The necropolis of Apamea, too, has provided useful information on the occupation of the city after its abandonment in the Hellenistic period. Located on the north-east of the modern village of Tilmusa (modern Keskince) the necropolis extended from the slopes of the hills to the plain where the city was laid out. A considerable number of rock-cut pit graves and inhumations dating to the Hellenistic period were discovered in the necropolis (Abadie-Reynal, Bucak and Bulgan, 2000: 327-337; Abadie-Reynal and Bucak 2001: 234-235). In contrast, except for a few splendid tombs of hypogeum type, Roman tombs and pottery finds were rarely unearthed. The scarcity of evidence indicated that the necropolis of Apamea was out of use, at least partially, in the Roman period. In the Byzantine period, nevertheless, it was densely occupied one more time. Many Byzantine finds, including a considerable number of lamps, have been brought to light so far (Abadie-Reynal and Bucak, 2001: 235). Two Syriac inscriptions, one dated to the 2nd c. AD and the other to the 9th c. AD, found in one of the rock-cut tombs, testified the use of the necropolis in later periods as well (Abadie-Reynal and Ergeç, 1999: 407).
5.1.2 Urban Layout and Development of Seleucia on the Euphrates/Zeugma
Seleucia on the Euphrates/Zeugma was situated on a series of hills and had a naturally secure position (fig. 116). The orthogonal grid plan, which was the main feature of the urban layout of the settlement on the eastern bank of the river, was most probably challenged by the steep topography on the western bank. Except for some parts of the city with buildings similarly oriented, Zeugma displayed an irregular urban layout.
131
Like Apamea on the Euphrates, Zeugma was surrounded by fortification walls in the Hellenistic and Roman periods. Several stretches of the walls were uncovered in different parts of the city according to which the whole circuit could have been hypothetically determined (fig. 117). In the Hellenistic period, the top of Belkıs Tepe was surrounded by an irregular polygonal circuit. Two stretches of walls were attached to this circuit from the north-east and north-west corners and descended towards the river. In the Roman period, two additional circuits were added to the Hellenistic walls to fortify the new urban areas formed by the expansion of the settlement mainly towards the east and west.
The highest point in Zeugma was Belkıs Tepe that rose approximately 300.00 meters over the Euphrates. It had a dominating view of the city and the river and was easily noticed from afar (fig. 118). For this reason, some scholars have suggested that the “fortress of Mesopotamia” to which Strabo referred while mentioning Seleucia was probably this hill (Wagner, 1976: 33; Kennedy, 1998a: 37). Belkıs Tepe comprised the most important sacred area of the Hellenistic and Roman Zeugma. Some archaeological finds revealed in this spot were dated back to the Early Iron Age and indicated that this mound might have had a sacral nature in pre-Hellenistic times as well (Görkay, 2012a: 277; Görkay, 2015: 28). The main structure within the surrounding walls of the hill was a temple that had three architectural phases. The initial construction of the temple was dated to the 2nd or 1st c. BC. In the 1st c. AD, probably after the Commagene Kingdom was annexed to the Roman province of Syria, the earlier temple structure became enclosed by a new one that measured approximately 19.46 m x 42.10 m (Görkay, 2012a: 282). The Roman temple must have been one of the most important and monumental structures of Zeugma since it was depicted on the city coins of the 2nd and 3rd centuries AD (fig. 119). Within the remains of the temple, there was archaeological evidence for a 9th c. or 10th c. structure whose function could not have been determined (Görkay, 2012a: 282).
The Hellenistic settlement of Zeugma was laid out on the bank of the Euphrates to the south-east of Bahçedere. In this zone, Karatepe had a prominent location commanding
132
the river and the city. According to Abadie-Reynal, the origin of the Hellenistic city was the possible acropolis with a defensive role situated on Karatepe, which she argued, might have been the “fortress of Mesopotamia” instead of Belkıs Tepe (Abadie-Reynal et al., 1996: 313; 1997: 349-351; Abadie-Reynal, 2001b: 13). Excavations have revealed that there was a continuous settlement at Karatepe from prehistoric times to the Ottoman period (Abadie-Reynal, et al., 1996: 311; Görkay, 2020a: 24).
The agora and the urban center of Hellenistic Zeugma were located on a flat area, on a spur extending towards the river, to the north-east of Karatepe, named today as İskele Üstü. In that area, several buildings were brought to light so far among which a temple and a “market building” were identified (Early, 2003: 11-17; Görkay, 2012a: 278; Aylward, 2013b: 119-123). Although the other buildings could not be identified securely, their scale, construction type, and architectural features indicated a monumental character and public functions. Residential structures of the Hellenistic period, on the other hand, were scattered among the public buildings, to the east and west of the agora.
Excavations in the valley called Dere 1 to the east of the Hellenistic center showed that the first settlement in that area dates back to the 2nd c. BC. There is no proof of the primary phase of Hellenistic occupation further east. This indicates that Zeugma did not expand much in its early history and preserved its nature as a small settlement. The city saw a visible growth when Apamea was abandoned in the 2nd c. BC.
The bridge that connected the twin towns, as well as the land routes, on both sides of Euphrates should have been an important urban feature of Zeugma, as well as of Apamea. There are several ancient sources mentioning the passage at Zeugma, which indicate the historical significance and recognition of both the city and the bridge in antiquity.59 Nevertheless, the physical appearance of the bridge is not clear in these
59 For the ancient sources that mention Zeugma see Kennedy, 1998b.
133
sources. Nor, any remains that can be precisely associated with river crossing have been revealed yet in archaeological surveys. The bridge of Zeugma is widely considered to be a pontoon bridge or bridge of boats anchored between two piers on either side of the river, rather than a masonry or wooden structure (Abadie-Reynal et al., 1996: 319; Kennedy, 1998a: 41; Abadie-Reynal, 2001b: 18; Aylward, 2013c: 17-21). For the bridge at Zeugma, ancient sources refer to damages and destructions which may have caused several reconstructions, complete rebuildings, and even perhaps shifts in its location (Kennedy, 1998a: 41; Kennedy, 1998b: 139-162 nos. 10, 12 and 74). A permanent structure, on the other hand, would have necessitated not only piers with sufficient height on the banks of the river but also intermediate piers raising over the water level, none of which have been identified so far in the city.
Based on the geographical attributes of the territory, and the results of urban and archaeological surveys Abadie-Reynal proposed that the pontoon bridge of Zeugma is likely to have been located further upstream, in the vicinity of Bahçedere (Abadie-Reynal et al., 1996: 319; Abadie-Reynal, Ergeç and Bucak, 1999: 331; Abadie-Reynal, Bucak and Bulgan, 2000: 281; Abadie-Reynal, 2001b: 18). This zone with immediate proximity to the city had a commanding view of the route coming from the west and also a view of the city to the east. There exists a series of islands in this spot, which makes it so easy to pass the river especially when the water level is low. The continuity and density of the occupation in the vicinity of Bahçedere from the Hellenistic to Early Islamic periods may also be related with the existence of the bridge in that sector. Kennedy suggested that since it was the only bridge for many miles, its approach is likely to have been marked by monumental structures (1998a: 41). The remains of monumental constructions and the presumed city gate in this sector, strengthen the possibility of the existence of the crossing point in this spot (Abadie-Reynal et al., 1996: 311; Abadie-Reynal, Ergeç and Bucak, 1999: 331).
In the 1st c. AD, especially after the arrival of the Legio IV Scythica, Zeugma flourished in terms of its population and prosperity. From this time onwards, new urban sectors were created to meet the accommodation needs of the new community
134
while the existing Hellenistic core of the city continued to be occupied with some changes and transformations. The city expanded towards the south, to the upper slopes of Belkıs Tepe, towards the east comprising the Hellenistic necropolis, and towards the south-west, occupying the At Meydanı plateau (Görkay, 2020b: 29-34).
Roman Zeugma probably had a second public center laid out on the flat area at the northern slopes of Belkıs Tepe. Remains of a barrel-vaulted structure spotted in this area may have belonged to a Roman arch or a bath building (Kennedy, 1998a: 37). To the north-west of this public center, the theatre of the city was located. Excavations conducted in this region revealed that the theatre was built no earlier than the beginning of the 2nd c. AD and was abandoned after the Sasanian invasion in the mid-3rd c. AD. Based on the dimensions of the upper cavea, whose contours are still visible and approximately 55.00 m in diameter, the seating capacity is estimated to have been 5,000-7,000 people (Kennedy, 1998a: 37; Görkay, 2012a: 279). The stadium of the city was presumably located approximately 400 m west of the theatre (Hartmann and Speidel, 2003: 100-126; Hartmann and Speidel 2013: 381-392). This upper part of the city, where monumental public buildings like theater and a possible stadium stood, was close to the Roman military sector which occupied the eastern edge of the city.
Archaeological and geophysical surveys conducted so far at Zeugma revealed Roman military installations in two different locations; on the eastern and western limits of the city. On the upper terrace of the plain east of Belkıs village, two superimposed military camps each covering an area of approximately 11 ha were identified. They were occupied for a short period of time around the mid-1st c. AD. A second and larger military settlement was spotted at the At Meydanı plateau and the surrounding fields extending between Bahçedere and Belkis Tepe. This area comprising 15 -20 ha of terraced land at the eastern edge of the city was bordered by two nearly parallel hills running roughly west to east, enclosed by a fortification wall and separated from the rest of the city. Excavations in this area revealed remains of a variety of buildings and numerous finds including military equipment, coins, Latin inscriptions, and stamped tiles. Most of these tiles belonged to the Legio IV Scythica indicating that the 4th legion
135
was responsible for most of the building activity. The results of the archaeological surveys confirmed several phases of construction and a long-term use in this military sector of the city. Evidence suggests an occupation between the late 1st and mid-3rd c. AD for the At Meydanı plateau and between the end of the 1st c. and at least 4th c. AD for the area north of it. The size of these military installations, however, indicates that they were temporary camps.
The eastern extension of the Roman city was dominated by residential plots which diffused into the Hellenistic necropolis. Remains of substantial public buildings were also brought to light within this sector, in the eastern limits of the city. Among these buildings were a possible temple or stoa, and two baths one of which was considered to belong to a gymnasium complex (Ergeç and Önal, 1998: 419-443; Başgelen and Ergeç, 2000: 32-33; Abadie-Reynal, 2001a: 271; Tobin, 2013: 73-75).
The results of the excavations at Zeugma offer vivid evidence for the destruction of the city in the mid-3rd c. AD. Most of the buildings were burned and demolished during the Sasanian invasions circa 253 AD. The revival of the city was not immediate after the sack. Most of the ruined buildings were never reoccupied and slowly buried under colluvium coming from the slopes of Belkıs Tepe. Several rooms exposed in the western end of the cavea of the theatre showed that a domestic structure occupied that part of the building probably at the end of the 3rd c. or the beginning of the 4th c. AD (Abadie-Reynal, 2006b: 331-333). The lack of finds dating to the 5th and 6th centuries AD prove that this higher part of the city was totally desettled in this period.
There is scarce evidence of habitation in the 4th c. AD in the city. It was in the 5th c. AD that a considerable amount of resettlement in certain parts of the city took place. In this period the occupation area was reduced and concentrated at the core of the Hellenistic and Roman town along the river and in the close vicinity of Bahçedere.
Procopius, in his famous work De Aedificiis, mentioned that Justinian had restored the old walls of Zeugma since they were too low and narrow, resembling walls of loose
136
stones (Aed. 2.9.18-20). Although no evidence for this later restoration has been revealed yet, it is supposed to be a partial renovation, restricted to certain sectors of the city.
Archaeological surveys have suggested that occupation at Zeugma in Late Antique and Byzantine periods were more scattered than in earlier periods. Moreover, there is a hiatus between most of these later structures and the ones of earlier phases which show that the zones that were unoccupied in the Imperial period were preferred rather than building on destruction debris (Tobin, 2013: 72). Most of the late antique and Byzantine remains excavated at Zeugma so far belong to the residential or artisanal structures except a religious building in İskele Üstü and a possible bath building in the western edge of the city (Abadie-Reynal, Ergeç and Bucak, 1999: 323-331). In these constructions, spolia was widely used.
After approximately two centuries of a relative revival, Zeugma saw a violent destruction one more time, in the 7th c. AD, which was probably caused by the Arab invasions. In the early Islamic Period, the earlier city became a much smaller settlement. Traces of Islamic occupation within the city were revealed only in the eastern and western limits of the city.
5.2 Domestic Architecture in Zeugma
The ancient remains in and around Zeugma are named “Belkıs Harabeleri” by the local population who, unaware of their historical value, were living among these remains for hundreds of years. The remains were also noticed by a number of foreign travelers and researchers visiting Anatolia and Mesopotamia in the 19th and 20th centuries (Ainsworth, 1842: 304; Ritter, 1843: 944; Sachau, 1883: 178; von der Osten, 1930: 68). Nevertheless, the remains then could not be associated with the city of Zeugma, since the ancient city had been considered to be situated across Birecik. The American researcher J. R. Metheny and the French archaeologist F. Cumont were the first scholars in the early 20th c, who suggested that these ruins may have belonged to the
137
ancient city of Zeugma (Metheny, 1907: 158-159; Cumont, 1917: 119-142). The German archaeologist Jörg Wagner, who participated in the excavations of Commagene in the 1970s, was the first researcher who carried out detailed surface surveys of the ancient remains in the modern villages of Belkıs and Tilmusa. He found large roof tiles with the stamp of the Legio IV Scythica which has long been supposed as the garrison of Zeugma in the 1st and 2nd centuries AD. With this discovery, the remains in this region became associated with the ancient city of Zeugma (Wagner, 1976).
The history of excavations at Zeugma goes back to the 19th c. when the excavations in Karkamış and Devehöyük were illegally extended there. During these excavations, some mosaics and some sculptural work from Zeugma were sent overseas, to western collections and museums (Wagner, 1976: 85; Ergeç, 2000: 20; Acar, 2000: 9). Though legal measures against illegal cultural trading were introduced after the foundation of the Turkish Republic, the illicit excavations in Zeugma continued until the 1990s especially by the local people since they were encouraged by the traders who could not reach the Euphrates region anymore.
First legal excavations in Zeugma started in 1987 and continued in the following years, by the Gaziantep and the Malatya Museums. In 1994, when it was officially declared that the Birecik Dam will be constructed in the area that would comprise the site of the ancient city of Zeugma, the Gaziantep Museum urged Turkish and foreign scholars to make rescue excavations before the site became flooded with the dam water. As a response to this call, French, Swiss, and German teams contributed to the excavations in Zeugma and Apamea.
In 2000, when the construction of the dam was completed and started to collect water, Zeugma appeared in the headlines of the local and foreign press and attracted unprecedented attention in both the national and international arenas. In the same year, the USA-based foundation the Packard Humanities Institute (PHI), launched and funded a comprehensive rescue campaign in Zeugma. With the support of this
138
foundation, Oxford Archaeology (OA) and Centro di Conservazione Archeologica (CCA) were engaged in concurrent excavation and conservation works. In the late 2000, independent of all other groups, a campaign was organized by a civil initiation called the Zeugma Initiative Group. The rescue excavations at Zeugma were finalized when the reservoir reached its maximum level in October 2000. In 2005 the Turkish Board of Ministers assigned Ankara University to coordinate the excavations and all the scientific works in Zeugma. Since then, more comprehensive and organized excavations have been undertaken by Ankara University under the direction of Prof. Dr. Kutalmış Görkay.
The rescue excavations of 2000 in Zeugma were planned in three zones that were defined according to the timetable of inundation. Region A covered the shore of the river, region B lied between the level 373.00 m and 385.00 m, and region C comprised the area above the water level of the dam where further excavation is possible. Most of the remains unearthed so far at Zeugma came from region B. Today, regions A and B which cover almost one-third of the ancient city of Zeugma, and the whole Apamea, are flooded by water. Most of the mosaics, wall paintings, and finds obtained from the excavations are now displayed in the Zeugma Mosaic Museum in Gaziantep, while some ruins in the upper parts of the ancient city are visible in-situ.
The finds that were brought to light during the rescue campaigns and the subsequent excavations at Zeugma are dominated by residential structures. So far, more than thirty houses were spotted, some of which were partially or almost completely excavated (fig. 120). These houses yielded sumptuous mosaics, wall paintings, and numerous artifacts including numismatic finds, domestic utilities, and military equipment. The architectural, decorative and artefact finds have been issued in detail in both the annual excavation reports and in comprehensive volumes published mainly in Turkish, English, and French by different scholars who participated in the excavations.
Most of the houses excavated at Zeugma were constructed and inhabited in between the 1st and the mid-3rd centuries AD. Some were constructed in the Hellenistic period
139
and had a continuous occupation through the Roman period. Starting from the 1st c. AD, some of these earlier buildings were incorporated into new residential units according to the changing needs and the preferences of the new community. Thus, the Hellenistic houses unearthed so far, which are few in number, offer only little evidence in terms of their architectural features since they were restored or partially or totally destroyed and rebuilt in later periods. After the mid-3rd c. AD, when the Sasanian sack caused violent destruction all over the city, only slight evidence for repair and reoccupation of the houses was detected in the 4th c. AD. By the 5th c., nevertheless, the city witnessed a partial revival and a few houses were built as new structures. The more extensively excavated and well-preserved houses of Zeugma will be examined in the following part of this chapter, in a chronological order, according to their date of construction and phases of occupation (fig. 123).
5.2.1 The House of the Fountain
The best representative of the Hellenistic period houses in Zeugma, of which only a few examples have been brought to light so far, is the House of the Fountain since its original plan was not changed much by later modifications. The house was partially exposed in Trench 11 by Oxford Archaeology, during the rescue excavations in 2000 (fig. 121). Though the time constraint prevented complete excavation and exact dating, the house was dated to the Seleucid or Commagenian period according to its ashlar masonry construction technique (Tobin, 2013: 93).
The House of the Fountain is located on a valley, called Dere 1, just to the east of the Hellenistic public center. It was laid out on an artificial terrace built on the sloped topography, overlooking the Euphrates River. It was separated from another property to the south by an alley and bordered by another probable alley to the east.
The plan of the house was arranged around a central courtyard that is the largest uncovered space of the building. The west side of the courtyard was not fully excavated but assuming that the pattern on the later floor mosaic was symmetrical, the courtyard
140
is likely to have extended further west than room 11D on its south and measured 4.70 m x at least 4.95 m (Tobin, 2013: 93). This latter room was separated from the courtyard by a screen formed by two columns in the center and two pilasters attached to the east and west walls, all placed on a low stylobate.
To the east of the house, there was a row of three rooms. The central one, room 11B, was probably the vestibule since it had a doorway opening to the presumable alley to the east and gave way to the courtyard to the west. This space also communicated with room 11A to its north and perhaps with room 11C to its south, in its original layout.
The courtyard had direct communication with room 11E to its north. This room measured 2.10 m x 1.70 m, and the water features on its east side showed that it functioned as a private latrine. Parts of three other rooms and two cisterns were identified to the north of the courtyard which may indicate that this part of the house comprised the service spaces. The western part of the house was not excavated.
In Roman times, the plan of the House of the Fountain seems not to have changed much despite the modifications carried out especially in the circulation patterns and the decoration. Initially, in the 1st c. AD the southern side of the house was partly rebuilt and the ashlar masonry walls were replaced by pier and panel system. At that time a door was opened to the east of the wall separating rooms 11B and 11C.
In the first half of the 3rd c. AD the house underwent a more comprehensive refurbishment. The major change of that period was the blocking of the door that gave way to the courtyard from the vestibule, which indicates a change in the main access to the core of the house. In the meantime, room 11C received a new door opening to the eastern street. The existing door that connected rooms 11B and 11C was blocked and a new one was opened to the west of the same wall. A terracotta drain was installed across the surface of that room, which probably fed the new fountain in the courtyard.
141
Against the blocked door on the east wall of the courtyard a decorated fountain, after which the house was named, was installed. The walls of the basin were covered with limestone slabs on top of which was a molded cornice. Small Corinthian columns on both sides of the basin top probably carried the structure that covered the basin. The courtyard also received a mosaic pavement of which only the geometric border was exposed on the east side. Traces of wall paintings that belonged to the same period were observed on some parts of the walls.
Rooms 11D, too, was redecorated in that period. The floor was covered with a high-quality figural mosaic depicting Nereids riding on sea monsters (fig. 122). It was laid to be viewed from the south. Dunbabin proposed a date in the 230s or the 240s for the laying of this mosaic (2013: 161). The walls of room 11D were decorated with painted plaster. A door to the east of this room was identified behind which the traces of a mosaic floor was discerned. It is likely that room 11D in the form of a loggia, communicated with the reception spaces that were to be located on the west side of the courtyard. The latrine continued to be used in the Roman period. Its water system was renovated in the first half of the 3rd c. AD and it received a simple geometric mosaic. Nevertheless, at some point before the Sassanid destruction in the middle of the 3rd c., it went out of use, the mosaic floor was covered with mortar and the door between the latrine and the courtyard was blocked. At the same time, probably the fountain, too, went out of use since the two installations shared the same water system (Tobin, 2013: 95).
The House of the Fountain was destroyed, probably set on fire, in the middle of the 3rd c. AD. The destruction layers of the house, especially of Room 11D, contained rich finds including copper alloy armor scales, iron bindings from a chest or box, an iron ladle, and a scattered layer of eight coins dated to 244 AD. After a long period of inoccupation, some building activity, dated to the late antique or early Islamic period, was observed above the destruction layer of the house (Tobin, 2013: 96).
142
5.2.2 The House of the Helmets
In Trench 2, some 160.00 m to the south-east of the House of the Fountain, four contiguous houses were partially excavated by Oxford Archaeology in 2000 (fig. 123). The houses were laid out on three rock-cut terraces, each of which was about 10.00-13.00 m wide. Two of these houses, both built on the lower terrace, the House of the Helmets and the House of the Bull were relatively well preserved. The House of the Helmets is noteworthy as it preserves the traces of its initial Seleucid period construction.
During the rescue excavations, only the southern part of the house was partly uncovered. The south-western part of the building rested against the terrace wall while the south-eastern part was bordered by a dead-end alley which separated it from The House of the Bull to the south.
The House of the Helmets was initially organized around the east and north sides of a colonnaded courtyard. The terrace wall formed the southern limit of the southern portico of the courtyard which indicated that the terracing system belonged to the earliest phase of settlement. The courtyard measured about 6.50 m x 8.50 m and the peristyle had three columns on every four sides resting on a stylobate (fig. 124). The limestone Tuscan columns of the courtyard were replaced by the earlier ones in the Roman period (Tobin, 2013: 78). Beneath the court were three rock-cut cisterns. The north portico of the peristyle, which was wider than the other three, was arranged in the form of a loggia. In the initial configuration of the Seleucid period, a large room to the east of the courtyard seems to have communicated with the peristyle through two doorways and with the alley through a single door on its south wall.
In the early imperial period, according to some archaeological evidence belonging to the Flavian era, renovations took place in the House of the Helmets in which some of the existing walls in ashlar masonry were replaced by pier and panel masonry, and new walls were added to create new spaces (Tobin, 2013: 77). The room to the east of the
143
courtyard was subdivided to create at least two small rooms and the door opening to the alley was blocked. A new doorway was opened on the wall separating the courtyard and the alley. To the west of the courtyard, new rooms were added. Room 2G with a trapezoidal shape and measuring about 3.70 m by 4.60 m opened directly to the courtyard via two openings and was probably a reception room. To its north, it gave way to the rectangular room 2F which had three more doors, one communicating with the courtyard and two with room 2E.
In a later period, probably in between the mid-2nd c. and the early 3rd c. AD, based on the style of the mosaics, some rooms of the house were redecorated (Dunbabin, 2013: 153). The floor of the porticos and the central space of the colonnade in the courtyard and room 2H received simple geometric mosaics. Traces of wall paintings with vegetal motifs from that period were also preserved in room 2G. Some rearrangements were realized in the door opening to the alley from the courtyard. A new threshold to this door was added and its east side on the alley was decorated with a black and white mosaic pavement.
Shortly after these improvements, the House of the Helmets witnessed a certain decline. Sometime before its destruction in the mid-3rd c. AD, several doorways were blocked and many rooms were given over to food preparation and storage. In that period, the northern colonnade of the courtyard and the east door of room 2H was blocked. The new access to the room had to be provided from a doorway on the north wall which was not excavated. A pithos containing vestiges of walnuts and pomegranates and an amphora was found inside the room. In the courtyard, against the north-east corner of the blockage of the colonnade, a mudbrick shelf was built. Several vessels and a stone quern were found on this shelf at the time of destruction. Some fragments of pottery and cooking ware were found on the floor of room 2F and a group of ten vessels was found in Room 2I. One of the two doors of room 2G was also blocked.
144
Clear evidence for a widespread destruction by fire dated to the mid-3rd c. AD was found throughout the house. In the deposit of the destruction layer, especially in the central area of the courtyard, a considerable number of military objects were found. These finds, including a nearly complete iron helmet and fragments of two others after which the house was named, an iron spearhead and a knife, indicate a military occupation in at least the latest phase of the house. There is no evidence for the reoccupation of the House of the Helmets after its destruction.
5.2.3 The House of the Bull
The House of the Bull was located on the same building terrace with the House of the Helmets and the two properties were separated by the alley in between (fig. 123). The House of the Bull was limited by a north-south street on its east and it adjoined the partially excavated House of the Peopled Plaster on its west. The southern part of the house remained unexcavated.
The exact chronology for the construction and the later phases of the house could not be determined due to the lack of stratigraphic data and datable finds, but the Trajanic era was suggested as ante quem for its construction (Tobin, 2013: 81). At least two phases of refurbishment were determined for later phases of occupation. These changes mainly involved the addition of mosaic pavements and painted wall plasters to the existing rooms. Based on the style and execution of the mosaics Dunbabin suggested a period between later 2nd c. and mid-3rd c. AD for these renovations (2013: 157).
The excavated part of the house had a compact plan with rooms organized around a central courtyard that measured about 6.80 m by 6.00 m. The floor of the courtyard was surrounded by a curb and a cistern was located on its south-east corner. The inner part of the curb was covered with a plain mosaic with two surrounding black bands. The narrow areas between the curb and walls were also paved with white mosaics. The eastern wall of the courtyard, as well as a large part of the northern façade wall of the
145
house, was torn down by robbers’ trenches. The other walls carried the traces of two layers of painted plasters. The first layer that consisted of red squares and framed by red and green borders was later covered with a second layer of painting that imitated marble.
In the northern limit of the house, rectangular room 2M opened to the courtyard through a colonnade of two columns (fig. 125). The location and the layout of this room recall rooms 2H in the House of the Helmets and 11D in the House of the Fountain. Two mosaic pavements, one on top of the other were discovered on the floor of this room. The latter one had a polychrome geometric pattern while the earlier one was highly damaged. Room 2M provided communication between rooms to the east and west of the courtyard.
The western part of the House of the Bull comprised a suite of three rooms. Access to this suite was provided through a wide doorway from room 2M. This passage was provided with an arch in a later period, probably during a refurbishment. The remains of a threshold with sockets also indicate the existence of wooden doors. Room 2K, with which room 2M directly communicated, gave way to the two other rooms in that section and probably served as an anteroom (fig. 126). To the south of room 2M, was the slightly larger room 2J, whose floor level was about 1.00 m higher than the former. The door of this room was framed by large ashlar blocks. A graffiti depicting a bull on one of the jambs of the door gave the house its name. To the east, room 2K opened to the vaulted room 2L (fig. 127). The floor of this partially unearthed room was about 0.20 m lower than that of 2M. The lower parts of that room were carved from the bedrock, while the upper walls were made of ashlar masonry. It had a vaulted roof above which was built room 2C of the neighboring house. All three rooms had geometric mosaic pavements and at least two layers of painted wall plaster often decorated with geometric or vegetal motifs, ribbons, and garlands. The limited communication of this suite of three rooms with the other parts of the house provided this eastern wing a certain level of privacy and this may indicate that it has been reserved for the family.
146
To the east, room 2M opened to relatively large room 2N. Since the wall between the two rooms was torn down, the type of passage between them could not be determined. The floor of room 2N was covered with the only figural mosaic found so far within the house. The central rectangular panel of the mosaic, depicting a fishing Eros, was surrounded by a frame with square panels representing alternating Dionysiac masks and birds. The central figure of the mosaic was oriented towards the east of the room. Considering its size and decoration, this room is likely to be a reception room.
Located on the south of room 2N was rectangular room 2O. Two steps on the north-west corner of this room suggest that its floor level was higher than the surrounding rooms. Traces of a mosaic pavement were found on the floor. At some time, definitely after the laying of the mosaic around the courtyard, a latrine was inserted into the south-west side of room 2O. The exterior walls of this barrel-vaulted structure, which were visible from Room O and the courtyard, had painted plasters.
The main reception room of the House of the Bull was probably located on the unexcavated southern part of the house. Two large openings in the southern wall of the courtyard, and a fragment of a mosaic discovered at the south-eastern corner of the trench indicate the existence of at least one decorated room in that part of the house.
The House of the Bull underwent modifications sometime before its destruction in the mid-3rd c., most probably due to the Sasanian attack. In the latest phase of occupation, the north-west part of the colonnade separating room 2M from the courtyard was blocked by reused material. Similarly, the two doorways on the southern wall of the courtyard were blocked. The walls of several rooms and the vault of room 2L were repainted with a rough layer of monochrome plaster in purple-blue color. The graffiti scratched into the walls of the courtyard as well as the ones on the walls of rooms 2J and 2K were attributed to this period too (Tobin, 2013: 83). In the destruction layer of different rooms, several finds were found including a large ceramic jar, an iron ax-head, a door hinge, several nails, fragments of iron bars and grilles to secure windows,
147
and a nearly complete millstone of black basalt. No evidence for rebuilding after its destruction was found in the House of the Bull.
5.2.4 The House of Dionysus and Ariadne
In 1992 a robbers’ tunnel leading to a mosaic on the northern slopes of Ayvaz Tepe, at the eastern sector of the ancient city of Zeugma, was noticed by the guard of the ruins. In the same year, the Gaziantep Museum started a rescue excavation in that area that would continue for several seasons. During this work, six contiguous houses were spotted and partially excavated. Among these, the so-called House of Dionysus and Ariadne was relatively well preserved and published. The remains of this house and that of the adjacent the House of Danae are covered with a modern sheltering structure and open to the visitors today (fig. 128).60
The construction of the House of Dionysus and Ariadne was dated to the 1st c. AD, but the date for later modifications could not have been precisely determined (Ergeç, 1998b: 89). It was laid out on the artificial building terraces partly carved from the bedrock and had different floor levels following the sloped topography (figs. 129, 130). It rested against the bedrock to its south and east and adjoined the House of Danae to its west. It had an upper story located on the upper terrace to the south. The upper part of the house was identified as a separate dwelling and named House A by Görkay who argued that the two houses were incorporated sometime before the Sasanian sack of the city in mid-3rd c. AD (2017: 200, footnote 26).
The southern and the more monumental part of the House of Dionysus and Ariadne was organized around a central courtyard (figs. 131, 132) (room D3 in fig. 131). In its original plan, the courtyard measured about 10.00 m by 11.50 m and had three columns with fluted drums and Doric capitals on each side (fig. 133). The central area of the
60 The mosaic pavements of both houses, except some small ones that were left in situ, were carried to and are exhibited in the Zeugma Mosaic Museum at Gaziantep.
148
colonnade was covered with a geometric mosaic and there was a cistern below. A slight slope on the pavement and the channels carved on the rock indicated that the mosaic floor once functioned as an impluvium and collected water for the cistern (fig. 134) (Ergeç, 1995: 6; Görkay, 2020b: 46).
In a later phase of occupation, the southern colonnade that was originally wider than the other three was blocked by mud-brick walls and the portico behind it was turned into a rectangular room (room D2), of about 7.60 m by 3.80 m (figs. 135). This new room is likely to have functioned as an anteroom for the rock-carved and vaulted room to the south (room D1), and gave way to it through an arched doorway of 2.30 m wide and 2.70 m high. On both sides of the arch remained the rod sockets of a probable wooden door. Just to the east of the doorway, there was a small window, measuring 0.65 m by 0.50 m, to provide light for the carved room. Both rooms were decorated with floor mosaics and painted wall plasters. Room D1 had a geometric floor mosaic while room D2was covered with a figural one representing the wedding of Dionysus and Ariadne that gave the house its name (figs. 136, 137). The figural panel was not oriented towards the room but to the opposite direction. Based on its stylistic features the mosaic was dated to the late 2nd or early 3rd c. AD (Campbell and Ergeç, 1998: 109). The niche on the western wall of room B is likely to have been a lararium, though in the latest phase of occupation it seems to have been used for storage. During the excavations, some metal finds including a bronze spoon, a knife, a padlock, and a dagger were found in the destruction debris near the niche (Başgelen and Ergeç, 2000: 18-29). The small bronze statues of Hermes and Eros that were found in an upper layer of the debris in the same area were attributed to a possible upper room (Ergeç, 1993: 327).
For the function and organization of the suite of two rooms on the southern wing of the house, Görkay suggested a connection with a wedding ceremony (2017: 200; 2015:72). He argued that the carved room might have been the nuptial chamber which might have changed its function and served as a reception room sometime after its
149
initial use (fig. 138).61 The rectangular room in front, according to him, could have served as the vestibule of the chamber where the mosaic would function as a salutatory scene for the newlyweds and would function as a commemorative scene for their entire life.
The western portico of the courtyard of the House of Dionysus and Ariadne gave way to a larger room with a slightly higher floor level (room D10), through two steps and a single doorway. The western part of the room has not been preserved and its floor was destroyed by a water channel in a later period. The standing parts of the walls of the room preserved traces of painted plasters. This room came into prominence with its location and size and may have functioned as a reception room, though no mosaic pavement has been detected on its floor.
Below the western portico of the peristyle was a carved and vaulted room, with a height of 1.85 m, which probably functioned as a cellar. Access to this space was provided via the steps at the northern end of the portico where there was a circular hole closed by a basalt block on the compressed floor.
Located to the south-west of room D19, a staircase with five steps that were carved from the rock led into an upper level where two communicating rooms (rooms D11-D12) were excavated (fig. 139). Painted plasters were identified on some parts of the walls of both rooms. On the southern wall of the western room D12 were two rows of holes that were likely to have carried the beams of the ceiling and a shelving system. On the floor of the same room was carved a 0.40 m deep hole measuring 0.90 m by 0.45 m. Ergeç related this hole with a weaving loom (1998a: 408). Below the collapsed northern part of that room was identified a rock-carved cellar. This part of the house which was located at a middle level is likely to have been used for domestic activities.
61 Görkay based his argument on a text inscribed on the mosaic pavement of a large chamber in a poorly preserved Roman period house unearthed during the salvage excavations at the shoreline of the dam reservoir in Zeugma. The text on the mosaic was an epithalamium, a wedding song.
150
To the east of the courtyard was a trapezoidal space that rested against the bedrock and formed the eastern limit of the house. This space opened to the eastern portico through multiple entrances. Though there is no information about this space on the excavation reports, its location and layout recall the loggia-like spaces in the previously studied houses. To the north of this space were two interconnected rooms (D4-D5), one of which had a cistern.
The floor level of the northernmost part of the house was about 2.50 m lower than the one of the southern parts. One of the entrances of the house was provided from this part. According to the plan on the site room D18 was the vestibule. The floor of this room was of compacted earth on the bedrock and yielded fragments of pottery and iron, bronze and lead objects as well as a bronze coin. Similarly, many pottery fragments were found in room 15. Room D16 that provided the connection between these two rooms was the only room that had a mosaic floor pavement in that section of the house. It seems likely that, at least in its latest phase of occupation, many rooms in this part of the house were used for storage purposes.
The third section of the House of Dionysus and Ariadne that was laid out on the higher terrace to the south presented a more modest layout (fig. 140). This part had its own entrance on the southern façade. The entrance door opened to a small vestibule-like space that gave way to the colonnaded courtyard to the east and to a trapezoidal room to the north. In the south-western corner of the entrance space was located a small latrine.
The upper level of the house had a smaller colonnaded courtyard than the one at the lower level with three columns in total, on two sides, and a simple mosaic in the central area (fig. 143). On the south-west corner of the colonnade was a niche, probably the lararium that intruded on the adjacent space. On the south wall of the courtyard, near the entrance was located a cistern. Two columns to the south of the main colonnade of the courtyard emphasized the space behind (fig. 141).
151
The entrance door of the house provided axial access to room 6 located just across. This room was paved with a geometric mosaic and gave way to the western rooms through a corridor-like space (room 3). To the north of that part were two larger interconnected rooms that could be accessed from both the courtyard and room 3. The lack of published information for the remains prevents us to comment on their function. Still, their size and location overviewing the courtyard below indicate a degree of prominence, and suggesting that they were likely to have been used as reception rooms.
The archaeological evidence demonstrates that the House of Dionysus and Ariadne witnessed destruction and fire in mid-3rd c. AD. After that date, some building activity was observed in some parts of the house. In rooms D10 and D20 new walls were built that randomly divided the earlier spaces. The wall in D20 diagonally cut the room and included a reused column part, which once belonged to the courtyard (Ergeç, 1998a: 410) (fig. 142). Similarly, a water channel which was dated to post-destruction passed through the house in south-west-north-east direction (Ergeç, 1998b: 87).
5.2.5 The House of Poseidon
During the rescue excavations in Zeugma, the archaeologists from the Gaziantep Museum and the Nantes University excavated a row of four contiguous houses in Chantier 12/Trench 8, at the eastern limits of the ancient city. Two of these houses, the House of Poseidon and the House of Euphrates, were almost completely unearthed while the third one named the House without Mosaics was partly revealed (fig. 143). A few rooms at the eastern end of the trench were associated with a fourth dwelling. In fall 2000, all four buildings remained under the water of Birecik Dam. Before the flood, most of the mosaics, wall paintings, and archaeological finds were brought to light and carried to the Gaziantep Archaeological Museum.
The House of Poseidon, the House of the Euphrates, and the neighboring two houses were laid out on an artificial terrace which was first built and occupied in the 1st c. BC
152
or 1st c. AD, probably during the expansion of the city in the early Roman rule (Abadie-Reynal, 2001a: 245; 2012: 233). The teeth and bones of a human skull that were found under the inner court of the House of Euphrates, suggested that this area was a necropolis before the houses were constructed. The terrace witnessed destruction in the 2nd c. AD, soon after which the second phase of occupation took place on it. The overall plan of the first buildings on the terrace could not be specified, without further excavation, but it appeared that their general layout and the orientation were followed by later dwellings that were built or rebuilt in the late 2nd or early 3rd c. AD (Abadie-Reynal, 2001a: 247). No water supply network was found for the first phase which means that the water was provided mainly by cisterns in the early buildings. On the other hand, there was already an organized wastewater evacuation network which partly continued to function in later periods.
The House of Poseidon, the westernmost house on the terrace, as well as the other three houses, were bordered by and accessed from a wide alley, overlooking the Euphrates to the north of the terrace. To the south, it rested on the natural rock. The House comprised two parts that presented similar plan layouts with certain differences in spatial quality and decoration (figs. 143, 144). The eastern part or Part A covered a surface area of approximately 520 m2 while the western part or part B covered an area of about 430 m2 on the ground level.
5.2.5.1 Part A of the House of Poseidon
The eastern part, part A, of the House of Poseidon was organized around a central peristyle courtyard (A6/P9) that was directly accessed from the vestibule (A9). The courtyard measured 17.80 m by 9.35 m and had colonnades standing on a stylobate on its three sides (fig. 145). Its west side had no colonnade since it rested on the eastern wall of the western part of the house. In the earlier phase of the house, the peristyle courtyard was larger and then reduced in size by the construction of western rooms of Part B (Abadie-Reynal, 2012b: 214). The composite columns of the peristyle were 4.38 m high with grooves that started from 1.45 m above the ground level. Fragments
153
of an architrave, column bases, and drums with smaller diameters, Doric capitals, and fragments of mosaic found fallen in the same area indicated that the southern colonnade was more monumental with the second row of columns above and there was an upper gallery above the porticos. Inside the stylobate was a shallow basin of 0.10-0.20 m deep with a mosaic pavement representing Poseidon on his chariot pulled by two seahorses. Below the chariot were the busts of Oceanus and his wife Tethys. The characters were surrounded by several marine animals like dolphins, shrimps, and fishes. The composition on the mosaic was oriented towards the east for the best view from the eastern portico and the rooms behind. The pavement of the porticos, if there was any, had totally disappeared. On the south side of the peristyle, between the two columns, was placed a fountain (figs. 146, 147). The back walls of the fountain, having the height of 1.53 m, and of its basin, having the depth of 0.70 m, had painted plasters decorated with marble imitation and palmette motifs. Painted fragments of a vault with a shell motif found above the ground level of the basin indicated that the top of the fountain was complemented by a half-dome. The back walls of the porticos were also decorated with paintings for which two phases were identified (Barbet, 2005: 75-92). In the first phase that was probably contemporary with the paintings of the fountain, the walls were decorated with imitations of marble panels and of opus sectile. In a later phase, they were repainted with animal and human figures which, according to Barbet, represented the themes that were related to the renewal of nature and of the richness of its gifts (2005: 299). Among the finds discovered in the courtyard were a chariot wheel rim, a broken limestone sundial, and bone fragments decorated with reliefs of garlands and Dionysian scenes.
The communication between part A and part B of the House of Poseidon was provided through the peristyle courtyard A6/P9. A door of 1.10 m wide on its western wall gave way to room B3/P12 of part B. This door was blocked in a later period. Similarly, room B4/P23 of part B opened to this courtyard via a window of 2.20 m wide located 1.25 m above the ground level. Near this window were found iron window grills and fragments of flat glass showing that the window had a glass pane and was secured by an iron grill.
154
To the east of the central peristyle courtyard was a row of four communicating rooms. The second of these (A2/P4) was an open space with an impluvium (figs. 148, 149). It was entered from the main courtyard through two wide doorways, having the width of 2.00 m and 1.73 m. This inner court received a shallow basin of 0.10-0.15 m deep sometime after its construction. At a later date a fountain that rested against the eastern wall and decorated with niches, marble veneers, and paintings was added to that space. The bottom of the basin was decorated with a figural mosaic depicting the mythological scene of Achilles in Skyros. The figure was oriented towards the entrance from the courtyard to the west. In the center of the basin was a bronze water jet surrounded by white marble. The water collected on the basin was directed to the cistern through channels. On the western end of the basin, two columns monumentalized the space. The floors of the porticos, on the northern and western sides of the basin, had mosaic pavements with geometric patterns. On the walls of this space, traces of paintings with depictions of a garden and a fountain basin were preserved (Barbet, 2005: 299). Within the debris of this room, a broken marble statue of a seated man, a ceramic lion head, an iron spearhead, a stone scale, and bronze coins were found.
One of the largest and most ceremonial spaces of the House of Poseidon was the triclinium (A1/P3) (fig. 150). It measured about 9.50 m by 6.80 m and was located at the northern façade of the house. The triclinium did not directly communicate with the peristyle courtyard. Still, the guests had to enter first this central space after the vestibule, then pass through the inner court to enter the triclinium. As such, they were expected to perceive the splendor of the peristyle and then to enjoy the freshness of the inner fountain court (A2/P4) before experiencing the luxury of the triclinium itself. The northern, eastern, and western walls of the room were almost completely torn-off, so it is not possible to know if there was an opening on the street facade through which the banqueters enjoyed the view of the river too. The fountain court gave way to the triclinium through a 2.00 m wide central passage and two 1.60 m wide lateral passages that were separated by 0.78 m wide quadrangular pillars. The passages once had wooden doors with double leaves. The estimated dimensions of this largest room of
155
the house were approximately 9.50 m x 6.80 m. The T-shaped arrangement of its floor mosaic, which led the room to be identified as the triclinium, had two figural panels (fig. 151). The larger, rectangular panel on the south of the room was oriented towards the entrance to be seen by incomers. It represented Pasiphae and Daedalus with three other figures. Beyond this panel a square one depicting Dionysus, Nike, and Bacchus was turned towards the back of the room for the view of the diners. The square figural panel on the mosaic was surrounded by a geometric pattern on three sides to receive the klinai. In the demolition layer of the room numerous fragments of painted plaster and stucco moldings, some belonging to the ceiling, were found (Abadie-Reynal, Bucak and Bulgan, 2000: 289). One of the pillars standing in-situ on the southern wall preserved the traces of paintings imitating marble and stucco decoration.
Room A4, located between vestibule A9 and triclinium A1/P3 is likely to have been an annex room of the latter but since its east wall was totally demolished no communication between them could be identified. The floor of the room had a mosaic pavement with geometric patterns that was highly demolished. Based on its location close to the entrance of the house, Önal suggested that this room had to be a vestibule or waiting room for the honored guests (Önal, 2013: 15; Önal, Jean and Barlak, 2012: 91). Nevertheless, even if there was a passage between this room and the triclinium, it would only provide a lateral entrance to the room from the backside of the diners which looks rather unlikely. Assuming that it communicated with the triclinium and taking into consideration the large number of ceramic fragments found in the demolition layers, this room is more likely to have been an annex room which at a given time could have functioned as a service room.
To the south of fountain court A2/P4 were two successive rooms. The first one, room A5/P5, communicated with southern room A11/P6, and court A2/P4, through two doors and two windows that were symmetrically placed on its north and south walls. It also opened to the peristyle courtyard through a wide window of approximately 3.00 m wide at a height of 0.90 m high from the floor (fig. 152). Near this window, just like the window of room B4/P23, an iron grill and flat glass fragments were found. The
156
threshold of the passage from that room to court A2/P4 was decorated with a geometric mosaic (fig. 153). The floor of the room itself was covered with a figural mosaic depicting Eros and Psyche. This panel was oriented towards the east of the room where a rectangular part with a simpler mosaic pattern was probably reserved for the installation of furniture. The walls of this room were decorated with stucco moldings and paintings imitating marble. In this room, two stone scales, three bronze coins, pieces of marble wings of a statue, legs of a coffee table, and fragments of glass, tiles, and ceramics were found.
Room A11/P6, the last one of the four successive and interconnected rooms, was entered from the southern portico of the peristyle courtyard and from room A5/P5. Its floor was formed by leveling the bedrock. It did not have a mosaic pavement but its walls had painted plasters with plant and garland motifs. Beneath one of the garlands, a painted inscription invoked to Zeus a wish for long life for Germanicus, who was probably the owner of the house (fig. 154). Among the finds in this room were a bone statue of Aphrodite, a part of an amulet, an iron ox, a hammer, and spear points.
The axiality, provided without any relation with the peristyle, in the westernmost part of the House of Poseidon is noteworthy. The windows and the doors on the partition walls of the successive rooms provided a visual and axial continuity. The open space A2/P4 functioned as an annex to the triclinium and enhanced its spatial and climatic quality. At the same time, it separated the most public space of the house from the two more private rooms to the south. According to Abadie-Reynal, the decorative water features were added to this space after its construction to add a touch of western influence (2012b: 235). She associated this space with the atrium “that served as an anteroom to dominus’ apartments” in the western Roman period houses (2012b: 235).
The south-eastern corner of the peristyle courtyard of the house led to a rectangular corridor with painted walls (A12/P36) that gave way to three different rooms. The largest and the axially located of these was carved room A14/P38, measuring approximately 7.20 m x 5.60 m. The floor of the room was paved with a figural mosaic
157
representing Zeus, in the guise of a satyr, seducing Antiope (fig. 155). This central panel was surrounded by a frame in which various animals like peacocks and roosters were depicted. The walls of the room were also decorated with stucco panels and wall paintings, from which red stripes on the white ground were preserved. A figurine of hermaphrodite or Apollo was found in this room.
Room A14/P38 communicated with room A13/P37 through a passage on its north-west corner. The latter could be entered from the corridor too. It was decorated with a mosaic pavement and molded stucco decorations. The figural panel of the mosaic, depicting the birth of Aphrodite, turned towards the south (fig. 156). In the upper part of the panel an inscription, saying “Zosimos of Samosata did”, was most probably the signature of the mosaic artist. The central panel is framed by a mosaic decorated with amours chasing deer, lions, and tigers. The north wall of this room had a 1.95 m wide window overlooking the peristyle (fig. 157). On its west wall a door that was blocked in a later phase, gave way to an open space with an impluvium (A15/P39). This open space was also accessed from the peristyle court. It measured about 5.10 m x 3.30 m and had a central shallow basin covered with a geometric mosaic. There was a cistern on the western end of the basin and latrines were placed to its north.
To the south of the open space with the basin was carved room A16/P40 that was entered by an arcade that bears the traces of a door to close this passage. The height of this room was 1.85 m and there was a niche on its south wall. Traces of a white coating and graffiti representing stars and triangles were found on the walls. This room must have been used as a cellar.
Room A17/P41, the third room that was accessed from corridor A12/P36, was partly carved on the rock. No mosaics or wall paintings were recorded for this room. Three legs of a seat and the broken parts of a marble satyr figurine that had fallen from an upper level were found on the ground.
158
This southernmost zone of the eastern part of the House of Poseidon, located on the farthest distance from the entrance, is likely to have been the more private sector of the ground floor. It might have been reserved for the household. Nevertheless, the two interconnected and highly decorated rooms A13/P37 and A14/P38 could also have been used to receive more intimate guests.
The last room in part A of the House of Poseidon was room A7 that was located next to the vestibule on the front façade of the house. The demolition layer of this room yielded a limestone stele depicting Heracles, several flat tiles, glass, a large number of ceramic fragments, and sixteen bronze coins. Its location on the street façade and the finds it yielded led this space to be identified as a shop (Önal, Jean and Barlak, 2012: 92).
5.2.5.2 Part B of the House of Poseidon
The western part or part B of the House of Poseidon, presented a plan layout similar to that of part A in some respects. The rooms in this zone, too, were organized on three sides of a peristyle courtyard and the placement of the eastern rooms followed an axial order. Nevertheless, the decoration and the layout of these rooms did not display as much splendor as the ones of the adjacent part.
This zone had a separate entrance from the northern street and had a vestibule in the form of an L-shaped corridor (B5a/P28). The peristyle courtyard and through it the main rooms of the house were accessed via a wide passage on the south wall of the horizontal bar of the corridor and hence were approached in a non-axial fashion. The same corridor also gave way to the service spaces to the north.
The peristyle courtyard (B6/P13) had a rectangular shape measuring about 13.00 m x 10.00 m. At its center was a shallow basin, measuring 5.00 m x 5.80 m, with a mosaic pavement displaying a simple geometric pattern (figs. 158, 159). The three sides of the basin were surrounded by columns. The column drums were smooth up to the height
159
of 1.45 m and fluted from this level above, just like the ones that formed the colonnade of in part A. Some twisted column drums found in that area showed that the southern façade of the peristyle had an upper colonnade, too. A fountain made up of limestone was placed between the two columns on the north colonnade. The inner and outer walls of the basin of the fountain that measured about 2.30 m x 1.10 m, were covered with painted plaster while the parapet was covered with marble plates. The mouth of the water outlet was decorated with a bronze lion’s head similar to the one found in A6/P9. On the west and north walls of the courtyard traces of two phases of painted plasters were identified. On the paintings of the first phase, geometric patterns and panels imitating marble were preserved. For the second phase, decoration with a garland and a vase was preserved in the northern wall. The floor of the porticos around the central basin of the courtyard was of beaten earth. There were two cisterns dug into the rock below the peristyle court. One was located in the south-west corner of the court and was fed by the water collected in the basin. The second one was on the south-west corner of the colonnade. On the west side of the basin were found two hearths made of bricks which probably dated to a later phase of occupation. In the peristyle court six stone mortars, several basalt ground stones, and a limestone oil press were discovered.
On the east side of the peristyle court a row of four interconnected rooms ending with a carved room was placed (fig. 160). The northernmost room of this row was room B1/P10 that was located on the street façade between the vestibule and the triclinium and recalled room A4 of part A. It was entered from the vestibule and communicated with room B2 via a door on the western side of its south wall, which seems to have been opened, after the initial construction of the room (Önal, Jean and Barlak, 2012: 107). Since its north wall has not been preserved, the means of communication with the street, if there was any, could not be determined. The floor of this room had a mosaic pavement with a square figural panel surrounded by geometric frames (fig. 161). The central panel representing Dionysus, Skirtos, and Telete was oriented towards the east where a rectangular panel with a simple pattern reveals the placement of the furniture. Traces of wall paintings were observed on the south and the east walls. In this room, several coins and seals as well as fragments of glass and ceramic and an
160
iron padlock were found. The location and decoration of this room, its communication with the triclinium, and the nature of finds, may lead us to suggest that it could be a tablinum, where the dominus of the house accepted his clients and did matters of business.62
Room B2/P11 was the largest and the most lavishly decorated room of part B and probably served as a reception room (fig. 162). It measured about 7.60 m x 5.50 m and communicated with all the adjacent rooms. Besides B1, it opened to the peristyle courtyard through a wide door and to room B3 via two windows and a door. The windows were later turned into built-in cabinets and then totally walled up in subsequent periods. This room was decorated with a floor mosaic and painted wall plasters. The central panel of the mosaic representing Perseus and Andromeda was oriented towards the north of the room (fig. 163). This central panel measuring 1.85 m x 1.80 m was surrounded by a wide frame of geometric bands. The preserved parts of the walls of the room were decorated with male figures standing on a podium in front of closed doors that were framed by columns (fig. 164). The upper parts of the figures were not preserved but their attitude with raised arms indicated that they were servants carrying dishes (Barbet, 2005: 297). The wall paintings on the south wall of the room were covered with a thin layer of whitewash in a later period. In the latest phase, depictions of gladiators were scratched on the same wall. On the east wall, a graffito representing a walking man was drawn. During the excavations, fragments of a limestone altar with a Latin inscription were found in the north-east corner of the room. In addition, the head of a female statue, fragments of figurines, a spear point, and a large number of fragments of wood, flat tiles, and joint covers were discovered in this room.
Room B3/P12 functioned as a passageway between the eastern and western parts of the House of Poseidon in its initial state of occupation. It connected the peristyle
62 For the social and architectural features of the tablinum in Roman hosues see Ellis, 2000: 5-6, 170; Dwyer, 1991: 27; Clarke, 1991: 4; Wallace-Hadrill, 1989: 63; Leach, 1997: 56.
161
courtyards of the two parts through a door of 1.10 m wide on its east wall and a wider door on its west wall. The eastern door was blocked in a later phase of occupation and the circulation between the two parts was confined, at least on the ground floor level. This space also gave way to room B2/P11 to its north and to room B4/P23 to its south (fig. 165). On its south wall, to the east of the passage to room B4/P23 was a 1.00 m wide opening located 0.50 m above the ground. Many ceramic fragments found near this opening led it to be identified as a built-in cabinet instead of a window (Önal, Jean and Barlak, 2012: 115; Önal, 2013: 70). The floor of the room was obtained by leveling the natural rock. The walls were decorated with wall paintings for which three phases were distinguished. The first layer had geometric motifs, the second was monochrome in gray color with some graffiti of Greek letters and the last had garland and bird motifs.
Room B4/P23 was accessed from room B3/P12 and communicated with the rock-carved room B13 to its south. This room overlooked the peristyle of the eastern part of the house through a wide window which remained in use after the door in B3/P12 was blocked. Thus, it can be assumed that the two parts of the House of Poseidon continued to belong to the same owner or to two different owners from the same family, even after their separation. The floor mosaic of this room had two rectangular panels with figural representations that were surrounded by a large band of geometric motifs. On the western panel that measured 0.90 m x 0.85 m, mythological figures of Antiope and a satyr were depicted. The same theme was represented in room A14/P38 too, with much better quality in style and workmanship. The walls of the room were covered with painted plaster. The surface was divided into quadrangular panels each representing garlands and female figures. One of the figures had a gray pale halo. The fragments of painted plasters, mosaics, and roof tiles found fallen on the ground indicated that this part of the house had an upper story. In addition, the head of a terracotta female figurine, coins, weights, an iron folding seat, a bronze lamp, and an iron lock were found in the room. The carved room B13 could not be excavated since its entrance was obstructed by fallen rocks. This suite of two rooms seems to have been used as the restrooms or the bedrooms by the family members.
162
To the south of the peristyle courtyard, two rooms were carved out of the natural rock. The western one, room B9/P27, had a trapezoidal shape with an area of approximately 35 m2. It was entered by two arched doors and through three steps since its ground level was 0.40 m lower than that of the peristyle (fig. 166). The walls of the room were covered with painted plaster and stucco decorations. Room B9/P27 provided passage to another carved room B11 that could also be accessed from the peristyle. The 1.30 m wide door between the two rooms was partly blocked and turned into a window in a later period. Room B11, like room B13, could not be excavated because of the fallen rocks.
The north-western zone of part B of the House of Poseidon was occupied by the service spaces. These were accessed either by the L-shaped vestibule or the peristyle. The north-west corner of the peristyle courtyard gave way to a vaulted room (B10/P29). On the ceiling of this room were two terracotta conduits that served as chimneys. This room gave way to a narrow and long space, B7 that measured 1.14 m by 4.15 m, through an arched door. In these rooms, a marble figurine head, a scale, a vase, a jug, and a terracotta banquet figure were found. In a later phase of occupation, an adobe wall enclosed the north-western corner of the courtyard near the entrance door of room B10/P29. Five pithoi, an amphora, two vases, a stone mortar, three terracotta oil lamps as well as a bronze statue of Mars lying on the floor between the pithoi were found in this room, indicating that this space, probably together with B10/P29 and B11, was used for storage (fig. 167).
The L-shaped corridor B5a/P28 ended with a small rectangular space of approximately 4.00 m2. Near a niche on the north wall of this room were found 3,750 silver coins in two different bags.63
63 The earliest dated coins in the hoard were those of Iulia Domna (193-217 AD) and the latest dated ones were those of Volusianus (251-253 AD) (Önal, 2013: 86).
163
Room B8 was located at the end of corridor B5a/P28 and entered from a door on its north wall. The floor of this room was of beaten earth and it yielded many fragments of tiles and ceramic sherds, terracotta statues, fragments of bone needles, oil lamps, scales, two crashers, and mussel shells that led it to be identified as a kitchen (Önal, Jean and Barlak, 2012: 131; Önal, 2013: 87).
Next to room B8 was room B12 and the adjoined latrines. Room B12 was paved with terracotta tiles, under which a pipeline ran from north to south. An oil lamp, an iron comb, an intact jar, and a large number of ceramic fragments were found in the room and thus it was identified as a bathroom (Önal, Jean and Barlak, 2012: 136; Önal, 2013: 87). Adjacent to the north wall of the bathroom were located the latrines (fig. 168). They were entered from a door on the west wall of B5a that was very close to the entrance door of the western part of the house.
To sum up, the House of Poseidon is one of the largest houses excavated so far in Zeugma. It covered an area of about 950 m2 on the ground floor. The archaeological evidence including column parts, fragments of mosaics, painted plasters, and objects found fallen from an upper level, indicated that at least some parts of the house had a second story. Nevertheless, no staircase was recorded in the excavations. The finds also proved that the house had a wooden roof structure that was covered with both flat and corrugated tiles.
The coin finds discovered in room 5b suggested the date 249 AD for the terminus post quem for destruction for the House of Poseidon. The ashes in the destruction layer pointed out that a fire that probably happened in the mid-3rd c., which could well be associated with the Sasanian raid in c. 253 AD. In its latest phase of occupation, the house underwent some modifications which were more apparent in part B. In this period peristyle courtyard B6/P13 seems to have been used mainly for domestic activities. The oil press, stone mortars, ground stones as well as the furnaces found in this open space suggested that it was used for production purposes and food preparation. The carelessly built storage area on the north-western corner of the
164
courtyard can be associated with the new function of this space but it may also indicate an increased need for storage. The statue of Mars found laid down between the pithoi and the bronze coins hidden in a cabinet in room 5b may well support this idea since they point out anxiety and preparation for an upcoming threat.
After the Sasanian destruction, the building terrace of the House of Poseidon seems to have been largely abandoned permanently. There was only slight evidence for partial reoccupation. The animal bones found in the rock-carved rooms A14 and A16 and the two large cut stones found in front of the entrance of the latter indicated that these rooms were probably used as animal shelters in later periods (Önal, Jean and Barlak, 2012: 101).
5.2.6 The House of Euphrates
The House of Euphrates was located to the east of the House of Poseidon to which it was partly adjoined. The northern parts of the two houses were separated by a dead-end alley while they shared the same wall on their south (figs. 143). The floor level of the house was about 1.00 m lower than that of the House of Poseidon. According to the archaeological finds the initial constructions of the two buildings were more or less contemporary (Abadie-Reynal, 2012b: 191).
The plan of the House of Euphrates was developed around two distinct but communicating zones (fig. 169). The rooms on the northern part were organized around a longitudinal inner court and did not have direct communication with the peristyle courtyard that was the central element of the southern part. There were two entrance doors that provided access to different zones in the building.
The northern part was accessed from the alley to the north of the terrace. The entrance door opened to a linear corridor (C3/P21) that gave way to the surrounding rooms (fig. 170). Traces of paintings with vegetal motifs and two superimposed graffiti on white
165
plaster, depicting a rider on a horse and a man with a sword in front of an altar, were observed on the walls of the corridor (fig. 171).
The south-west corner of corridor C3/P21 led to the largest room (C1/P19) of the house through a door. The room measured 7.40 m x 6.35 m and the T-shaped layout of its floor mosaic attested to its use as a triclinium (fig. 172). The mosaic had two figural panels that were surrounded by geometric patterns. The rectangular one, representing muses, was located to the south of the room and oriented towards the south. In the square panel there were three female figures with their names Paideia, Arete, and Sophia written in Greek which were oriented towards the north. On the south wall of the room traces of two phases of paintings were identified. In the destruction layer of this room a bronze bust of a double-faced Hermes, fragments of bronze vases and a candelabra, sconces, and coins were found. In addition, a large number of mosaic fragments with geometric patterns found fallen from an upper level indicated that the room had an upper story.
The triclinium opened to an elongated court (C2/P20) measuring 8.90 m x 3.00 m, through a door and a window on its south wall. This central court communicated with all the surrounding spaces and provided light and air to them. It was also entered from corridor C3/P21, through an axial path from the entrance door. There was a window on the west wall of the room overlooking the alley. The court had two impluvia with interconnected shallow basins that were decorated with mosaic pavements (figs. 173, 174). The larger basin had three panels each representing a river god. The representation of the God Euphrates in the center panel gave the house its name. This central panel was oriented towards the north, to a blind wall, while the lateral ones were oriented towards the east or west. On the smaller panel of the second basin that was oriented towards the south, Gaia was represented. None of these figural panels were centered to the entrances of the surrounding rooms. Paintings with geometric motifs were preserved on the walls of the court (fig. 175). The area surrounding the basins was about 0.60 m wide and covered with cut stones. This walking path was rather narrow for ambulation but it seems that this inner courtyard had been used as a
166
passage between the northern and southern parts of the house. According to Görkay, the shallow basin was probably used for a practical purpose, such as washing guests’ feet before they reclined on the kline for a banquet (2020b: 48; 2020c: 73).
The eastern part of the northern zone was composed of a row of three interconnected rooms. The northernmost one, room C9/P18, was the first room that could be accessed from the corridor and its location recalls rooms B1/P10 and A4 of the House of Poseidon. Nevertheless, unlike them, it did not communicate with the triclinium and did not have a mosaic pavement. Its floor was of beaten earth but its walls had paintings. The second room of the row, room C8/P17, communicated with room C9/P18 to the north and with room C10/P24 to the south. It could also be accessed from the corridor and it overlooked the inner court through a large window located 1.28 m above the floor level. The floor of the room was covered with a mosaic pavement that was decorated with simple geometric motifs. On the south-west corner of the central mosaic panel, corresponding to the window on the west wall, a figural panel, representing Dionysus and Ariadne was placed (fig. 176). The unusual placement of this panel indicates that the view towards the court was probably the main concern in the decoration and the furniture layout of the room. Among the finds of this room were fragments of ceramics, glass objects, roof tiles, pithoi, and a limestone vase as well as a grindstone and limestone mortar. The southern room C10/P24 had no decoration and was probably used as a store, at least in the last phase of occupation. The floor of the room was beaten earth and there was a bench of 0.90 m wide and 0.70 m high resting against the east wall. In the room, an in-situ pithoi with olive seeds around, fragments of many other pithoi, amphorae, pots, baked bricks, and iron were found.
To the south of the inner court was a row of three rooms, two of which provided passage to the southern part of the house (fig. 177). Larger room C4/P16 could only be accessed from the court. It had a geometric pavement. At the ground level of the room numerous fragments of ceramics and glass, a marble vase as well as the bases and bottoms of other vases and an iron ox were unearthed. Besides, the fragments of
167
pithoi and tiles were found at about 2.00 m above the ground level which were supposed to have fallen from the upper floor (Önal, Jean and Barlak, 2012: 146). Next to room C4/P16 was room C5/P15 that gave access to both the inner court to the north and the peristyle court to the south. It was divided into two by an adobe wall and its floor was of beaten earth. This room probably functioned as a store, too, since it yielded a large number of ceramic fragments, tiles, pieces of a bowl, iron hooks, and nails in the demolition layer. The westernmost room of this row was room P12/C5 that had a beaten earth floor and painted walls with floral motifs. This room is likely to have been a passage that gave way to the northern part of the house, from the second entrance on the western alley. Nevertheless, a large number of ceramic and tile fragments, iron objects, bronze coins, and a ring found in this room also suggested a storage function, at least for its last phase of occupation.
The southern part of the House of Euphrates was accessed from the dead-end alley that ran in between the two adjacent houses-the House of Poseidon and the House of Euphrates. The alley was initially of beaten earth and measured 17.00 m long and 2.00 m wide. At the end of it was the second entrance door of the house that gave direct access to the peristyle courtyard. The 1.27 m long threshold of this door was made up of two steps and its frame bore traces of white plaster. The two mortises on the step showed that the door had two wooden leaves opening inside. In a later phase, this part of the house underwent significant modifications. The open space at the end of the alley was closed and a new entrance door was installed at 3.20 m north of the existing door. As such, by encroaching upon the public space, a vestibule (C16/P1) was created. The new entrance door was more monumental than the earlier one. Its frame was decorated with moldings and there was a cippus of 1.18 m high on one side, which probably had its pair on the other side. The treatment of this door recalls the monumental entrance doors of the residential structures at Apamea on the Orontes. Probably contemporary with these arrangements, the alley was paved with concrete.
The new vestibule (C16/P11) of the house gave way to a small space through a new door on its east wall, while the existing door on the south wall also remained in use.
168
This small space seems to have served as a modest inner vestibule or passageway that gave access to both the peristyle courtyard to the south and to the inner court with the impluvia, through room C15/P2, to the north.
The peristyle courtyard of the House of Euphrates (C6/P22) rested against the natural rock formation to the south of the terrace. It measured about 10.00 m x 12.00 m and had colonnades on its three sides (fig. 178). The columns of the peristyle had smooth drums and Doric capitals and had a height of 3.28 m. In the central area of the colonnade was a shallow basin with a fountain on its west side. The basin measured 9.60 m x 5.80 m and was decorated with simple geometric mosaics. Near the basin, a limestone statue of Athena Parthenos was found. The fountain had the dimensions of 2.20 m x 1.50 m and had a 0.55 m deep basin. On three sides of the basin were hung smaller circular ones. The floors of the western and northern porticos of the peristyle were of beaten earth, the floor of the south portico was cut from the bedrock and the eastern portico was paved with stone. On the walls of the porticos, traces of wall paintings with marble imitation were preserved. On the north-western corner of the courtyard, near the entrance door were placed the latrines. The floor of the latrines was covered with large tiles, on sides of which a mosaic pavement was preserved in places. To the south of the latrines, on the western side of the courtyard was a cistern.
The peristyle courtyard of the House of Euphrates was surrounded by several rooms on its east and south sides. Room C12/P25 on the east had a two-partite layout. The partition wall of the room and the eastern wall of the north section was constructed with reused material including a fragment of a marble statue. Thus, this room should have been repaired and divided into two in a later phase of occupation. Both parts of the room opened to the peristyle. Their floor was of beaten earth and they yielded a large number of ceramic fragments, a terracotta oil lamp, an amphora, a vase, and a basalt hand mill.
Room C12/P25 communicated with room C13/P26 to the south. This latter room overlooked the peristyle through a 1.40 m window (fig. 179). The floor of the room as
169
well as its south and east walls were formed by leveling the bedrock. The walls of the room were preserved up to 2.90 m in some places and had painted plasters in relatively good condition. On the paintings, several standing young women, two with names of Penelope and Deidamia, were represented (fig. 180). Barbet interpreted this decoration as the representation of the seven famous women known for their fidelity to their husbands and chosen from the heroines of Iliad and Odyssey (2005: 292). The decoration and placement of this room opening to the peristyle in a relatively secluded part of the house, recalls room B4/P23 of the House of Poseidon. In this room fragments of a marble statue, glass objects and ceramics, iron stool legs, scales, a jug, an amphora, a marble bust of Serapis, and bronze and terracotta oil lamps were found.
To the south of the peristyle were three rock-carved rooms. Two of them could not be searched since they were blocked due to the landslides. The third one, room C14, measured 4.00 m x 4.50 m and opened to the peristyle through a door. It had a niche on its west wall.
Archaeological evidence pointed out the mid-3rd c. AD for the destruction of the House of Euphrates. The function of several spaces seems to have changed just before the violent destruction by the Sasanian troops. The considerable number of finds, belonging to objects used for storage and domestic utilities revealed in rooms with mosaic pavements and/or in rooms with considerable dimensions and locations, like rooms C4/P16, C12/P25, and C13/P26, indicated an increasing need for storage perhaps because of the anticipation of the siege. After its destruction and a long phase of abandonment, the House of Euphrates witnessed a partial reoccupation. The new habitation was concentrated on the peristyle courtyard. A new wall was constructed on the north-west corner of the courtyard to determine the southern limit of a new space. A furnace found in the courtyard was also associated with a later inhabitation. Abadie-Reynal dated this later reoccupation to around 7th c. AD (2012b: 188).
170
5.2.7 The House without Mosaics
To the east of the House of Euphrates, a unit comprising at least ten rooms was excavated (fig.143). The westernmost rooms of this unit were adjoined the east wall of the House of the Euphrates in which no passage was detected. The floor level of the southern rooms of this unit (rooms D1-D2-D3) was at most 1.00 m higher whereas the floor level of the northern rooms (D4/P35-D5/P34-D6/P30) was at most 0.90 m lower than the floor level of the House of Euphrates (fig. 181). For these reasons, the French archaeologists who participated in the excavation of the area accepted this residential unit as the third house on the terrace and named it the House without Mosaics (Abadie-Reynal, 2012b). The Turkish archaeologists, nevertheless, supposed it to have been a part of the House of Euphrates and named the unit as Zone D (Önal, 2013: 131-138).
Rooms D1, D2, and D3 were similar in their dimensions and layout. Their floors were of beaten earth and the walls had no paintings. Room D1 was a carved room but the excavation of the rock to the south remained incomplete. Room D2 communicated with room D3 and with an eastern room that was not excavated. Rooms D4 and D5 were in the northern zone of the house that had a floor level at about 1.90 m lower than that of the southern one. Room D4/P35 opened to room D5/P34 and to an eastern space. In these two rooms many fragments of ceramics and pithoi, fragments of blue glass and a terracotta oil lamp were found.
Room D6/P30 differed from the other five rooms with its decoration. It had three layers of wall paintings of which the second layer was better preserved. In this layer, full-length female figures were represented on rectangular panels (fig. 182). In the destruction layer of the room fragments of pots, bricks, an oil lamp, and iron objects, and a complete jug were found.
Room D6/P30 opened to room D7a/P33 to the east. This small space functioned as a passage between room D6/P30 and the probable corridor D10/P43. At the same time, it functioned as an anteroom for room D7b/P31 to its south to which it gave way
171
through double bays. The floor of the room was paved with a mosaic of large tesserae. Room D7b/P31 was relatively small, measuring about 3.00 m x by 3.00 m. Its floor was about 0.10 m higher than the anteroom and was covered with stone tiles. A partial wall painting, representing a bird was preserved on the south wall. Rooms D6/P30, D7a/P33, and D7b/P31 came into prominence with their layout and decoration. Nevertheless, fragments of pithoi and ceramics found in these rooms, too, indicate a storage function, at least for the last phase of occupation.
D10/P43 was an elongated space that marked the eastern limit of the northern part of the house. To its east, several rooms were partly excavated. The floor levels of these rooms were about 0.80 m lower than the western ones and there was no passageway in between. The corridor led to room D8/P48, a small room with a beaten earth floor, and to D4/P35.
Under rooms D7a/P33 and D7b/P31, two vaulted rooms were exposed. Access to these rooms was provided from a door on the east wall of room D7b/P31. The benches and the fragments of amphorae found in these rooms led them to be identified as cellars.
A structure with three steps, measuring 1.00 m x 2.20 m was attached outside the north wall of room D6/P30. This was probably a staircase that led to an upper level. The considerable difference between the floor levels on the northern and southern parts of the building makes it likely that at least the northern part of the House without Mosaics had an upper story. Considering the modest size and layout of the house and the location of the stairs outside the building, Abadie-Reynal pointed out that there might have been a separate dwelling unit for rent on the upper story (2006a: 5).
The current state of excavation makes it difficult to comment on the general layout of the House without Mosaics and the means of communication between different levels. The lack of a courtyard, assuming that there was not any on the unexcavated south-eastern part of the house, is noteworthy since the open courtyards were the main
172
elements in the organization of the spaces in the neighboring houses. The modest size of the spaces and the lack of decoration point out a much less wealthy residence.
5.2.8 The House of Synaristosai (The House of Zosimos)
In summer 2000, French archaeologists from Nantes University excavated a large part of a domestic building, in Trench 6, at the easternmost limits of the ancient city. The building was named the House of Synaristosai after the theme of one of its floor mosaics representing a scene from Menander’s play with the same name. It was also called the House of Zosimos, after the signature of the mosaic artist inscribed on the same panel.
The House of Synaristosai, just like many other domestic buildings in Zeugma, was built on an artificial terrace that was partly carved and leveled from the natural bedrock and partly backfilled. The building rested against the bedrock on the southern limit of the terrace leaving a vast circulation space, of at least 20.00 m, to the north. To the east, the house was bordered by a public space, measuring about 20.00 m in the east-west direction. Further east was a probable public building that could not be excavated due to the lack of time before inundation. The architectural remains and the terracotta figurines found in the adjacent public space suggested a cultic function for that building (Abadie-Reynal, 2001a: 271; 2012: 106).
The House of Synaristosai had several phases of occupation that were marked by a number of rearrangements. The area was occupied by a necropolis at least in between the beginning of the 1st c. BC and the early 2nd c. AD (Abadie-Reynal, 2012a: 16; Abadie-Reynal and Darmon, 2003: 81). After the abandonment of the necropolis, a rural type habitation gradually invested the terrace by reutilizing some burial chambers. By the end of the 2nd c. AD, there were presumably two adjoining dwellings at the terrace that were incorporated, reorganized, and redecorated in the first half of the 3rd c. AD, to obtain a larger and more elaborate house, the so-called House of Synaristosai.
173
The plan and the functioning of the earlier houses on the terrace are difficult to grasp due to later modifications. It has been assumed that the eastern part of the site was occupied by a housing unit that comprised rooms P1, P2, P4, P8, P10 and E256 while the western rooms belonged to a second unit (fig. 183). Room P4, the largest space of the eastern house with the dimensions of 6.70 m x 7.80 m, has been assumed to be the courtyard (Abadie-Reynal, 2012a: 24). It seemed to have provided access to the house, following the scheme in many rural houses of northern Syria (fig. 184).64 The other rooms were aligned on the southern and western sides of the courtyard and all but P8-P10 opened to it. Room P8 was a rectangular room that was accentuated towards south by incorporating room P10 that was an earlier hypogeum tomb. This room was either accessed from room P9 or through an axial path from rooms P1 and P2. The size, layout, and relatively indirect access of this room may indicate its prominence among the other rooms. The door between rooms P8 and P9 was blocked, and an apse facing room P8 was designed on its place, in a later phase, probably in the late 2nd c. AD. More or less contemporarily a door was opened on the south wall of room P9 that gave way to older tomb E256 and a new means of communication was arranged between rooms P9 and P8. A staircase placed in the small room E256 presumably gave way to an upper level.
The western part of the terrace initially saw a loose occupation with a few carved rooms that were preceded with the large courts P12 and P15, on its southern limit. These courts were later covered and, with the carved rooms, were incorporated into a larger and more complex domestic structure built in the second half of the 2nd c. AD. The northern part of this structure was highly damaged due to the later constructions and the systematic erosion that affected the entire northern part of the terrace. At the north-east corner of the excavated area was probably an open courtyard that was partly invested by room P17 of the House of Synaristosai, as indicated by an old cistern in that room. The limits of this courtyard could not be specified but a slightly raised part of the leveled rock floor that made a right angled return, as well as the right angled
64 For rural residences in Syria see Tchalenko (1953); Tate (1992) and Sodini and Tate (1984).
174
canalization, to the south of the mosaic pavement of room P17, was associated with a stylobate and implied the existence of a possible peristyle in that courtyard (Abadie-Reynal, 2012a: 81-84). Room P3 might have been a covered room opening to this courtyard in the same phase of occupation. The westernmost rooms of that house were not excavated.
From the second half of the 2nd c. AD, the rather rural type occupation on the terrace at the eastern limits of the city, gradually gained a more urban character, obviously due to the urban growth. In the first half of the 3rd c. AD, the habitation on the terrace underwent major modifications and saw its most prosperous phase. The two existing residential units were united and enlarged to provide a unique house of at least 500 m2 (figs. 185-188). The new unit, the so-called House of Synaristosai, was redecorated with sumptuous mosaic pavements and wall paintings. Its circulation patterns were also reorganized by providing new openings or blocking the existing ones. As such not only the spatial quality of the rooms was enhanced but a particular hierarchy among them was (re)defined.
The plan of the House of Synaristosai seems to have been organized in mainly two parts, more or less following the distinction between the two units of the earlier phase. The eastern part, with a more elaborate layout and decoration comprised the public rooms, while the western part was rather reserved for service spaces, perhaps with the exception of room P17. The traces on the bedrock overhanging room P10, the staircase in room E256 and the fragments of painted plaster, mud bricks, and beams found in the destruction layer of room P9, and fragments of white mosaic found fallen in room P3 suggested an upper floor in the southern part of the house. The upper rooms might have been used for the daily activities of the family members.
Room P4 remained the spatial focus of the eastern wing of the House of Synaristosai. It played a key role in the circulation and functioning of this zone. Nevertheless, this earlier courtyard was probably transformed into a covered space as evidenced by the large number of tiles found in the destruction layer as well as the absence of vertical
175
pipes inside (figs. 189, 190). The probable opening of 1.40 m wide on the eastern wall of this space that was considered to be the entrance door of the earlier house, was blocked by the installation of the apsidal niche of the fountain in this phase of occupation. Since the northern wall of the room that was also the façade wall of the house, was totally torn down, it was not possible to know if there was direct access to that room from the public space to the north, but it seems not likely. Yet, a window grill found in the demolition layer on the north part of the room suggested a 1.30 m wide opening on that wall. Room P4 received a geometric mosaic and a fountain most probably during the same decorative program in the first half of the 3rd c. AD. The main panel of the mosaic pavement was placed offset towards the west with respect to the north-south axis of the room. This shift seems to have been related to the presence of the fountain and was roughly in alignment with the axis of room P9. A round marble base, pierced with a hole containing a stud and found in-situ on the mosaic in the center of the room was obviously carried a statue. The fountain with a rectangular basin was built on the eastern wall of the room against, but not axial with, rooms P1 and P2. It could also be visible from room P9 and from the street through the window on the northern wall. The basin measured 3.40 m x 1.30 m and its outer walls were covered with white-gray plaster. Fragments of marble slabs found nearby were probably used for the decoration of the fountain. Room P4 directly communicated with rooms P1, P2, and P9.
Room P1 was the first unit of a suite of four rooms that were aligned (P1, P2, P8-P10, P13-P11) to the west of room P4 (fig. 191). Located on the northern facade of the house, it is likely that this room was the vestibule with an entrance door on its north wall that was torn down up to the foundations. The room was paved with a geometric mosaic. Contemporary with or later than the installation of this mosaic, the existing passage between rooms P1 and P2 of the earlier phase of occupation was blocked (Abadie-Reynal, 2012a: 38). As such, all the visitors in the vestibule were initially led to room P4, the largest and sumptuous space of the house, and from there they were distributed to other rooms.
176
The carved room P9 was directly accessed from room P4 through a wide passage of 2.25 m (fig. 192). Its door was emphasized by decorative pilasters on both sides showing the importance of the space behind. The floor level of the room was slightly higher than that of P4 and was paved with a geometric mosaic. The walls were covered with painted plasters, two phases of which were identified. The second phase with marble imitations was dated shortly after the installation of the mosaic. Rich metallic objects found in the destruction layer testified the presence of elaborate furniture inside. The location, size, and decoration of this room suggested that it might have been a smaller reception space that served to accommodate the clients of the owner who could access this room from the probable family part located at the upper level and the reception suite to the west, through room E256.
At the far end of the circulation diagram of the eastern wing of the House of Synasristosai was room P13. Since the passage between rooms P1 and P2 was blocked in this phase of occupation, the guests in the supposed vestibule P1, had to enter first P4, then pass through P2 and P8 to reach the reception room P13. In this layout, the small room P2 that measured about 2.40 m x 2.65 m and was decorated with a geometric mosaic pavement seems to have served as an inner vestibule or intermediary room that linked the most frequented space of the house, room P4, and the relatively isolated reception space, room P13 (fig. 193).
Room P13 was the most elaborately decorated room with a figural mosaic pavement depicting women at breakfast, a scene from Menander’s play of Synaristosai (fig. 194). This rectangular central panel was surrounded by geometric motifs on three sides, forming a U-shaped arrangement that led the room to be identified as a triclinium. On the east wall of the room traces of wall paintings representing a figural scene with architectural details were identified. The importance of the triclinium was also testified by the decoration of its entrance door. The main access to the room was from P8 through a wide passage of 2.30 m which was decorated by moldings, reminiscent of the entrance door of room P9 (figs. 195, 196). On the eastern side of the room, there was a secondary opening, most probably a service door that opened to service room
177
P12 on the western part of the house. Room P13 was accentuated by an earlier burial chamber, room P11, to which opened an alcove with acrosolium E253 (fig. 197)65. It seemed that when this funerary chamber was integrated into the house its rock floor was recarved and lowered (Abadie-Reynal, 2012a: 60). Similarly, two identical niches, 0.69 m high and 0.40 m deep and arranged at 0.80 m above the floor, on the east and west walls of room P11 were dated to a later phase than that of the necropolis. Fragments of painted plasters with elaborate decoration were found in the demolition layer of this room. This space with an arcaded back wall and refined decoration, despite the lack of a mosaic pavement, recalls the reception rooms in the houses of Apamea on the Orontes.
Room P8, apparently an important reception space by itself in the earlier phase of occupation, seems to have functioned together with triclinium P13 in the House of Synaristosai. It preceded the triclinium and gave way to it through a monumental entrance (fig. 198). The floor of that room had a mosaic pavement representing different species of fish on a central panel that was bordered by geometric motifs. Traces of wall paintings with marble imitations were observed on the walls. The decoration of the apse on the northern part of the east wall, at the place of the blocked door between rooms P8 and P9, was also noteworthy. This 1.40 m wide and 0.30 m deep apse was painted with marble imitations in the form of narrow rectangular panels. According to Barbet, this decoration imitated a marble veneer used for fountains and hence the apse might have been the exedra part of a nymphaeum (2005: 209, 212). Room P8, with its mosaic pavement representing a marine theme and the apse imitating a nymphaeum, recalls the fountain courts to which the triclinia opened in the houses of Antioch. Room P10 was added to room P8 in the earlier phase of occupation and continued to function as an extension of that room. It communicated with P11 and provided lateral communication between rooms P8 and P13.
65 Acrosolium is an arched recess or sepulchral cell used for entombment in a Roman burial place or catacomb.
178
To the west of the triclinium was a large room (P12) measuring approximately 10.00 m by 6.25 m. It was probably a service space that communicated with the triclinium and the two rock-cut rooms on the southern limit of the terrace (fig. 199). A door of 1.15 m wide door on its south wall led to a carved room that was not totally exposed. The floor level of that room was about 1.00 m higher than that of P12 and it was accessed by a rock-carved staircase. Room P14 was probably a utility space and communicated with P12. Room P15, on the other hand, seems to have functioned as a service corridor that provided passage between the service spaces on the south-western part of the house and the northern rooms.
Room P3, to the north of room P13, was probably an open space as suggested by the existence of a basin (fig. 200). It measured about 8.50 m x 4.00 m and was limited by the façade of the house to the north. Since that part of the house was poorly preserved, it is not possible to specify the circulation patterns but it seems probable that this open court had provided the link between the eastern and western parts of the House of Synaristosai.
Room P17 was the only decorated room in the western wing (fig. 201). In the first half of the 3rd c. AD, this room replaced the presumed peristyle court of the earlier occupation phase. The area of this room was also enlarged by relocating the façade wall on this part of the house further north. The floor of room P17 was covered by a large mosaic pavement measuring 7.40 m x 9.75 m and comprising two figural panels. The larger panel on the west side, which measured 5.00 m x 3.70 m and highly damaged, depicted a sea snake and a fish biting a hook. The smaller panel measuring 4.50 m x 2.10 m represented a marine scene with amours, too. The eastern panel was not centrally aligned to the western one, and was placed around 1.00 m to its north, probably because, according to Abadie-Reynal, the earlier cistern on the south side of this space was still in use when the mosaic was installed (2012: 86). Based on the layout of the mosaic, that was suitably arranged to place the klinai on three sides of the smaller eastern panel, this room was identified as a triclinium. Since the walls on this part of the house were not preserved, the entrance door and other openings could
179
not be spotted. Yet, assuming that the klinai were placed to the east, the main entrance door should have been on the west wall. Given the proximity of service rooms and courtyard P3, Abadie-Reynal suggested that this triclinium was rather intended for the family, whose privacy could be preserved by a clear separation from the more official part of the house (2012: 85-86). Nevertheless, since it was almost the largest decorated space within the house and was located just behind the façade, just like the triclinium A1 of the House of Poseidon and C1 of the House of Euphrates, it can be suggested that this room was intended to impress the visitors, instead of being reserved for the family members.
Traces of destruction dating from the middle of the 3rd c. AD were present, especially in the western rooms. Unlike many other houses in the ancient city, this building was reoccupied not immediately but soon after its destruction and inhabited for a relatively long period. Initially, probably at the end of the 3rd c. or the beginning of the 4th c. AD, the terrace was reinvested with a concentration around the carved rooms to the south (Abadie-Reynal, 2012a: 16). Then, in the 4th c. or 5th c. AD a rural habitat gradually occupied a large part of the earlier house (fig. 202). During this later phase, only a few walls of the earlier house were used while new ones were constructed independent from the earlier layout. In that period rooms P10 and P11 were separated from rooms P8 and P13 by the addition of new walls on the southern side of the latter rooms. A new room, P16, was built on the north-eastern corner of room P12. In rooms P13 and P17 new walls were built over the mosaic pavements. At the end of the 6th c. AD the terrace seemed to be completely abandoned until a sporadic Islamic occupation took place between the 9th and 11th centuries AD (Abadie-Reynal, 2012a: 131-137).
5.2.9 The Late Imperial Peristyle House
During the rescue excavations, conducted by Oxford Archaeology, remains of an impressive late antique house were revealed in trench 7B, in the western part of the ancient city (fig. 203). According to the archaeological evidence, it was built in the second half of the 4th c. AD and remained in use until its destruction in the first half of
180
the 7th c. AD (Tobin, 2013: 100-102). This house, of which only a small part was excavated, is noteworthy since it is the only unearthed example of a domestic building that was originally built in the late antique period, rather than being adapted from an earlier house. Besides, unlike many contemporary houses, it seems not to have suffered from subdivision but was enhanced during its occupation.
The house was organized around a colonnaded courtyard (fig. 204). In its initial phase, the colonnade surrounded an area of 5.00 m x 5.00 m with columns on its western and southern sides and pillars on the eastern side. The original arrangement of the northern side was obliterated by later modifications. In the first half of the 6th c. AD the colonnade was extended towards north and west. At least two new columns were added to compensate for the new area.
The western portico of the courtyard was limited by the façade wall of the house beyond which a street ran along in the north-south direction. This portico also gave access to a staircase that led to an upper story above the rooms to the south of the courtyard that probably comprised the more private section of the house. The southern portico was aligned by three small rooms 7C, 7D, and 7E, the largest of which measured 3.40 m x 3.00 m. These three rooms were accessed from the southern portico of the courtyard and had windows on their southern walls probably opening to an open-air space. In the easternmost room, room 7E, a second doorway was located at the southern limit of the eastern wall, presumably leading to the outside (Tobin, 2013: 100). The modest size and the communication with a presumable open space to the south of these rooms, as well as some archaeological finds discovered in room 7D, suggested a utilitarian function. The northern part of the courtyard was only partially excavated. A doorway between two pillars in that part of the courtyard gave way to an unexcavated room. There was a stone shelf on the eastern side of that doorway. It seems probable that the reception room(s) of the house was located on this northern part, overlooking the river.
181
Everywhere in the Late Imperial Peristyle House, there was evidence of destruction by fire which led to the collapse of the building. Archaeological evidence suggested the mid-7th c. for the destruction, which could be related to the Arab invasions in that period (Tobin, 2013: 101).
5.3 Overview of Domestic Architecture in Zeugma
The sloped topography and the soft bedrock formation in the region of the ancient city of Zeugma challenged symmetry and order in the urban layout and architecture and gave rise to more practical solutions and adaptations. The residential structures of the city were built on artificial terraces that were partly carved from the bedrock and partly backfilled. The southern parts of the terraces were occupied by the buildings while the northern parts overlooking the valley were left empty for public circulation.
The size and layout of the buildings were restricted, to a certain extent, by the stretch of the artificial terraces. The houses of Roman Zeugma were relatively modest in size, compared to those of Apamea on the Orontes for instance, but they offered a certain degree of luxury with their distinct architectural features and decoration. Among the almost fully excavated houses in the ancient city, the House of Poseidon covered the largest surface area of about 1,000 m2 on the ground level, in its most grandiose phase. Nevertheless, this house had a two-partite arrangement which probably corresponded to two distinct earlier structures, each with a surface area of 500 m2 on average. Similarly, the House of Euphrates and the House of Synaristosai covered an area between 400 m2 and 500 m2 on the ground level. Remains of stairs evidenced that the houses of Zeugma had upper floors, at least in some parts of the buildings.
The earliest houses excavated in Zeugma were dated to the late Hellenistic or early Roman periods. They were organized around central open courtyards, sometimes with a complete peristyle or a single colonnade on one side forming a kind of loggia behind, recalling the pastas in Greek architecture. Carved rooms were also common in the domestic architecture of the ancient city from the early periods on. These spaces should
182
have been provided the desired cool environments in the hot climate of the city and they served diverse functions, like reception and service. In the earlier houses, the lack of decoration, other than wall paintings in some rooms, and the absence of running water were noteworthy. Until about the late 2nd c. AD, water for domestic use was mainly supplied from the cisterns that were carved on the rocky ground in the courtyards.
During the late 2nd and the first half of the 3rd centuries AD, seemingly in parallel to the economic prosperity in the city, as well as to the increase in population and urban growth, the houses of Zeugma underwent extensive modifications. They transformed in terms of room function and appearance thanks to the reorganization of circulation patterns, (re)definition of hierarchy between spaces, and the introduction of running water and decorative features. As such the differentiation between the public and the private parts of the domestic buildings became more obvious.66 In that period the domestic buildings of Zeugma reached their apogee in luxury and decoration.
The courtyard remained the main and the central element in the houses of Zeugma until the destruction of the city in the mid-3rd c. AD, but its function and impression changed progressively. The peristyle courtyards as well as some smaller inner courts received shallow basins with mosaic pavements, mostly with representations of marine themes, and in some cases with central dispositions like water jets and statues. Monumental and decorated fountains were usually added on one side of the shallow basins. As such the courtyards that used to provide passage and distribution to the surrounding rooms were transformed into spaces of pleasure and relaxation. Several surrounding rooms also took the advantage of the visual pleasure of the courtyard and the water features as well as the daylight and fresh air, as attested by the inner windows on their walls.
66 For studies on the ‘public’ and ‘private’ use of Roman houses see Wallace-Hadrill, 1988; Thébert, 1993; Grahame, 1997; Riggsby, 1997; Cooper, 2007; Tuori and Nissin, 2015. For late antique Asia Minor see Çonkır, 2005; Özgenel, 2007 and 2018; Uytterhoeven, 2022.
183
Two types of rooms came into prominence. They were differentiated by size, decoration, and/or location in the plan layout. The first was the triclinium that can be identified by the T or U-shaped arrangement of mosaic floor panels. The triclinia had the most sumptuous decoration in the houses, with their figural mosaic pavements, and wall paintings. Scenes from mythology or theatrical performances were chosen to be depicted in the mosaic panels. Such performances could well have been staged in the triclinia in Zeugma, to entertain guests after dinner, since they were popular in the Roman world including Antioch (Görkay, 2012a: 288-289; 2012b: xiii; 2011: 45-46). An inscription found on the wall plaster of a house in Dura-Europos supported this idea. The inscription listed the names of the artists and hetaera who traveled between Zeugma and Dura-Europos (Immerwahr, 1944: 203-265).
It is noteworthy that, none of the triclinia in the houses of Zeugma that have been studied here, had an axial placement with the peristyle courtyards or an axial opening towards them. The triclinium C1/P19 of the House of Euphrates, for instance, did not even have any direct communication with the peristyle courtyard. Nevertheless, the triclinia were often located on the façade and were preceded by inner courts with decorated water features, just as room C1/P19 of the House of Euphrates and room A1/P3 of the House of Poseidon, or by elaborate anterooms as room P13 in the House of Synaristosai. In the latter example, the anteroom P8 was not provided by any water feature but the marine theme was present in that space too; in the mosaic pavement and on the apsidal wall niche that was presumably intended to emulate the exedra of a nymphaeum.
The significance of triclinium in the domestic architecture of Zeugma in the late 2nd and the early 3rd centuries AD can be attested by the spatial transformation observed in the Houses of Poseidon and Synaristosai. In the House of Poseidon, the western portico of the peristyle courtyard A6/P9 was removed by the construction of a row of rooms, one of which was probably a reception room. Similarly, in the House of Synaristosai, the peristyle courtyard of the earlier phase of occupation was replaced with the large triclinium P17. This transformation was most probably a result of the
184
land limitations in the growing city but it also testifies the coming into prominence of the triclinium at the expense of the peristyle courtyard.
The second type of room can be distinguished with its location in the plan layout within the houses of Zeugma. The two most apparent examples of those were the carved rooms D1/19 in the House of Dionysus and Ariadne and P9 in the House of Synaristosai. Both rooms were preceded by an elaborately decorated larger space from which they were entered through relatively axially located and highlighted passageways. They were decorated with geometric floor mosaics and wall paintings. These rooms, with their distinct location and layout, seem to have provided a stage probably for the reception of more intimate guests and/or clients.
The introduction of running water brought about significant enhancements in the domestic context of Zeugma from the 2nd c. AD onwards. Besides its utilitarian function, water played an important role in the decoration of the houses. The placement of ornamental hydraulic structures in the key positions in the circulation pattern and/or in close association with the prominent rooms as well as the existence of installations imitating water features indicates that water had a strong social value, too, in the domestic architecture of Zeugma. Abadie-Reynal who compared the shallow basins representing marine themes with the fishponds linked to the triclinia in some excavated houses in Italy argued that such use of water and water features provided proof for the will of the local elite of Zeugma to imitate a western way of life (2010: 20; 2008: 109-115).
The florishment of the domestic context at Zeugma did not last long. Probably with the increasing threat of the Sasanians towards the mid-3rd c. AD, the renovations throughout the city came to an end and were followed by a period of decline. In this period, the houses underwent some changes which can be interpreted as adaptations rather than renovations. The archaeological evidence has reflected the anxiety and concern for a possible attack in the domestic lieu of the city. During that period, the houses were modified in terms of accessibility and functionality. Existing houses were
185
subdivided by blocking of the doors, some lavishly decorated rooms were given over to storage and food processing functions, latrines and fountains in some houses went out of use and some walls with painted plaster were covered by graffiti. Numerous locks and padlocks for doors and cabinets found in the deposits suggested an intense preoccupation with security, too (Dieudonné-Glad, 2006: 43-48). In addition to the subdivision of the houses, several military objects found in the destruction debris of several houses led some scholars to suggest that the houses were billeted by soldiers during the attack (Tobin, 2013: 110; Aylward, 2013c: 24).
In 252/253 AD, the army of the Sasanian King Shapur sacked Zeugma, burnt and destroyed a large part of the city. The archaeological evidence has suggested that the inhabitants of the city managed to escape before the sack, probably with the hope to return back to their houses, as several finds including hoards of coins, abandoned furnishings and household objects indicate (Aylward, 2013c: 30).
The excavations produced very little evidence for recovery just after the sack in the residential zones of the city. There is minor evidence for small-scale salvage in the late 3rd and 4th centuries AD. By the 5th c., however, new houses were built, some with rather high architectural standards like the Late Imperial Peristyle House in Trench 7B. At the same time, much smaller and simple houses were also built, sometimes with spolia, or the earlier ones were incorporated. The archaeological documentation suggests that there was a more pastoral life in Zeugma in late antiquity. Tandoor ovens built in a very simple style used for bread making were found in almost all the houses in this period. Spindle whorls and ivory needles used for spinning wool were usually found in contexts also dating to Late Antiquity (Parton, 2013: 295-344). The lavishly decorated rock-cut rooms of some 2nd and early 3rd c. AD houses were turned into stables in the last occupation phase of the city.
186
CHAPTER 6
PALMYRA: INHABITING THE DESERT
6.1 Urban Layout and Development of a Desert City
Unlike many other Greco-Roman cities in the East, Palmyra did not develop according to a predetermined Hippodamian plan, enclosed by fortifications. The city, however, adapted some other models of urban space organization in different districts (fig. 205).
The earliest defenses of the city date back to the 1st and 2nd centuries AD. These walls are considered to have been built as customs walls and barriers against the marauders from the desert rather than protecting the city from invasions (Juchniewicz, 2013: 193). They enclosed gardens as well as the residential areas but they did not form a complete ring and were built only where necessary (Gawlikowski, 1974: 231-242). For instance, no walls were built in the west of the city where the slopes of the hills formed a natural barrier.
The initial settlement of Palmyra is likely to have concentrated around the Temple of Bel from where it extended west along the southern bank of the Wadi al-Qubur (Gawlikowski, 1974: 232; Will, 1983: 77)67. This early Hellenistic district in the south of the wadi was occupied between the 3rd c. BC and 3rd c. AD (Schmidt-Colinet, Al-As’ad and Al-As’ad, 2013: 303). Archaeological records attest that, from the 2nd c. BC onwards, the area was dominated by smaller and bigger residential structures with
67 According to Zuchowska Wadi al-Qubur flowed in a different bed till the end of the 1st c. AD and the Temple of Bel was originally on the same bank as the Hellenistic settlement (2008: 229-234). She suggested that, in the late 1st or early 2nd c. AD, during the execution of the new urban plan, the course of the wadi was artificially moved to the south. As such the most important sanctuary of the city, the Temple of Bel, was integrated into the new monumental center that became concentrated in the north of the wadi.
187
courtyards organized irregularly around two principal streets forming a bifurcation (fig.206) (Will, 1983: 78; Zuchowska, 2011: 143).
In parallel to the increasing importance and prosperity of Palmyra in the late Hellenistic period, the urban area gradually extended towards the north of the wadi. Although a precise chronology for the initial phase of this extension is not yet set, it is assumed that the new urban quarter, the so-called south-western district of Palmyra, was developed during the 2nd c. AD when it was surrounded on three sides by colonnaded streets (Grassi, 2008: 3). This district, which is likely to be dominated by residential structures, showed a different layout than its neighborhoods (fig. 207) (Grassi and Al-As’ad, 2013: 117). It was organized around streets that were lying approximately in the north-south direction. The streets were neither in equal distance nor exactly parallel to each other but they were perpendicular to the wadi’s bed which is likely to have been the main axis of the communication of the city before the construction of the Grand Colonnade (Will, 1983: 79; Grassi, 2008: 1-11; Zuchowska, 2011: 143). Besides, the principal entrances of the theatre and the agora, as well as that of the Temple of Bel had been initially oriented south, towards the wadi. They were redesigned later with new entrances opening towards the city’s new artery to the north. The later colonnades surrounding this district on its three sides did not meet with the streets, but instead were just overlapped. Thus, it is obvious that the later development of the city had nothing to do with the urban layout of this earlier sector.
The Colonnaded Street of Palmyra consisted of three sections that were built in several stages in the course of the 2nd and 3rd c. AD (figs. 205, 208). The total length of the street was approximately 1.10 km and showed a non-linear alignment. The western section, with sidewalks and shops, was the oldest part and its construction had begun before 158 AD, as confirmed by some inscriptions (Baranski, 1995: 39). It started from the West Gate and was connected, at a right angle, to the Transverse Colonnade that led south to the Damascus Gate. The north-western district of the city seems to be contemporary with this section of the Colonnaded Street. The urban space in this sector showed a regular organization with streets almost perpendicular to the colonnade in
188
the north-south direction. Most of the area in this part of the city was occupied by residential units.
The eastern section of the Colonnaded Street started from the eastern limit of the Temple of Nabu and ended in the Temple of Bel. Its construction had started after 175 AD and continued until the beginning of the 3rd c. AD (Baranski, 1995: 43). The middle section, on the other hand, was built in the early 3rd c. AD and connected the western and eastern parts. This latter one also marked the monumental center of the city. Most important public buildings, including the theatre, agora, monumental arch, baths, and some temples were clustered around it. In order to create this center, some reorganizations were made in the already-built area. The temenos of the Temple of Nabu was reduced and some of the houses in the south-east area were restored to open space to the eastern section of the colonnade, the street that branched to the Grand Colonnade was linked with the theater and provided communication between the new center and the old city (Zuchowska, 2011: 192-193).
Zuchowska argued that there might have been a change in the master plan of urban planning in Palmyra before the western section of the colonnade was completed (2000: 192). According to her, this could be the reason for the non-linear arrangement of the colonnaded street, the difference in the column details between different parts, and the rearrangements necessitated in the southern district during the construction of the eastern and middle sections of the colonnade (2000: 187-193). Will and Grassi, on the other hand, suggested that the colonnaded street of Palmyra was implanted rather as an attempt to put an order in an irregular urban development than as part of a unique master plan (Will, 1983: 80; Grassi, 2008: 3).
The great building projects which gradually transformed the urban character and the appearance of Palmyra in the course of the Roman period were either initiated by imperial investment or more often euergetism undertaken by wealthy citizens. Many inscriptions testify that large scale urban projects including the construction of some parts of the colonnaded street or the rebuilding of the existing temples were financed
189
by the local elites who in turn had their statues placed on the columns of the city’s thoroughfares (fig. 209) (Zuchowska, 2000: 193; Ball, 2000: 76; Bryce, 2014: 278-279). They were mostly Palmyrene merchants, and embellished the city with elements of Greco-Roman architecture. The Temple of Bel, for instance, was rebuilt in the first half of the 2nd c. AD by a wealthy Palmyrene merchant called Male Agrippa who adorned the temple with imposing Corinthian columns (fig. 210) (Gates, 2011: 401-403; Bryce, 2014: 279). The Temple of Baalshamin, similarly, received a pronaos fronted by six Corinthian columns, representing an example of classic architecture (fig. 211). These buildings, nevertheless, did not always provide a setting for the Greco-Roman type of civic or religious rituals. Ball suggested that the building commonly identified as the senate house may have served as a place for the gathering of tribal elders or chiefs rather than for the meetings in the manner of the provincial Roman state (2000: 222). Similarly, the theatre itself, which was built in the late 2nd or early 3rd c. AD is believed not to have been used for theatrical performances in the traditional and Classical sense, despite its conventional Classical architectural features (Bryce, 2014: 283). Stoneman suggested that it was used less for dramatic performances than for public announcements, political meetings and speeches, and displays of oratory (1992: 65-66). Likewise, Bryce argued that even the Grand Colonnade might have served as an Oriental Bazaar with its numerous shops and booths, as well as the main thoroughfare and a processional way (2014: 282).
Monumental building stopped abruptly in Palmyra, after 273 AD when the city rebelled against Roman power and was destroyed by Aurelian. Structures under construction, such as the Theatre and the Grand Colonnade remained largely incomplete (Baranski, 1994: 9; Zuchowska, 2000: 191-192). Some of the existing buildings experienced disruptions in this period presumably by the Aurelian’s troops as archaeological evidence attests (Intagliata, 2018: 98).
From the 4th c. AD onwards Palmyra survived as a much smaller settlement of military importance. The city was involved in Diocletian’s substantial building program that aimed to strengthen the defenses of the eastern frontier. Probably under his reign or
190
slightly later Palmyra’s fortification walls were repaired and strengthened.68 The new circuit of walls enclosed only a smaller area, excluding the Hellenistic quarter in the southern bank of the wadi. The westernmost area of the city, beyond the Transversal Colonnade was converted into a military camp where a garrison was quartered. The building activities of this period also included the construction of Baths of Diocletian and some minor restoration works on the Grand Colonnade (Intagliata, 2018:99).
The advent of Christianity was also a significant phenomenon that reshaped the urban landscape of Palmyra in late antiquity. Although Christianity is attested in the written sources from the first half of the 4th c. AD the archaeological evidence comes mainly from the 5th and 6th centuries AD when several churches with their dependencies and chapels were built, mostly in the north-western quarter of the city (Intagliata, 2018: 30-31).
Palmyra continued to prosper to a certain extent throughout the period of Umayyad dominion. A suq was installed in the westernmost part of the Grand Colonnade and a mosque was built nearby. Evidence of continuity of occupation has been found in most of the residential districts of the city (Gawlikowski, 2008: 89-90; Intagliata, 2018: 106). By the end of the 10th c., the settlement presumably had shrunk into the Sanctuary of Bel which itself was converted into a mosque while its territory became a fortress (Gawlikowski, 2008: 91). Another fortress on top of Qal’at b. Man, overlooking the city from the north was constructed possibly in the 13th c. AD (Intagliata, 2018: 107). This small Islamic settlement survived as such until the early 1930s when the inhabitants were transferred to a new place by the French authorities.
68 There is not a consensus on the exact chronology of the ramparts of Palmyra. Seyrig supposed that queen Zenobia built the walls hastily against the assault of Aurelian, and Justinian modernized them (1950: 229-252). Gawlikowski and Kowalski associated them with the reign of Diocletian (Gawlikowski, 1974: 231-242; Kowalski, 1997: 44). Lately, Juchniewicz attributed them to the reign of Aurelian and some later reinforcements to the reign of Tetrarchs (2013: 193-202).
191
6.2 Domestic Architecture in Palmyra
Palmyra was rediscovered as early as the end of the 17th c. AD by English merchants traveling to Aleppo. In the 18th and 19th centuries, the ruins of the ancient city became a focus of interest and were visited by many European antiquarians, artists, photographers, and travelers. Some of them published accounts of the ancient city presenting some notes and drawings of the standing monuments.
It was in the 1860s that scientific studies were intensified at Palmyra. Many scholars came to the site for archaeological explorations, however, these were limited to few opportunistic clearances and were not conducted as real excavations. During these works, many tombstones and inscriptions came to light. The first real excavation at Palmyra was conducted by the German-Austrian archaeologist Otto Puchstein in 1902. Several non-sequential excavations followed in 1912, 1917, 1924, and 1925 by different researchers. With the establishment of the French Mandate and the creation of the General Directorate of Antiquities for Syria and Lebanon in the 1920s the scale of archaeological researches changed. Hasty explorations, often with no follow-up, were replaced by more systematic and comprehensive excavation and restoration campaigns, which were directed by Henri Seyrig, and continued until World War II. Between 1929 and 1932 the modern town of Palmyra was moved from the ruins to its present site. With the end of the mandate and the independence of Syria in 1946, the archaeological work resumed and expanded. Since then, different missions from different countries have been working at Palmyra, often in association with the General Directories of Antiquities and the Museums of Syria. The archaeological work at Palmyra, just like at Apamea on the Orontes, had to stop in 2011 because of the civil war in the region. In 2015 ISIS gained control of the area, destroyed the ancient city, and murdered several inhabitants of the modern city, including the archaeologist Khaled al-Assad, the head of the Antiquities of the ancient city of Palmyra.
Remains of several residential structures were noticed among the ruins of Palmyra in the early years of the archaeological explorations. However, only a few of them had
192
been reported and none of them had been fully studied. Only in the latter decades of the 20th c. some domestic structures were excavated and studied in more detail. Although the sample unearthed and published so far is limited, the study of the current data gives an overall understanding of the characteristics of the domestic context in Roman Palmyra.
The urban spaces of Palmyra, except the monumental center and the Temenos of Bel, are considered to have had a mainly residential character. Remains of ancient houses were detected in various quarters of the city that have different chronology and urban layout (fig.212). Some of these houses were built on the building plots determined by the parallel streets as was the case in the western quarter of the city, while some others were constructed in the free spaces between the public buildings just like the ones in the monumental district (Zuchowska, 2011).
Due to the lack of stratigraphic data for the houses, especially for the ones mentioned in the earlier studies, it is not always possible to analyze the present sample within a chronological sequence. However, the construction of most of the houses published so far is dated back to the 2nd or early 3rd c. AD in reference to their structural techniques or decoration. In the following part of this chapter, the Roman domestic structures of Palmyra will be studied according to their location within the urban layout that helps to follow an approximate chronological order too.
6.2.1 Domestic Structures in the Hellenistic Quarter
So far, no house dating to the Hellenistic period of Palmyra is totally unearthed and studied in detail. However, stratigraphic surveys in the so-called Hellenistic quarter in the south of the Wadi al-Qubur have revealed that the urban space around the two main streets in this area was filled with houses that did not follow any specific urban grid (fig. 206) (As‘ad and Schmidt-Colinet 2000). There were houses of various size that were organized around a courtyard. The most striking feature of these early houses is
193
the lack of colonnades in their courtyards. They were also not very regular in shape and their inner space organizations were also irregular (Zuchowska, 2011: 142-143).
6.2.2 Roman Residential Buildings in the Western Urban Quarter
6.2.2.1 Houses 38, 39 and 45
The earliest traces belonging to the domestic architecture of Roman Palmyra were concentrated on the houses visible on the surface of the ground in the western quarter of the city. In 1925, Albert Gabriel detected and surveyed twelve houses that spread over the ruins and published the plans of three of them -those referred to with numbers 38, 39, and 45 in his site plan (fig. 213) (Gabriel, 1926: 84-87). Two of these houses, referred to as 39 and 40 in Gabriel’s plan, had been noticed also by Daniel Krencker in the German expedition of 1902 and published in 1932 (fig. 214) (Krencker, 1932: pl. 19).
The early published plans were rather in the form of sketches or drafts instead of final drawings since the buildings were neither excavated nor scientifically studied. In comparison to the other houses excavated in the city, these plans look rather idealistic than realistic in terms of the regularity and symmetry of their layout. The lack of any stratigraphic work and chronological data makes it impossible to relate the plans with the occupational phases of the dwellings. However, they give a general idea about the architectural layout, distribution of spaces, and circulation patterns. These were smaller residences, constructed not earlier than the beginning of the 2nd c. AD, since this urban quarter is contemporary with the Grand Colonnade. They have several common features to which Gabriel called attention. They were organized around rectangular or square peristyle courts of varying dimensions. The peristyle of House 45, whose major colonnade remained standing, shows the characteristics of a Rhodian type with one side higher than the other three. In most cases, on one side of the peristyle was a room larger than the others, which was sometimes flanked by secondary rooms. In House 39, this large room is preceded by a double portico which Gabriel interpreted
194
as a Mesopotamian or Persian import (1926: 87). Both houses had a dis-axial entrance that provided a certain level of privacy and invisibility for the core of the houses whereas House 38 had more direct access. These residences were likely to share their north and south walls with their neighbors, as shown in Krencker’s plan. Instead of being individual residential structures, these smaller units may well have been part of a bigger structure that had to spread longitudinally since they were bordered by parallel streets on their two sides. This thesis is supported by the existence of a similar longitudinal house, House F, in the same quarter of the city.
6.2.2.2 House F
To the south-east of Houses 39 and 40, a large residential complex, named House F after the designation of the insula on Gabriel’s plan, was excavated by the Polish mission between 1988 and 1995. It was located between two parallel streets running northward from the Grand Colonnade: the so-called Diogenes Street, after an inscription on a column and the so-called Church Street leading to three Christian basilicas. The streets were connected with a narrow road at a distance of approximately 135 m from the Colonnade which marks the northern limit of the house. The area to the south of the house up to the Colonnade was left empty (Gawlikowski, 1996: 139).
The construction of the house was dated to the 2nd half of the 2nd c. AD thanks to the sherds revealed in the foundation layers and the style of the capitals and stucco work (Gawlikowski, 2007: 87; 1991a: 399, 403). The house was continuously inhabited until the late 8th or early 9th c. AD.
This was a large two-story structure, measuring about 26.00 m wide – the entire width of the insula-and 79.00 m long, covering an area of more than 2,000 m2 (figs. 215, 216). It was supported by a stone pedestal about 1.50 m above the floor level and had mud-brick walls above this pedestal. Entrance to the house was provided via three doors opening to rooms 3, 32, and 38 from Diogenes Street and via four doors opening to rooms 16, 17, 25, and 44 from Church Street. The inner space was organized around
195
several courtyards and was composed of mainly two parts having separate entrances. The northern section of the house was isolated from the southern one and is likely to have been the more private and secluded section, reserved for the family or for domestic activities (Gawlikowski, 2007: 87-89; 1997a: 161). This section was accessible via a narrow door from the eastern street and entered through a cramped recess under a flight of steps (room 16) from where the interior courtyard was not visible. The unpaved courtyard (room 13) measured 10.00 m by 14.00 m and had porticoes with three columns on two sides, forming a kind of pastas. Each portico gave access to separate suits of two or three rooms, one of which is larger than the others. In the southern end of the western portico a kind of iwan, marked by an arch was formed, commanding the rooms 5 and 12. Between two sets of rooms behind the western portico, an entrance from Diogenes Street (room 3) and a staircase, which was later removed, were set.
The northern section of the house had no connection with the southern one on the ground level and the communication between them is likely to have been provided from the upper levels. Gawlikowsi suggested a mezzanine level at the mid-height of the front columns, between the stairs in the rooms 16 and 25 which provided the passage (2011: 88; 1997a: 164). He argued that the two staircases could have only one flight since their width is about 1.20 m and they cannot reach all the way up to the level of the terrace of the peristyle. He added that the height of the orders, which was about 5.80 m in the colonnade, did not seem to be appropriate for the smaller rooms located between these two staircases. Thus, he suggested that the communication between the two separate parts of the houses could have been provided from a mid-level above the rooms 17, 18, 23, and 24 and was led by staircases 16 and 25.
The southern section of the house was composed of a number of units, each being organized around a separate courtyard. Colonnaded court 22 and the surrounding rooms seem to have had a more public character and were open to the visitors. This section was accessed via a door adjacent to the one in the northern quarter with which it shared the same monolithic threshold block. Through this entrance, one passed two
196
antechambers (rooms 17 and 18) and reached courtyard 22 that was paved with large slabs and adorned with a monumental colonnade on two sides. Next to the entrance spaces to the east of the courtyard were located rooms 23 and 24, probably with a service function, and a secondary entrance under a flight of stairs (room 25). At the other end of this courtyard was a suite of three rooms (rooms 19, 20, and 21). Room 20 was the largest one of the unit and had the most elaborate stucco decoration in the whole house. It was provided with niches and three doors leading to the neighboring spaces. To the north, room 20 had a connection with the adjacent room 19 and its distyle porch 21 that opened directly to the peristyle court. To the west, it had access to the service core of the house composed of rooms 27-32. This core had a separate entrance from Diogenes Street (room 32) and included a possible kitchen (room 27), a lavatory (room 31), and another room of uncertain purpose, probably a washroom (room 30). Court 29 was provided with a well and drainage but the latrine was not equipped with running water.
Further south was a second paved courtyard, court 35, with columns only on two sides. This courtyard could have been entered through a narrow passage under a staircase (room 39a). It had also access to the service entrance through room 34, to the adjacent peristyle courtyard 39, and entrance 38 through room 36. It was surrounded by several interconnected rooms (rooms 33, 34, 36, 37, 40 and 41) on mainly two sides. Considering the entrances and circulation patterns the section around courtyard 35 was likely to have served as the more secluded unit in the house while the adjacent section around courtyard 22 had a more public function. These two units did not have direct communication but they both had access to the service core.
To the further south of the house, two smaller courtyards (courts 39 and 46) with columns and one without (court 45) were located. Courtyard 39 had an independent entrance from Diogenes Street (room 38) and gave way to a relatively large room 42 and to another peristyle, court 46, through the passage room 47. There is a small latrine in the corner of room 44, too. Since this southernmost part of the house was
197
substantially altered in the Byzantine and Islamic periods, its original layout could not be followed (Gawlikowski, 1996: 143).
Staircases in different parts of the house indicate the existence of upper levels in different sections of the house except for the southernmost part. According to Gawlikowski, all the rooms around the two northern courtyards were likely repeated at the upper level with terraces above the porticoes (1997a: 163). However, the differences in the height of orders between the two peristyle courts and even between the two orders in the same courtyard had to necessitate level differences at the upper floor. In the northern peristyle, the order of the northern portico was 4.25 m while the western one was about 4.75 m, and hence there must have been a few steps on the upper level to provide the connection between the terraces above the porticoes (Gawlikowski, 1997a: 163). Similarly, the height of the orders in the southern colonnade is about 5.80 m which is at least 1.00 m higher than the northern part. Thus, it is difficult to assume a pre-planned and continuous second floor that stretched the entire structure. Yet, it is also possible that the stairs had led to a terrace or a flat roof instead of an upper story.
House F is a remarkable building with its relatively large size and elaborate layout with several living units organized around different courtyards with colonnades. It had a lavish decoration with decorated cornices and friezes and painted pilasters. Other decorative elements might have been deprived due to a very long period of occupation. What is noteworthy is that, unlike many other Roman period houses, it lacked spacious reception rooms and did not exhibit any axiality or symmetry. The lack of running water and water features within the house is also striking. There are five wells in various parts of the house but their depths are above the water table. Thus, the water supply was probably provided by transporting it from outside (Gawlikowski, 2007: 90; 1994: 139). The two latrines had outlets into the main conduit under the neighboring streets.
198
It is likely that this house belonged to a relatively well-to-do citizen of Palmyra. The discovery of a set of tesserae issued in the name of a priest called Yehiba in a test trench in 1996, may indicate that he was the owner or a prominent resident of the house sometime in the 3rd c. AD (Gawlikowski, 1997b: 197). The architecture of the house has some Classical features like columns and plaster cornices. On the other hand, its strict architectural separation into two, perhaps so as to serve both as a family residence and a public place with their own entrances, and its architectural layout that created suits of three to four rooms to accommodate a large family may reflect a more Eastern tradition.
House F survived in its more or less initial form until the late 6th c. AD when an earthquake devastated the structure. After that disaster, both its architecture and the function of many spaces were gradually transformed. In the later phase of occupation, the upper story fell out of use and the stairs were blocked (fig. 217). The house was divided into several parts each of which functioned as a separate dwelling sharing common entrances and courtyards which suggests that the owners likely were relatives (Gawlikowski, 1991b: 86-87; 1993: 563). An Arabic signature engraved on the cemented floor of a room may indicate that the latest inhabitants were Arabs (Gawlikowski, 1991b: 8; 2008: 90, 95).
Within the separate dwelling units of the later phase, the larger spaces were also further divided into smaller rooms, some doors were blocked and new entrances were provided. Most blocked doors were transformed into storage cupboards and new cupboard niches were carved into the walls. The colonnades of the courtyards and the porches were blocked with partitions, some to the mid-height of the columns to provide small inner courts to receive light to the rooms. On the other hand, the cornices and the friezes of the 2nd c. AD decoration remained in-situ in most parts of the house until its abandonment. In the later phases of occupation, some spaces of the house changed function. The original entrance (room 16) to the northern courtyard was blocked on its inner end and was turned into a shop. A new room (room 15) was added next to it with a new entrance which was later blocked too and the room was turned into a stable as
199
suggested by a manger near the blocked street door. The former entrance to the adjacent courtyard (room 17) was blocked and the room became dependent on the northern part. To the south, a large courtyard was formed adjacent to the service core by abolishing the walls between rooms 32, 33, and 34. An oil press was installed in room 37 and probably also in room 42. In the southernmost part of the house several stone benches, tannours, and storage jars sunk in the ground were detected (Gawlikowski, 1996: 143). With these subsequent transformations, the elaborate residence of an affluent family turned into a cluster of dwelling units that accommodated several families, perhaps the relatives, in a more rural setting until its abandonment in the late 8th or early 9th c. AD.
6.2.2.3 The Peristyle House
One of the two houses that Gabriel detected to the south of the Grand Colonnade in the western part of the city, and referred to as number 61 in his plan, has been partly excavated by the Italian-Syrian archaeological mission (Pal.M.A.I.S) between 2008-2010. This was the largest structure visible on the ground and was located very close to the late rampart (figs. 218, 219).
The construction of this dwelling was dated to the end of the 2nd c. AD or the beginning of the 3rd c. AD and had several occupational phases until the 8th c. AD (Grassi, 2011: 190; Palmieri, 2010: 183). It was aligned between two north-south oriented alleys but its orientation was different from the orientation of all other structures revealed in the south-west quarter of the city (Grassi and Al-As’ad, 2013: 117). Thus, this large peristyle house seems to have been lately installed into the existing urban fabric of the district that formed the earliest extension of the city towards the north of the wadi.
So far, only a part of the peristyle and some northern and western rooms of the house were excavated which cover a total area of 560 m2. The excavations reached the level of the oldest phase of the building only in room A, so there is little information about the original state of the building. It had an almost square peristyle, measuring
200
approximately 12.00 m x 12.00 m, with six columns on each side. To the north of the peristyle the relatively large rooms I and H were situated which probably constituted the monumentally designed reception part of the dwelling. The northern portico of the peristyle fronting this monumental section was wider than the western one. On the western side, there were two smaller rooms, A and B, communicating with each other and also with the peristyle courtyard. The backfill of the latter floor in room A yielded rich pilaster fragments and various colored marbles of different origins including Greece, Asia Minor and Africa which were likely to have been part of the wall and floor decoration of the house in its first state of occupation (Nava, 2015: 241-252; Palmieri, 2010: 175-186). The archaeological data, although rather limited, indicates an extensive and luxurious house planned as a large peristyle residence with rich and elaborate decoration.
In the later phases of its long occupation, this house, similar to many other dwellings in Palmyra, witnessed some architectural and functional changes, until its abandonment around the 8th c. AD. The floor levels of the building were raised around 0.90 to 1.00 m, probably to cover a collapse or destruction layer (Grassi and Al-As’ad, 2013: 119). The porticoes of the peristyle were blocked to form new rooms, larger spaces to the north were subdivided and the doors between the interconnected rooms in the west were blocked. During the latest phase of occupancy, a considerable amount of the architectural elements and construction materials, mostly from the same building, were reused. Among these reused materials, an entablature or console with a Greek inscription and two small altars with Palmyrene inscriptions are notable (fig. 220). There is not sufficient data to determine the exact function of the altered spaces and the structure itself but in reference to some arrangements in rooms A and B and a number of coins found in room A Grassi pointed out the possibility of productive or commercial use in addition to the residential one, in the later stages of the building (2011: 195).
201
6.2.3 Residential Buildings in the Eastern Urban Quarter
6.2.3.1. The House(s) of Achilles and Cassiopeia
Remains of a large and elaborate residential complex to the east of Temple of Bel were partially brought to light in 1939-1941 by the Directorate of the Department of Antiquities of Syria and with the initiation of Henri Seyrig. Unfortunately, all the documents of this excavation which were kept in Damascus disappeared before being published during World War II. The in-situ remains were later restudied and the plan of the edifice was published by Edmond Frezouls in 1976 (fig. 221) (1976: 29-52). More recently, a French archaeological mission reexamined the remains without doing further excavation (fig. 222) (Delplace, 2013: 37-48).
Frezouls suggested that this complex might have occupied the southern edge of a residential quarter which was bordered by perpendicular streets that defined a vast insula (1976: 35). If hypothetically completed, the whole structure would have fit into a rectangle of at least 65.00 m x 45.00 m and hence would result in a total surface area of more than 3,000 m2. The excavated part of the complex covers an area of approximately 1,300 m2. Initially, the exposed remains were interpreted as parts of two distinct residential structures which were named the House of Achilles and the House of Cassiopeia, after the iconography depicted in the floor mosaics (Frezouls, 1976: 35-44). The later studies, on the other hand, suggested that the remains belonged to a single large structure and hence the house was renamed as the House of Achilles and Cassiopeia (Zuchowska, 2011: 148; Delplace, 2013: 37-48). Based on its construction technique, and a comparative study of the decorative elements, the house was dated to the end of the 2nd c. or the first third of the 3rd c. AD, while the iconographical analysis dated its mosaics to the mid-3rd c. AD (Delplace, 2013: 38).69
69 Traces of an earlier mosaic were also identified under the level of the Achilles mosaic (Frezouls, 1976: 50).
202
The house was organized around several courtyards and had two relatively distinct parts which had caused the complex to be identified initially as two separate dwellings. The southern section having a surface area of about 1,000 m2 is better preserved and more elaborate than the poorly preserved northern section, with about 300 m2 unearthed surface area. The highly demolished eastern limit of the house does not allow to restore this part of the plan while the western side largely remained unexcavated except the entrance spaces in the south-western quarter.
The house was entered through a triple door in its south-west corner providing access for both the pedestrians and the carriages. The entrance led to a large open vestibule, room 2, which had a direct connection with some service premises. Room 3 next to the entrance door might have been a porter’s lodge. Room 4 measured roughly 3.00 m x 5.00 m and had an axially positioned and later collapsed wall in the direction of its length which recalls a staircase. This room gave way to other spaces to the north which were ruined. The proximity to the entrance makes it likely that some service quarters including food storages and kitchens, and even a dining room might have been placed to the north of the entrance unit which constitutes the western border of peristyle courtyard 5 (Frezouls, 1976: 38).
The entrance door and the vestibule did not have a direct visual connection with the core of the house. From the vestibule one could reach the eastern portico of the peristyle courtyard, passing through a corridor between rooms 3 and 4. The central peristyle courtyard with six columns on each side and measuring about 22.00 m x 19.00 m constituted the largest space of the house around which the other rooms were organized. The numerous stucco fragments found in the court indicate the opulent decoration once adorned this space. The northern portico of this peristyle was larger than the other three and gave access to the most monumental and elaborate section of the house exposed so far. To the north of the peristyle and centered with respect to its axis, was located the spacious room 16. This room was distinct in terms of its vast size and prominent location and was most probably the major room that served as a reception hall. Next to it was a second smaller peristyle court with three columns on
203
each side and whose portico was paved with the mosaic of Achilles (fig. 223). This small court communicated directly with the large courtyard to the south, with the large reception room 16 to the east, and probably with another room to the east which is totally ruined now. It was also connected, via room 15, to a separate suite of rooms to the east of the main peristyle courtyard.
To the east of the large peristyle court was a square unit consisting of small rooms and an interior courtyard with a central basin. This assembly was a closed unit in itself and according to Frezouls it might have been destined for the living quarters of the family or women, despite the several openings to the peristyle court (1976: 40-41). He argued that rooms 9, 10 and 11 which took advantage of the fresh air and light coming from the courtyard could have been bedrooms, while the slightly larger and the richly decorated room 13, communicating with all the surrounding spaces, might have been a common hall of this wing.
In the south-east corner of the house was a service court, court 8, with an irregular pavement that probably gave way to the utilitarian quarters in the east which are not preserved. Next to this court is room 7 that had a latrine and a possible staircase leading to the terrace above. The function of large space 6, on the other hand, could not be determined but its proximity to the entrance and service spaces and the limestone pavement it had could indicate its function as a service space too.
The northernmost section of the house, which might have belonged to another dwelling, the so called the House of Cassiopeia, was separated from the southern part with a common wall. This wing was organized around a large courtyard too, but it lacked a peristyle.70 Among the six rooms that were partly preserved in the north and east of this courtyard, room 22 is distinct in its size and location. Similar to room 16, it was located on the axis of the courtyard and had larger dimensions. It opened to the peristyle with a wide central door and to two of its neighboring rooms, 21 and 23, with
70 There might have been a portico in the western part of the courtyard from which nothing survives.
204
smaller symmetrical side doors. Room 21 had the same width with 22 but was less deep than the former. It revealed remains of a raised platform in its eastern part and its floor was paved with the mosaic of Cassiopeia. With their location, dimensions and decoration, adjacent rooms 21, 22 and 23, and the unpreserved room with which the latter one communicated to the east had a prominent look and are likely to have been the reception suite of the northernmost part of the house. Room 24, located symmetrically with room 20, opened to the courtyard by a 3.00 m wide door and it might have been a part of this suite.
Room 21 opened to a small space, room 20, that was separated from the courtyard by two columns, one of which is preserved. Since the back wall of this room is demolished, its layout and function are difficult to understand. It may have been a porch dependent to the courtyard or if its back wall had a door, it may have been a passage room leading to the southern part of the house through corridor 18, or to a vestibule providing entrance from the east of the house. Next to room 20 was a long room or perhaps a corridor (room 19) that was poorly preserved.
Different than the above-mentioned examples of Palmyrene houses, House(s) of Achilles and Cassiopeia demonstrates a more articulated layout and an opulent decoration, and was much bigger in size. Availability of larger building plots and the suitability of the urban layout in this sector of the city may have given a chance for more free and spacious planning. In addition, it should also be kept in mind that this house was built at least a century later than the others.
6.2.4 Residential Buildings within the Monumental Centre of the City
6.2.4.1 The House to the South-east of the Theatre
Excavations in the theatre district revealed that several residential structures were scattered in the small area between the Theatre and the Temple of Nabu. One of these houses was partially excavated in the early 1960s by the Service of Antiquities of Syria
205
during the conservation works in the portico around the theater. The plan and the description of the house were published by Frezouls in 1976 and a corrected version of this plan was republished by Gawlikowski in 2007 (fig. 224) (Frezouls, 1976: 44-51; Gawlikowski, 2007: 84).
This was a large peristyle house with a total excavated area of about 1,000 m2. Its eastern, western, and southern limits were either unexcavated or not preserved. It was adjoined the curvilinear portico encircling the cavea of the theater which indicates that the house was built later than the theatre, which itself had not been built earlier than the late 2nd or early 3rd c. AD (fig. 225) (Frezouls, 1976: 50).
The north façade of this house had a curvilinear layout following the line of the portico. This facade extended parallel to the central part of the colonnade at a 4.40 m distance. The ten columns facing the unearthed part of the house were distributed at equal intervals except for the ones in the middle that corresponded to the entrance. The span between these two central columns was wider than the others, accentuating the entrance of the house. Beyond the entrance was a large central vestibule, room XI, with two probable service spaces on both sides, rooms IV and X, each having direct access to the peristyle courtyard. There was also a shop, room VII, communicating with the vestibule, in the same façade. Unlike other domestic structures excavated at Palmyra, this house had an entrance axially located with respect to the street and the courtyard. Hence, a visitor in the vestibule or in front of the door, or even a passerby in the street could have a view of the core of the house when the doors were open.
The vestibule led directly to the large central peristyle courtyard that measured about 18.00 m x 15.50 m with five columns on each side. Around this courtyard were arranged groups of rooms with different layouts. The larger width of the eastern portico indicates the prominence of the eastern section of the house. It is striking that the two corners rooms, XIX and V, were cropped at one corner, to the advantage of the courtyard which makes it likely that this portico was enlarged sometime after the initial construction of the house.
206
Large room XIII was centered on the axis of the peristyle court, behind the larger eastern portico, and opened to it with a wide central entrance. It was served by two doors opening to corridors on both sides. This room also communicated with a relatively secluded and smaller courtyard of four columns designed as a tetrastyle space (Room VI). Since the easternmost part of the house remains unexcavated, it is not possible to picture out the total layout of this section. The dimensions of Room XIII and its communication with the central courtyard, service corridors, and the smaller courtyard, nonetheless, make it likely to have been a dining room or a reception room. There may well have been some service spaces unexcavated in this part of the house and room I or another one further east may have served as a service entrance.
To the south of the large peristyle courtyard a second, similarly spacious room was situated (Room XX). Unlike room XIII, however, this room was placed in a non-axial location with respect to the courtyard and had two smaller doors placed close to the corners. This off-centered door arrangement provided relative visual privacy to the interior; when they were open they exposed only the side areas of the large hall to the peristyle. In almost all the residential structures examined in this study, prominent rooms -that were likely to have been the reception rooms- were situated to flank the courtyards on their short sides. Thus, their entrances, acting as a threshold between the courtyard and the hall that extended beyond, framed a spatial depth and visually oriented the visitors towards the other end of the room. Romm XX in House to the South-East of the Theatre, on the other hand, flanked the courtyard on its longer side and the doors close to its corners may have been planned as such to create a similar experience of capturing the spaciousness of the hall in an indirect way. In a very similar arrangement with room 16 in the House(s) of Achilles and Cassiopeia, room XX opened to a smaller tetrastyle courtyard on one side. On its other side, it was connected to Room XIX, and perhaps to one more space further east.
Three smaller interconnected rooms XV, XVI, and XVII were located to the west of the large peristyle courtyard. All these rooms directly communicated with the large peristyle courtyard and also with the tetrastyle courtyard XXI, via the southernmost
207
room XVII. Though planned in a different layout, the location and intercommunication of these rooms, perhaps together with room XXII, recall the probable living units of the House of Achille and Cassiopeia as well.
The unit of three rooms on the north-east of the excavated part of the house is also striking. Small rooms II and III that had direct access from the street led to larger room V. This room had the remains of two columns probably forming an arch that emphasized and, in a sense, separated one part of the room from the other. Room V is also one of the few rooms that did not directly communicate with the central courtyard. Nevertheless, it communicated with the more secluded courtyard VI with a side door. Although it is not possible to define the exact function of this unit, due to lack of archaeological investigation or finds, room V recalls the audience chambers of the late Roman period houses that were located more often in the peripheral zones to preserve the privacy of the rest of the house. As such, it is a strong possibility that Rooms II and III functioned as a secondary vestibule and/or the waiting rooms for the clients and visitors while Room V was the place where the owner of the house accepted them. The location of the house as attached to the portico of the theater, and in the monumental core of the city is a strong indication of a prominent owner engaged with public issues. This suggestion indirectly supports the idea that he organized an audience hall in his house to manage the public interaction.
With its magnitude and architectural design that integrated a number of colonnaded courtyards, one being a spacious peristyle, and the type of physical relationships between its spaces, House to the South-East of the Theatre resembles more House(s) of Achilles and Cassiopeia than the other houses excavated so far in Palmyra. However, it is also distinguished in certain aspects, such as its axially planned street entrance and the presence of a spatial unit possibly functioning as an audience hall and its dependencies. Unfortunately, there is yet not any evidence or mention concerning staircases, water features or decoration aspects neither in the plan nor in Frezouls’ report. Nonetheless, for such an elaborate residence it will not be wrong to assume that the house had mosaic floors and decorated walls that were most likely disappeared. In
208
the unpreserved parts of the house, in addition, there might have been one or more latrines, kitchens, and staircases leading to the upper level.
6.2.4.2 The Houses to the East of the Temple of Nabu
Excavations done in 1964-1965 in Palmyra brought to light partial houses situated between the Temple of Nabu and the eastern section of the Colonnaded Street leading to the Temple of Bel. So far, only Saliby published a plan showing a small section of this residential area, accompanied with very brief information (fig. 226) (1996: 289-290).
This part of the city was most probably inhabited since the Hellenistic period and underwent several rearrangements throughout the centuries. Some parts were expropriated for new urban developments, particularly for the construction of the Colonnaded Street, throughout the late 2nd and early 3rd centuries AD, as mentioned above. Totally fourteen rooms were opened in this sector, belonging to at least two different dwellings separated by an alley that provided access to this sector from the colonnade. Some of these rooms touched the exterior wall of the Temple of Nabu to the west, while some others were adjacent to the shops behind the colonnade.
The current state of excavation does not allow gathering detailed information except for some general features. Several rooms were organized around courtyards but no colonnades were spotted. The floors of many rooms were paved by cut limestone. There were many installations that served for production. The largest room of the architectural assembly to the west (room 3), measuring 7.20 m x 5.00 m contained a reused column shaft and a trough from which flows a channel cut into the pavement that ends with a circular section sewer. According to Saliby, existence of such an assembly may indicate that this room was used for dying wool or silk (1996: 290). In the adjacent room 7, there existed two basalt blocks forming a part of a mill that could have served to grind spices. Similarly, there was a column shaft in room 1 and a circular trough and a well in room 14. The presence of such installations suggested
209
that the owner of the house could have been a grocer treating the products that were imported to Palmyra (Saliby, 1996: 290).
6.3 Overview of Domestic Architecture in Palmyra
The current sample of the excavated houses at Palmyra is a limited one, comparing it with, for example, the sample from Apamea on the Orontes. The published examples are not many and the knowledge about them is restricted. There is no firm chronological data and archaeological finds are not recorded. The absence of detailed excavation reports and drawings other than plans in most cases, furthermore, prevents making a comprehensive study. Yet, the available documentation provides an insight into the distinctive aspects of domestic architecture in Roman and Late Antique Palmyra.
First of all, Palmyra did not develop according to a regular urban plan, unlike many Greco-Roman cities planned with a Hippodamian plan in the eastern empire. Consequently, the domestic structures were not planned to fit into a traditional insula system that is an urban layout pre-determined by virtual lines and grids. Instead, they adjusted to the available space in an organic way. In the north-western and south-western quarters of the city where the urban space was organized along more or less parallel streets built at irregular distances, the houses exhibited a linear extension between these streets, as in the case of House F. While the width of the structures was determined by the distance between the streets, the length was determined more freely, according to the architectural necessities. Within the monumental center of the city where the public buildings dominated the urban space, on the other hand, the house plans were adapted to the available free space between these structures and hence took some design references from them, just like House to the South-east of the Theatre which had a curvilinear façade following the line of the portico. In the districts where space was probably more available, the domestic structures extended more freely in width and length as was the case in the House(s) of Achilles and Cassiopeia.
210
The basic structural unit of the Palmyrene house was a central courtyard surrounded by rooms. Bigger structures were designed by the multiplication of this generic unit and received more courtyards serving the varying needs of crowded households or extended families. These courtyards received colonnades probably in the 2nd c. AD, whereas their Hellenistic predecessors spotted in the south of the wadi were designed without columns.71 The colonnaded courtyards were planned either as peristyle courtyards with columns on all four sides or with colonnades in only one or two sides, reminiscent of the pastas or prostas arrangements in Greek houses.
Central courtyards also constituted the main medium of circulation within the house and the clusters of rooms were arranged around them. Functions of different sections or rooms required different spatial treatments such as being open, visible and accessible or else more secluded and thus invisible and indirectly accessible. This duality was managed by using secondary rooms, smaller courtyards, passages and interconnected rooms. The houses themselves were also provided with a certain degree of privacy from the street by means of indirect entrance patterns through the vestibules, except the House to the South-east of the Theater whose vestibule opened axially and directly to the street.
The use of Greco-Roman features is more apparent in the architectural layout and decorative scheme of the relatively later houses of Palmyra. The House(s) of Achilles and Cassiopeia and the House to the South-east of the Theater, both of which date to the late 2nd or the early 3rd c AD, exhibit a more elaborate layout and decoration with larger peristyle courtyards, reception rooms of considerable size, axial locations and mosaic floors displaying Greco-Roman iconography (fig. 82). On the other hand, House F, which dates to the second half of the 2nd c AD and is likely to have belonged to a relatively well-to-do and crowded household, displays a different architectural design approach. The apparently strict separation between the units reserved for the
71 The earliest unearthed domestic structures in the north of the wadi dates to the 2nd c. AD, and thus it is not possible to know if the 1st c. AD structures had colonnaded courtyards.
211
more private areas and for the reception suits, the relatively modest layout of the smaller and irregular rooms, and the lack of right angles in many walls seem to have responded to the local practices of daily life and the associated architectural traditions. The use of columns in courtyards and the use of friezes and painted pilasters as architectural decorations on the walls indicate the adaption of Classical elements, probably as a reflection of the current architectural trends that manifested in the urban monumental architecture.
Despite a certain degree of splendor and luxury in the residential structures in Palmyra, the lack of running water and related amenities is notable. Except for a few cisterns which were likely to have been filled with transported water and small latrines, the residential structures of Palmyra were not provided with running water neither for utilitarian needs nor for decorative purposes.
The changing political and economic dynamics in Palmyra after the events of 273 AD had a certain impact not only on the social sphere and the urban context but on also the private domain in the city. After this date and coinciding with the cease of major public building activities, no new elite residences were built while the existing ones continued to be inhabited. Among the excavated residential structures, House F and the Peristyle House provided chronological data according to which they were occupied for about six centuries with several changes and transformations.
The urban residences in Palmyra showed a different transformation process than their contemporaries in other cities, in the course of late antiquity. The common tendency of the Roman aristocratic elite towards the elaboration of their domestic settings and their will to display power and wealth through their residences, especially in the 4th and 5th centuries AD, cannot be observed in Palmyra. None of the structures excavated so far were redesigned in that manner, nor were adorned with large apsidal halls, lavish floor mosaics, marble revetments or, sumptuous water installations that were common in many late antique houses in both the eastern and western provinces.
212
Palmyra, nevertheless, underwent a more radical transformation towards the ruralization of both the urban and the domestic contexts, and this happened earlier than its contemporaries in the eastern provinces. Encroachment of the existing public monuments with poor residential buildings and workshops started at least as early as the end of the 4th or the beginning of the 5th c. AD as the evidence from the Temple of Baalshamin, the Temple of Nabu, and the annexes of the agora suggest (Intagliata, 2018: 37-41).
The elaborate urban residences of the well-to-do families, on the other hand, gradually turned into smaller, poor-quality, and multi-functional units with rural settings. Larger houses, as in the case of House F, were split into smaller apartments to accommodate more families. Upper floors were abandoned, small shops were added to the facades, porticos of the courtyards were blocked to make new rooms or inner courts, larger rooms were subdivided into smaller units and some spaces turned into stables for sheltering animals and into workshops for production and storage activities as evidenced by the installation of olive-presses, mills, ovens, storage jars, and cupboards. Gawlikowski and Intagliata related this phenomenon of transformation, though in different chronological frames, to the replacement of the landed aristocracy of Palmyra with a middle class of merchants, artisans, and shopkeepers. They, however, propose different chronologies. According to Gawlikowski this transformation process was triggered by an earthquake in the 6th c. and continued in the 7th c. AD (1993: 136; 2008: 90). Intagliata, on the other hand, underlined the possibility of a much earlier change in the social composition of the Palmyrene community, considering the encroachment on public buildings occurring as early as the 4th c. AD (2018: 45).
213
CHAPTER 7
DURA-EUROPOS: FROM SHOPS TO PALACES
7.1 Urban Layout and Development of a Fortress City
Michael Rostovtzeff, after several years of excavations in the ancient city of Dura-Europos, identified the early Seleucid settlement as “…not a mere stronghold or a military village, but a combination of strong fortress and regularly-planned, and well-constructed city” (1941: 483). According to him the urban plan of the city with fortification walls was implemented shortly after its foundation (1938: 11; 1937: 197). The recent excavations and the review of the earlier work by the Franco-Syrian mission, nevertheless, put forth that there were two distinct phases of settlement at Dura-Europos (Leriche and al-Mahmoud, 1994; Leriche, 1996, 1997; Kosmin, 2011). According to current archaeological, stratigraphic, and epigraphic data, Dura-Europos remained as a small military occupation for about a hundred and fifty years after which it witnessed an urban expansion and a substantial reorganization in the 2nd c. BC.
The initial Seleucid settlement of Dura-Europos was situated on a hilltop, at the south-eastern extremity of a rocky plateau directly overlooking the Euphrates (fig. 227). This promontory hill, commonly identified as the citadel by the excavators, was separated from the plateau by a deep ravine and was relatively isolated from the rest of the city. The top of the hill was fortified by stone walls and occupied by a palatial structure. The dwellings of the military officers were clustered in the periphery of the citadel (fig. 228) (Kosmin, 2011: 99).
From the 2nd c. BC on, in line with the population growth, the settlement of Dura-Europos gradually extended towards west, through the distinctive area between the river to the east and two parallel deep ravines lying at a distance of about 1.00 km to
214
the north and south. In its largest extent, in the Roman period, the city covered an area of about 75 ha (fig. 229). It was around the middle of the 2nd c. BC that a large-scale urban planning and construction program was initiated in Dura-Europos. This program comprised the implementation of the Hippodamian plan and the construction of city walls and large-scale public buildings.
According to Pierre Leriche, the fortification walls of Dura-Europos were initially conceived as a prestigious work but in the course of the construction they took rather a defensive character (Leriche, 1993: 126; 1996: 165; 1997: 191-210; Leriche and Al Mahmoud, 1994: 400). That is the reason, he argued, why many gates were blocked and the construction technique changed during their execution. The particularly neat cut stone masonry used in the defensive walls gave way to mud brick construction on a stone plinth in some sections of the wall. Besaac and Leriche explained this change by the need to accelerate the construction of the fortifications due to the incoming Parthian threat (Besaac: 1988: 297-313; Leriche; 1996: 165-166; 2003: 176-177).72 The fortification walls of Dura-Europos were strengthened by 26 towers especially on the desert front to the west. The main gate of the city, called the Palmyra Gate, was placed on the west wall and there was a secondary gate on the south wall.
The Hippodamian plan of Dura-Europos was contemporary with the fortification walls and was established with reference to the main street that started from the Palmyra Gate and ran towards the river in the north-west south-east direction (fig. 230). The 12.00 m wide main street was approximately twice the width of the regular streets. The city blocks or insulae formed by the intersection of the perpendicular streets within the city walls measured approximately 35.00 m x 70.00 m. In the center of the city, just to the north of the main street was the agora which occupied a large area covering eight
72 Von Gerkan and Rostovtzeff, who studied the fortifications of Dura-Europos during the excavations of Yale University, explained the change of construction technique of the walls in a totally different way. According to them the defensive walls of the city were initially constructed with mud brick on a stone plinth in the Seleucid period, and they were replaced by cut stone in the Parthian period to strengthen them against a possible Roman attack (von Gerkan, 1939: 4-61; Rostovtzeff, 1938: 11).
215
city blocks. The width of the streets surrounding the agora was about 8.00 m wide (Leriche, 2003: 181).
Besides the citadel palace and the agora, only a few excavated buildings were dated to the Hellenistic period of Dura-Europos. These included a second palatial structure that was called the Redoubt or the Strategeion, found at the edge of the inner depression to the south-east of the city, the Temple of Zeus Megistos, just to the south of the Strategeion and the Temple of Artemis located in the southern part of the city. It is also noteworthy that no remains of a theater, one of the essential components of a Greek city, nor a stadium or gymnasium were found in the ancient city of Dura-Europos (Kosmin, 2011: 102).
The grand urban program of Dura-Europos remained unfinished when the city was conquered by Parthians circa 113 BC. After that date, since the settlement lost its significance as a military fort, the fortifications were abandoned. In some parts of the defensive walls where the construction work was not completed, the unused material was left on the ground. The accumulation of debris in some places along the rampart reached more than two meters in height and prohibited access to the towers (Leriche, 1993: 126; 1996: 166).
It was especially after the middle of the 1st c. BC that the city of Dura-Europos flourished and expanded to fill the entire area within the walls. During the Parthian period, the Hellenistic Hippodamian plan was highly respected in the development of the city, despite an increasing irregularity of the blocks towards the periphery. This phenomenon may attest to the continuity of the Greek administrative and municipal structures well after the conquest (Leriche, 2003: 182).
Being laid out on the Hellenistic grid layout, the urban appearance of Parthian Dura-Europos was gradually transformed into that of an oriental city, with the use of regional forms and traditions in public and private architecture. No new administrative or secular public building, except a bath in the north-east sector of the city, was
216
constructed during the Parthian reign but the existing buildings were maintained and continued to function with some transformations.
The large open square of the Hellenistic agora was covered by new buildings with commercial and/or domestic functions. The existing commercial blocks were also expanded to the detriment of the public space. As such the Hellenistic agora with a large open square lined up with regularly planned shops and offices was gradually transformed into a rather oriental type of marketplace with narrow streets lined with small shops and workshops (figs. 231, 232).
The Parthian era of Dura-Europos was dominated by the construction of many temples dedicated to Parthian, Syrian, and Palmyrene Gods. Besides the newly erected religious buildings, the existing temples were also renovated in that period. Two towers, 1 and 14, of the Hellenistic rampart, that were located farthest away from the main gate were integrated into religious buildings. According to Downey, the temple plans of this period were mostly based on Babylonian forms (1985: 116-129; 1988b: 88-129).
With the Roman conquest circa 165 AD, Dura-Europos received a permanent military garrison and regained its military importance. Consequently, with the increased concern for security, the fortifications came into prominence once more. The walls were restored and a mud-brick glacis was built to support them. A series of stairs was added against the walls to provide access to the rampart walkway and to the top of the towers which became inaccessible due to the accumulation of debris in the Parthian period. Similarly, the main gate was repaired and a forecourt was added to protect it (Leriche, 1993: 126).
After 210 AD, when the garrison was strengthened by new detachments, fundamental changes appeared in the urban layout of Dura-Europos. To provide the needs of the enlarged garrison and the increased number of soldiers, the northern part of the city was separated, at least in part, from the rest of the city by a mud-brick wall and
217
transformed into a regular military camp. Some of the pre-existing buildings in that area were razed to the ground while some others were transformed and received new functions.
Roman influence was well reflected in the building types and their architecture within the military camp (fig. 233). In the center of the area, over the remains of earlier structures, was built a monumental praetorium. In front of it ran a colonnaded street with a triumphal arch. Two new temples dedicated to the gods of the Roman army were erected while some of the existing temples in the same area went out of use. Two monumental baths were built near the praetorium and the earlier Parthian bath was reconstructed for the use of the soldiers. A modest amphitheater was also built within the camp area to serve only the garrison. Several graffiti indicated that the amphitheater was frequently used for gladiatorial shows (Rostovtzeff, 1938: 25). On the north-eastern edge of the city close to the military camp, but not within it, a large residential complex, the so-called the Palace of Dux Ripae or the Roman Palace was constructed. Several houses within the military camp were converted into barracks for the soldiers. Except for the Roman Palace and one of the bath buildings, the orientation of the new constructions of the Roman period was fitted into the Hellenistic urban grid.
The Roman military presence was not confined to the camp and the traces for the occupation of soldiers were found throughout the city. Inscriptions and graffiti provided ample evidence for the presence of Roman soldiers especially in and around the Palmyra gate (Pollard, 1996: 212-215). A number of houses in that part of the city were taken over for the billeting of soldiers. Roman presence, nevertheless, did not bring major changes in the urban layout and appearance of the southern ‘civilian’ section of Dura-Europos. That part of the city was provided with very few new buildings. These included a new large bath near the citadel and two new cult centers, a synagogue, and a Christian church, both of which were converted from the existing houses. No new temples or large residences were built in that period, but the existing ones were restored or transformed.
218
In the later years of the Roman reign, some parts of the agora and the main street were endowed with colonnades. According to Pierre Leriche, they were part of an urban project which was started later than those of the other cities in the eastern empire and could not be completed due to the Sasanian incursion (2004: 157-158). Susan Downey, conversely, argued that there was no attempt to turn Dura-Europos into a ‘Roman’ city (2000: 170). In any case, Dura-Europos never looked like other Roman cities in the region, such as Apamea, Antioch or Palmyra, all of which were adorned with colonnaded streets and monumental public and private buildings. This difference could be related to several reasons such as the relatively late incorporation of the city to the Roman sphere and the lack of time for transformation before the Sasanian invasion, the social and political structure of Dura-Europos where the private enterprises overtook the public ones and the apparent lack of euergetism in the city (Butcher, 2003: 261; Baird, 2014: 62).
7.2 Domestic Architecture in Dura-Europos
The site that houses the remains of the ancient city of Dura-Europos was visited by some scholars and travelers in the 19th and the early 20th centuries but its significance was not recognized until 1920 when a detachment of British soldiers camping in the ruins incidentally revealed the wall paintings of an ancient sanctuary. Professor James Henry Breasted, who was the director of the Oriental Institute of the University of Chicago and was in Baghdad at the time, made a one-day visit to the site and inspected the remains. The impressive paintings he recorded, and published after his brief survey gave rise to an immediate interest in the ancient city among the archaeologists and the public (Breasted, 1922; 1924).
In 1922-1923, Franz Cumont, on behalf of the French Académie des Inscriptions and Belles-Lettres, undertook two brief archaeological campaigns in Dura-Europos. 73
73 With the San Remo agreement in 1920, the borders of Syria and Iraq were established and the former was put under the protection of the French while the latter was under the British. That is why the early archaeological work in Syria was mostly undertaken by French teams.
219
Franz Cumont, probably because of his interest in ancient religions, extensively excavated the religious buildings and gave little attention to the residential structures. He published only a single house plan with brief information (1926, 241-249, plate 289).
The excavations restarted in 1928 under the joint auspices of the French Academy and the Yale University, with Michael Rostovtzeff acting as the scientific director. Throughout this expedition of ten seasons, more than a quarter of the surface area of the ancient city was uncovered. A large number of houses were excavated during this period, mainly with the hope to find such artifacts as paintings, inscriptions, parchments, and papyri, rather than with an interest in their architecture or domestic context (Baird, 2014: 14-15). As referred by Baird, a letter written by Henry Rowell, one of the archaeologists working at the site, to Rostovtzeff reveals the motifs for excavating the houses: “In sha Allah, the houses will yield us some good pottery and perhaps papyri.” (2014, 15). The results of the ten campaigns of the Yale University-French Academy joint expedition were published in the form of preliminary and final reports but the houses had never been the main topic of study for the excavators, and thus no full-length study or a final excavation report on domestic architecture was ever published. Some of the excavated houses were only treated in varying focus and details in the preliminary reports. Today a large collection of the documents including the field records, notebooks, object registers, plans, reports, and photographs belonging to this expedition, most of which remained unpublished, are preserved in the archives of Yale University Art Gallery. The archaeological finds from the site were shared between the excavating institutions and Damascus. Some of them are currently being exhibited in different collections and museums in the USA, France, Canada, Syria, and Lebanon.
In 1982, the Mission Franco-Syrienne d’Europos-Doura (MFSED) was established and the excavations recommenced in Dura-Europos under the directorate of Pierre Leriche. The mission stated its aim as to reexamine the existing archaeological data, to save the monuments from destruction, to make available the documentation from
220
the previous excavations and to make new surveys (Leriche and MacKenzie, 1996: 589). This expedition continued annually until 2011 when the civil war in Syria caused all the archaeological work in the region to a halt. It was only during and after this latest mission that the houses of Dura-Europos attracted more scholarly attention and several publications have been produced mainly by the reexamination, in some cases reexcavation of the remains that were previously revealed on the site, and the restudy of the unpublished excavation documents of the earlier expedition that are kept in the Yale University archive (Allara, 1986; 1988; Allara and Saliou, 1997; Baird, 2014).
The residential buildings comprise the largest part of the excavated area of Dura-Europos. So far, more than a hundred houses have been unearthed that provide a tremendously valuable sample for the study of domestic architecture, not only in the ancient city but also in the eastern territory of the Roman Empire. This sample is notable because of the unique condition of the preservation of the site. Since the city was abandoned and was not reoccupied after the Sasanian incursion of the mid-3rd century AD, the buildings in their latest phase of occupation were not damaged by later constructions and largely remained intact.74 Furthermore, the houses revealed in Dura-Europos were not only those of the elites, as was the case in many other excavated Roman cities in the region, but they present a variety of types and sizes, ranging from modest residences to palaces.
The extent of the excavations and the content of the publications of the earlier expeditions in Dura-Europos, present a very restrictive context for the study of domestic architecture. First of all, the excavated houses and the artifacts were not systematically recorded and documented. As mentioned above some of the excavated houses have never been published and some were only briefly mentioned with schematic or sketchy plans in the preliminary and final reports. Architectural plans were frequently prepared at a much later date than the excavations that took place in
74 The rich and well-preserved sample of houses excavated in Dura-Europos led the early scholars call the city “the Pompeii of the Syrian Desert” (Rostovtzeff, 1938: 2-3).
221
the related buildings. Similarly, numerous finds were not always recorded and/or published in context. The artifacts of everyday life were mostly ignored, while finds like paintings, inscriptions and papyri were given importance. The nomenclature of the excavated houses in Dura-Europos also differed in the publications. In some studies, the naming system established by the Yale University-French Academy team in 1931 was followed. In this system, the city was divided into sectors, each containing about eight insulae (fig. 230). These sectors were designated alphabetically. Within the sector, each insula was given a number and identified with a two-character code such as C7 or G1. Within this scheme, a letter was assigned to each house and a number to each room. The courtyard of house B in block C7, for instance, was identified as C7B1. Besides this alpha-numerical system, some houses were named according to one of their features or the name of estimated owners, or the name of its excavator, like the “The House of the Large Atrium”, “The House of Lysias” or “Cumont’s House”. In some other cases, the same house was referred to by different names in different publications.
The lack of stratigraphic studies in Dura-Europos also makes it difficult to identify the chronology of the houses in most cases. The virgin soil was reached only in a few areas during the excavations. Unless epigraphic evidence existed, the excavators dated the structures according to the construction materials and techniques or with reference to the ceramics found in the cisterns of courtyards. Thus, as Allara stated, the dating of the structures given in the excavation reports is mostly hypothetical (1986: 43; 1988). Two excavated structures with a partial residential function that can certainly be dated to the Hellenistic period of Dura-Europos are the palatial buildings-the so-called the Citadel Palace and the Redoubt Palace or the Strategeion. Some remains below the later houses were dated to the Hellenistic Period, too, but their state of preservation did not give much idea on their architectural layout. Most of the residential buildings excavated so far in the city were most probably built or rebuilt in the Parthian period and continued to be used, with some changes and transformations, throughout the Roman period. The only residential building securely dated to the Roman period is another palatial building, the so-called the Roman Palace or the Palace of Dux Ripae.
222
In the following part of this chapter, some of the relatively well preserved and published houses unearthed so far in the city will be examined in more detail, in a chronological order. The sample to be examined is chosen among many houses with the aim to illustrate the chronological and architectural diversity, and to present the major urban features (fig. 234).
7.2.1 The Citadel Palace
The fortified top of the citadel hill on the eastern edge of Dura-Europos was occupied by a building with two successive phases. The building was excavated during the 2nd season of the Yale University-French Academy expedition that took place between October 1928 and April 1929. It was identified as the probable residence of the Seleucid governor of the region and was called the Citadel Palace (Pillet, 1931: 13-15). The uncovered remains of the building were only briefly mentioned in the excavation reports (Pillet, 1931:13-15; Hopkins, 1931: 53-55). In the 1980s, the remains in-situ and the excavation documents kept in the Yale archive were reexamined by Susan Downey who published the results of her studies in several articles (Downey, 1985; 1986; 1988a; 1992).
The initial phase of the Citadel Palace was dated to the 3rd c. BC. The remains of this phase were too fragmentary to reconstitute its plan but they revealed that the exterior walls of the initial building were built in parallel to the enclosure wall of the citadel. In a second phase, probably in the 2nd c. BC, the palace was rebuilt with a totally different plan and orientation (Perkins, 1973: 15; Downey, 1986: 34).75 Since a large portion of the cliff overlooking the Euphrates fell into the valley and took away the northern part of the building, only the remains of the south-western part of the second phase of the citadel palace has been preserved (fig. 235).76
75 Rostovtzeff claimed that the second phase of the Citadel Palace was dated to the Parthian period (1938: 10).
76 According to Perkins the cliff fell down in the 1st c. BC whereas for Downey it was probably the result of the earthquake that took place in 160 AD (Perkins, 1973: 16; Downey, 1986: 27).
223
The unearthed remains revealed that the Citadel Palace was divided into two parts -the northern and the southern parts- that were separated by a corridor of about 2.12 m wide (fig. 236). The southern part of the building was organized around a peristyle court with Doric columns on at least two sides and with a cistern at the center. The courtyard probably had a square form, with sides of 20.00 m. On the south side of the courtyard was a row of interconnected rooms. The one at the center was the largest, measuring about 9.50 m x 7.50 m. This room was axially located with respect to the courtyard and opened to it via a wide opening with three columns. A row of five pillars at the center of the room probably supported the roof. This large central room, most likely a reception room, was flanked by at least two smaller rooms of different sizes on either side. The side rooms did not open to the courtyard. Remains of a rectangular feature, which could have been the support of a table, according to Downey, was revealed in one of the side rooms to the east (1992: 151).
On the western side of the courtyard were the remains of five interconnected rooms that were arranged in two rows. Two of the rooms, and probably a third one too, on the eastern row opened to the courtyard. The two larger and adjacent rooms had doors on four sides while the others had doors on at least their two sides.
Remains of a 2.12 m wide corridor running along east-west direction was found to the north of the eastern wing. Beyond this corridor was the corner of a large space and an adjacent north-south running corridor with the pilasters engaged inside the walls. The width of the latter one that formed the western façade of this part of the building was about 1.05 m, except between the pilasters where it narrowed to 0.75 m. There were three preserved entrances to this corridor, one was 0.90 m wide and located on its southern wall while the other two were on the eastern wall and 1.12 m wide; the latter opened to a large adjacent space to the east. This large space was also entered from the east-west corridor, through at least two doors of 1.02 m wide. The absence of interior supports in the preserved section of the space suggested that this was an open court but the wall thicknesses in the northern part of the palace that were wider than those in the southern part challenged this idea.
224
Frank Brown, the field director of the excavations in Dura-Europos in 1936-1937, took some notes on the Citadel Palace and drew a hypothetically restored plan of the building that are all kept in the Yale University Archives (fig. 237). In his schematic plan, Brown supposed a symmetrical arrangement for the building. For the disappeared eastern wing of the southern part, he suggested two rows of rooms, similar to the ones found in the western wing. He also added two more rooms to the southern wing. Considering the wall thickness (1.40 m wide), Brown suggested a row of three iwans that functioned as reception halls for the northern part. In between the halls, he placed narrow corridors and smaller rooms. Brown’s proposal of the reception halls in the form of iwans was found convincing by Rostovtzeff and Perkins but not by Downey and Nielsen (Rostovtzeff, 1938: 36; Perkins, 1973: 14-15; Downey, 1985: 116; Nielsen, 1994: 121). Since iwan has not been identified as an architectural form before the Parthian period, Downey found it unlikely that it was used by the Greek colonists in Dura-Europos and suggested instead the existence of a large open court in the northern section of the palace (1986: 32; 1985: 116).
To conclude, the Citadel Palace, one of the earliest examples of the unearthed buildings with at least a partly residential function in Dura-Europos, was clearly planned in two sections that were divided by a long corridor. The southern section with small rooms located on at least two and most probably three sides of a peristyle courtyard with indirect access was probably the residential part of the building with a certain amount of privacy. The layout of this part exhibits some Greek architectural features. The northern section which had large halls or an open court(s) was probably the official and ceremonial part of the building. The narrow corridor with pilasters in this part could have been a local architectural practice since it was not common in Greek architecture. Downey resembled this elongated corridor with inner pilasters to those in the Babylonian temples (1985: 116). She also found similarities between the Citadel Palace and the residential buildings excavated at the Hellenistic city of Aï Khanum which were also designed with two sections; the residential part on one side and a large open courtyard on the other (fig. 238) (1986: 32).
225
The citadel of Dura-Europos had been abandoned sometime before the desertion of the city, probably just after the collapse of the northern part of the cliff. The area turned into a burial site, probably at the beginning or before the Roman occupation as the archaeological evidence suggested (Pillet, 1931: 56). According to Perkins, when the citadel was abandoned, the functions that the palace served should have been transferred to Redoubt Palace built across the ravine (1973: 16).
7.2.2 The Redoubt Palace/The Strategeion
The south-eastern edge of the plateau on which the city of Dura-Europos extended had a peculiar geologic formation. This area was surrounded by steep rocky slopes on three sites and protruded like a bastion above the ravine. It had a dominating position over the river and the ancient road passing through the valley, as well as the citadel across the ravine. This rocky spur was identified as a redoubt by Cumont, and considered to have been the acropolis of the Hellenistic city by Rostovtzeff (Cumont, 1926: 19-20; Rostovtzeff, 1938: 35, 47). On the northern edge of the summit of the spur, raised a spacious building on a sumptuous retaining wall that was supposed to be the official residence of the chief civic magistrate of the city, the strategos (Rostovtzeff, 1938: 35; Perkins, 1978: 13; Nielsen, 1994: 116). Hence, the building was named the Strategeion and also the Redoubt Palace.
The Redoubt Palace was largely excavated in 1931, during the fourth season of the Yale University-French Academy excavations and briefly described with an accompanying plan in the excavation reports (fig. 239) (Pillet, 1933: 21-27, pl. III). In 1938 Rostovtzeff published the partly hypothetically reconstituted isometric drawings of the building that were drawn by Pearson (fig. 240) (1938: plate V). The remains were reexamined by MFSED after 1985 and provided some new information on the chronology and the development of the building. In 1986 a plan of the building that was slightly different from that of the Pillet, was published in the comprehensive volume Haus und Stadt im Klassichen Griechenland by Hoepfner and Schwander (fig.
226
241). Still, little is known about the chronology and the architectural development of the building.
The Redoubt Palace had a similar chronology with the Citadel Palace. It was probably built in the 3rd c. BC and rebuilt in the first half of the 2nd c. BC (Leriche, 2003: 176; Rostovtzeff, 1938: 37). In its second phase of construction, the building was enlarged but preserved the main principles of its initial plan layout (Rostovtzeff, 1938: 37). More or less contemporary with this reconstruction, a temple dedicated to Zeus Megistos was built to the south of the palace. The palace and the temple were presumably associated and communicated by means of a large peristyle court in between (Rostovtzeff, 1938: 14; Nielsen, 1994: 116).
The palace had a square plan with a central courtyard and had a surface area of approximately 1,300 m2. On the southern side of the façade were the remains of a row of four columns which were assumed to have been the members of a longer colonnade of a large peristyle forecourt (Rostovtzeff, 1938: 47; Hoepfner and Schwander, 1986: 220). Remains of a staircase and, according to Pillet the thickness of the walls, indicated that the palace had at least one upper story (1933: 21). In Pillet’s plan, the palace was surrounded by and communicated with residential units on three sides. There is no information on the chronology of these units, but it is probable that they were adjoined the main building in later phases of occupation.
The inner court of the palace measured about 17.00 m x 12.00 m and was surrounded by rooms on four sides. On the southern and western sides were the larger reception rooms that opened to the court via porticos, each with two columns. On the north and eastern sides were smaller spaces, most probably the service rooms and the family units that communicated with the courtyard through longitudinal narrow corridors.
The main access to the palace was provided from the southern façade. The entrance door was located behind the colonnade of the large forecourt and gave way to a two-partite entrance hall (F-E). On the south-western corner of the building, in connection
227
with entrance hall F and entrance hall D to the north, were two small interconnected rooms that probably functioned as guard rooms to control the entrances. To the east, the entrance hall F led to a reception suite that was formed by two, or perhaps three, large and communicating rooms (J-K-M). The largest room had almost the same width as the courtyard and opened to the southern portico through a wide central door. The annex room (K) measuring 8.10 m x 7.70 m was also one of the largest rooms of the house. It opened to a small corridor to its north through which it provided access to the inner court (L). A third room (M) which also opened to that small corridor, probably served to this suite as a service room, a reception room for more intimate meetings, or as the restroom of the master of the house.
On the western front of the palace, an open court gave access to the building through two different entrance doors. This court was situated on the edge of the steep slope of the ravine and was presumably accessed by a staircase that had totally disappeared before the excavations (Pillet, 1933: 21). The architectural layout and the circulation patterns of the western wing of the building were similar to the southern one. A rectangular entrance hall (D) gave way to a large hall (A), most probably a second reception room. This hall could also be entered directly from the western court. To the east, it opened to the colonnaded portico of the courtyard through a wide central and two narrow lateral doors. It had an annex room (X) to the north. Similar to room K, annex room X communicated with the inner court and with room W through a small corridor (V). Room W had a window looking towards the ravine and was connected to the adjacent room T with a narrow door.
The northern and eastern sides of the inner court were surrounded by walls. Behind the walls, longitudinal corridors (U-N) provided circulation between the back rooms and the courtyard as well as between the rooms themselves. At the northern end of corridor N was located a staircase with four flights. Under the first flight of the stairs was a small room that was equipped with latrines. Room O on the east side of the courtyard provided communication with the residential units that were adjoined the palace.
228
In the excavation reports, Pillet identified three different dwelling units on the north-east, south-east, and south-west sides of the palace complex. Since these units were only partially excavated and were not dated, they will be only briefly mentioned. According to Pillet, the unit located to the north-east of the palace was contemporary with the main building and was probably used as a pleasure pavilion or a festal hall by the family (1933: 25). This unit comprised an open court (Y) with a two-columned portico to the west and a gallery with three windows opening to the north. The northern part of the building was not searched since it fell into the ravine. To the south of the court was an entrance hall (22) that provided access from a longitudinal inner street or corridor (20) and two annex rooms (21-23). This unit was lower than the main palace and communicated with it through room O by means of a door equipped with several steps.
To the south-east of the palace, a dwelling unit of considerable size was partially uncovered. This unit was organized around an inner court (11). To the west of the court, a large room with benches (9) and its annexes to the south and north (7-8-10) were placed. To the north of the court was room 12 and a staircase (13) with four flights similar to that of the main palace. To the east of the court were some interconnected rooms (24-14-15-17-18) that extended along the eastern façade of the palace. Since this unit was adjoined the very thick walls of the palace and extended over the eastern part of the presumed large forecourt, it is likely that it was a later construction.
The Redoubt Palace continued to be occupied in the Roman period. In its later phases of occupation, some alterations were made especially in the circulation patterns. The door between the entrance hall F and reception room J was blocked. Similarly, the door of entrance hall D and that between the entrance hall and guardroom G was also blocked and room D was turned into a kitchen. The court to the west of the building was paved with small bricks and a passage between the forecourt and the lateral court was provided and controlled through the newly added room 2.
229
7.2.3 Insula C7
Insula C7 is one of the few housing blocks that was entirely excavated in Dura-Europos (fig. 242). It was initially excavated by the British army in the 1920s and totally cleared between 1930-1932, during the fourth and the fifth seasons of the Yale University-French Academy mission conducted first by Pillet and then Clark, under the directorate of Hopkins. In 1994-1995 Catherine Saliou resurveyed the insula and conducted several small-scale excavations (Saliou, 2005).
The insula was located on the eastern part of the city, to the south-east of the agora, on the intersection zone of the two main streets. The street coming from the Palmyra Gate intersected with the north-south extending street H at this point, which was adorned with an arch. The main street narrowed behind the arch and was aligned by a row of shops in block B8 to the north and by the housing block C7 to the south.
Considering its location within the urban fabric, Saliou suggested that the establishment of insula C7 could go back to the beginnings of the urban expansion of the Hellenistic city according to a grid plan (2005: 65). Excavations revealed that the houses in the block were built gradually and underwent several phases of construction and numerous modifications in their circulation patterns. A median longitudinal wall seems to have played a key role in the initial division of the building plots within the insula (fig. 243). The continuity of this wall, nevertheless, was interrupted in the northern extremity. The housing units in this northern part were also smaller, irregularly shaped, and were closely interlocked, unlike the large and more regular units built in the south part of the block. According to Hopkins, this scheme suggested that the houses in the northern part of insula C7 were built before the regular lot divisions were made, whereas according to Baird the wall was likely to have been removed from this part of the insula during a reconstruction (Hopkins, 1934: 35; Baird, 2006: 118).
230
Twelve housing units were distinguished and designated in the insula, relying on the presence and the layout of the courtyards as well as on the circulation scheme (fig. 244). Some of these units communicated with each other and probably belonged to the same house, at least in the latest phase of occupation. Eight separate houses were determined for the latest phase (A/A2/B/B2-C-G-G2-G3/C2/D/E/F) (fig. 245). The unearthed houses were labeled by letters following the Yale convention. The units that were thought to have communicated in any phase shared the same letters and were denominated as, for instance, G, G2, and G3. This scheme continued in the room numbers which were continuous in some communicated units.
7.2.3.1 House C7A
The houses excavated in the northern part of insula C7 presented a rather irregular plan layout presumably due to the several reconstructions and modifications in the circulation patterns throughout its long period of occupation. Three units, A, A2 and B, in that section occupied almost one-eight of the surface area of the insula which were supposed to be built according to the original division of the regular housing plots within the city (Benech, 2010; Hoepfner and Schwander, 1986: 225). The archaeological work revealed that these units, which functioned as separate dwellings in their latest phase of occupation, communicated with each other in at least one of the previous phases.
House C7A was accessed from the main street to the north. The entrance door opened to the narrow corridor 9 that gave access to courtyard 1 through an L-shaped path. A blocked door between corridor 9 and room 10 of House C7E explains the unusual form of the corridor that had a dead end to its west. The unpaved courtyard of House C7A measured about 40 m2 and gave way to three rooms on its east, west and south. An L-shaped staircase that rested against the east wall of the court gave way to either an upper floor, about which there is no archaeological evidence, or to a terrace roof (fig. 246). On the north-eastern corner of the court, between the staircase and a partition wall that screened the entrance of the house, a semi-open space was determined and
231
was identified as a kitchen in the excavation reports (Hopkins, 1934: 36). A plaster basin was placed in this area just behind the staircase. In the middle of the courtyard was a cesspool/cistern dug into the ground.77 Two massive plaster bowls were placed in the courtyard, one next to the entrance door of room 7 and the other in front of the staircase (fig. 247). These bowls, similar to those found in all the unearthed houses in Dura-Europos, were identified as coolers by the excavators as they were thought to have been used to keep water.
Room 7 on the western limit of the house is likely to have been built over the courtyard in a later phase of occupation, probably when room 9 and room 10 of House C7E were still communicating, as the plan layout of corridor 9 suggested. It had a wide and axially placed entrance from the courtyard that demonstrates the prominence of this room within the house. On the opposite side of the courtyard was located room 10 which had low and wide masonry benches on all its four sides, which probably received the carpets and cushions to sit, was located. The placement of room 10 on the corner of the courtyard, and its non-axial entrance, as well as the considerable width of the partition wall between this room and room 3 indicated a possible lateral modification in that part of the house. The blocked door between room 10 and courtyard 6 of house C7A2 may point out that room 10 originally belonged to the latter house. There were several niches above the benches on the walls of room 10. On the west wall of the room were scratched a series of graffiti.78 Across room 10, to the south of the courtyard, two larger and interconnected rooms were located. A door that was placed almost centrally on the southern wall of the courtyard opened to the north-eastern corner of room 3 and a second symmetrically placed door in the same room
77 Pear-shaped underground features with rectangular slabs covering their mouth in the middle of the courtyard were found in almost all of the unearthed houses in Dura-Europos. They were interpreted as cesspools or as cisterns by the excavators. In the house descriptions given in this section these features are mentioned as cesspool/cistern; they are discussed in more detail in the overview part at the end of this chapter.
78 Since the plaster of the walls of room 10 was badly damaged only a part of the graffiti could be read. In that part of the graffiti there were some letters of a name which, according to Hopkins, could be Tiberinus or Tiberianus (Hopkins, 1934: 36)
232
gave way to room 2. This latter room had two blocked doors on its east and west walls that provided communication between room 8 of house C7B and room 4 of house C7E at one time. Beneath room 2 was a small cellar that was closed by a plaster cover on the floor and reached via a small staircase of a few steps. This small and low cellar most probably served as a cool storeroom.
7.2.3.2 House C7A2
House C7A2, consisting of a courtyard (6), a small chamber (8), and two shops (4-5), and measuring about 75 m2, was one of the smallest units of the insula. Originally, it probably comprised room 10, too, as noted above. The main access to the house was provided from the eastern street, through a small entrance hall that gave direct access to the courtyard. The staircase that rested against the eastern wall of the courtyard gave way to an upper level, presumably to a flat roof over rooms 4 and 5. A figurine in the form of a camel used to burn incense and another one in the form of a bird standing on a small altar found in the courtyard indicated the occasional use of this open space as a cult area by the family, besides its other functions. Interconnected rooms 4 and 5 had separate entrances from the northern street while the former communicated with the courtyard, too. Both rooms probably functioned as shops that belonged to the owner of house C7A2. The immense quantity of coarse pottery fragments was found in room 4 supported this suggestion.
7.2.3.3 House C7B
House C7B was a small unit consisting of a small court and two chambers. The narrow entrance hall gave way to courtyard 7. Next to the entrance hall was the smaller room 9. Across the court was the larger room 8 that was adorned with a low bench on its four sides and a molded plaster frieze on the walls. As mentioned above, room 8 had previously communicated with room 2 of House C7A.
233
7.2.3.4 House C7B2
House C7B2 is distinguished with its unusual linear layout. Probably because of the fairly narrow plot allocated to the house, its rooms were not arranged around a central courtyard but they were aligned in axial order. The entrance hall (5) was on the south-eastern corner of the house and led to the small rectangular courtyard to which rooms 6 and 2 opened. Room 2 was the largest space and was the principal room of the house, with benches on four sides and decorated moldings on its walls. Two adjoined and communicating rooms, 3 and 4, opened to room 2, via two doors on its west wall (fig. 248). The staircase that rested against the east wall of the courtyard provided access to an upper level or to the roof. Three coolers and a large jar were found in the courtyard, one cooler was placed at the end of entrance hall 5 while the two others and the jar at the north-west corner. House C7B2 communicated with House C7C through a door between the two courtyards in the latest phase of occupation of the insula.
7.2.3.5 House C7C
House C7C was organized around a central courtyard, just like most other houses in the insula. An L-shaped entrance path that was formed by a partition wall across the entrance hall led to the courtyard. The partition wall also formed a corridor-like space for the passage between House C7C and the adjacent house C7B2. Behind that corridor, between the partition wall and the stairwell, was a deep narrow space that opened to the courtyard. The staircase of this house was more spacious than those found in the other houses. It had three flights and was enclosed by a wall. To the south of the courtyard was room 14, a rectangular space divided by a thin partition on either side of a central column. This room contained a high plaster socket for a wooden pole and was identified as a kitchen, though no trace of a cooking or food preparation installation was found (fig. 249) (Hopkins, 1934: 38). A door that was blocked on the back wall of room 14 once provided communication with house C7C2. Just outside of this room, on the south-east corner of the courtyard, a large jar was found. On the western side of the courtyard were located room 11 and its annex room 12. Room 11
234
was the largest room of the house with benches all around and opened to the courtyard via a wide opening. A door opened later at the back wall of this main room provided communication with House C7G in the latest phase of occupation. To the east of the house were rooms 3 and 4 which communicated with each other and also with the courtyard. There were two niches on the north and west walls of room 4, near the two doors. A series of graffiti was scratched on the walls of the same room. Among them were the representations of deer, a riverboat propelled by a man, a buffalo approaching an altar, and the front of a temple with two gladiators standing inside and two eagles next to it (fig. 250). According to Baird, the graffiti and the existence of some military equipment among the few recorded artifacts could attest to the use of the house by the army personnel which, she suggested, also explains the later addition of doors between the different housing units of the insula (2006: 128; 2014: 145-146).
7.2.3.6 House C7C2
The entrance scheme of House C7C2 was similar to that of House C7C. At the end of the entrance hall, was a partition wall that projected into the court and defined a small space like a secondary entrance hall which opened to the courtyard and also provided passage to House C7C, before the door on its north wall was blocked. The partition wall was pierced by a window, behind which was a tannour located in a roofed area that was separated from the courtyard by two columns. The small room 3, on the north-west corner of the courtyard, opened to that cooking area and seems to have served, presumably, as a food storage room. Several niches in the back wall of the room were installed in place of a blocked door which gave way to House C7G in an earlier period (fig. 251). On the western end of the courtyard was placed a linear staircase. The main room of House C7C2 was room 8. Measuring about 6.00 m x 11.00 m it was one of the largest and pretentious rooms within the insula. The room was located on the south side of the courtyard and opened to it via a monumental doorway with a few steps and a wide door of about 2.50 m. There were no benches in this room. Room 8 had two annexes, rooms 7 and 2. Room 2 was a small room located to the west of the courtyard. In a later phase of occupation, a door was opened between rooms 3 and 2. Room 7
235
located on the eastern side of the main room on the other hand was more spacious and it communicated with room 6 that also opened to the courtyard. A graffiti depicting a boat in room 6 and fragments of painted plaster wall decoration in room 2 was recorded by Saliou during her reexamination of the insula (2005: 76).
7.2.3.7 House C7D
House C7D was one of the most spacious houses of the insula, but its poor state of preservation made it difficult to ascertain its exact layout (Hopkins, 1934: 40). Since the house was entirely built of mud brick, as opposed to the rubble or rubble with mud-brick walls in other houses, its walls were completely demolished. The remaining walls and the foundations that were shown on the published plan indicated that House C7D had a different plan than its neighbors. It was entered via a fairly large entrance hall (5) that gave access both to the courtyard, through a door in room 11, and to a suite of rooms that were independent of the courtyard, through a narrow corridor. Room 24 to the west of the median wall of the block had apparently been taken from House C7G3. Not only its location but also the blocked door between this room and room 22 of House C7G3 and the lately opened door on its back wall that provided communication with room 3 of House C7D attested a reorganization.
7.2.3.8 House C7E
House C7E was located on the north-western corner of the insula and was accessed from the main street. The entrance hall (10) on the north-eastern edge of the house gave way to the courtyard (1) but any direct view of the court from the entrance was prevented by a partition wall built in front of the entrance hall. Beyond this partition wall was formed a semi-open space (9) which probably served as a service room. Next to the entrance hall, on the eastern side of courtyard 1 was a stairwell with a two-flight staircase, and room 2. At the back of room 2, a very narrow doorway opened to room 3, which, according to Hopkins, could have been used as a storeroom for valuables because of its remote placement (1934: 41). To the west of the courtyard was room 7.
236
The largest space of the house, measuring 8.30 m x 5.50 m, was room 5 that was entered from court 1 through a wide door and a few steps. This space was initially identified as an inner court by the excavators, but its location and layout make it more likely that it was the principal room of the house (Pillet, 1933: 35). To this room, another spacious room (4) and a narrow room (6) opened on its east and west sides. The blocked doors attested that room 4 once communicated with room 2 as well as room 8 of house C7B. Two Heracles reliefs, a painted glass fragment with the inscription “Thetis” and a number of coins dating from 225 BC to 256 AD were discovered in the house. Several features that are shown on the plan, nevertheless, were not described in the excavation reports.
7.2.3.9 House C7F/The House of the Frescoes
House C7F or the House of the Frescoes had its front façade on the main north-south street of the city. On this façade, the entrance hall of the house and three shops were aligned and opened to the street. The floor level of the house was lower than that of the street and the house was entered via a few steps. An elongated entrance hall (7) on the north-western corner of the building led to the courtyard. The west wall of room 8 to the north of the court defined the boundary of the entrance hall and prevented any direct view to the core of the house, just like the wall of annex 9 of House C7E. A large stone bowl, or cooler, was found in the entrance hall. To the east of the courtyard two interconnected rooms, 2 and 10, were located. On the walls of room 10 and room 8 traces of poorly preserved wall paintings were observed. In front of these rooms, resting against the east wall of the courtyard was a staircase under which a latrine was installed into a recess. The largest room of the house, measuring about 8.70 m x 5.80 m, was located to the south of the courtyard and opened to it via a wide door of 1.91 m. A low masonry bench with a height of 0.23 m and a width of 1.40 m surrounded the room. On the walls of the room were two layers of painted stucco. The best preserved and the most elaborate painting was on the south wall opposite the entrance, after which the house was named (fig. 252). In this figural painting, a battle scene with eight horsemen combatting and pursuing each other was represented. At the top of the
237
scene were depicted people reclining on a row of beds as if they were watching the battle, and at the bottom were some running animals. The scene was bordered by a geometric pattern of a checkerboard and triangles in red, white, and black colors. There were also inscriptions in Pehlevi in the painting.79 Based on the costumes and arms of the two groups depicted in the painting, Pillet interpreted this scene as an encounter between Parthians and desert nomads, while Rostovtzeff and Little suggested an encounter between the Iranians and Romans (Pillet, 1933: 36; Rostovtzeff and Little, 1933: 174-175). Traces of paintings were observed also in the east and west walls of room 4, among which only two horsemen and three ornate columns were distinguished. In the east wall, two graffiti on both sides of a niche were preserved. One of them included the name “Leg III Cyr” and the other was a sort of a menu listing vegetables, lentils, wine, milk, and desserts. The location of room 4 in the coolest south side of the courtyard, its size and layout as well as its decoration indicated that it was the principal room of the house, which should have functioned as a reception room.80 Room 4 had originally two annex rooms (3 and 5) to its east and west, the latter of which was turned into a shop in a later phase. House C7F had three shops, rooms 6, 5 and 13, on its front façade. The blocked doors indicated that these rooms communicated with each other and room 13 communicated with house C3G, in an earlier phase(s) of occupation. Shop 6 communicated directly with the courtyard of the house, too, which, according to Baird, was the result of later reorganization (2006: 140).81 During the excavations three large earthenware jars that were sunk into the ground, a large jar in room 13 and the fragments of a considerable number of vases were found in room 5.
79 Pehlevi is an Aramaic-based writing system used in Persia from the 2nd century BC to the advent of Islam in the 7th century AD.
80 Room 4 of House C7F was identified as an inner courtyard, just like room 5 of House C7E, by Pillet in the preliminary report of the fourth season of the excavations (1933: 35). Rostovtzeff and Little, on the other hand, identified this room as liwan (1933).
81 Rostovtzeff and Little identified room 6 as a chamber of the house with a window opening to the street, instead of a shop (1933: 168).
238
7.2.3.10 House C7G
The entrance and the circulation schemes of House C7G were similar to those of the other houses in the same insula. The entrance hall (16) led to a courtyard via an L-shaped path. Remains of four columns found fallen on the ground of the relatively spacious courtyard indicated the existence of a portico.82 The discovery of a gypsum relief of the god Hadad led the excavators to suggest that there was probably a domestic shrine in the courtyard that was sheltered by the portico (1934: 42-45). On the north side of the house, in front of the entrance hall, were a small room (15) and an adjoining stairwell. The main room of the house (4) with its monumental entrance towards the court was located to the south (fig. 253). The southern wing of the house presented an unusual layout with the corridor (8) located between the main room (4) and its annex room (10). Serving as a passageway between House C7G and House C7G2, this corridor was probably the outcome of a later reorganization in that part of the building. According to Hopkins, rooms 5, 7, and 9 of House C7G2 originally belonged to House C7G, and corridor 8 was designed to serve them (1934: 45). According to Baird, nevertheless, the houses were separate in origin and were connected in a later phase of occupation, which seems more likely (2006: 147). During this renovation, she argued, a part of room 4 was taken to build corridor 8. Likewise, the blocked doors in rooms 3, 4, and 16 of House C7G, which once gave way to Houses C7C2, C7G2, and C7F respectively, attested that there were at least two phases of communication between these houses. To the east of House C7G, rooms 2 and 3 opened to the courtyard. Room 3 also communicated with the main room and room 2 provided passage to House C7C to the east. Room 11 to the east was also a spacious room built at a little higher level than the courtyard. There was a row of niches in the back wall of this room and two coolers were placed on both sides of its entrance. Another cooler was found in the courtyard, which was placed close to the entrance door of room 11.
82 In the excavation report Hopkins recorded that the bases of the columns had disappeared (1934: 41). For this reason their placement in the courtyard is unknown and the portico has not been shown on any of the published plans.
239
7.2.3.11 House C7G2
House C7G2 was a small unit that communicated with Houses C7G and C7G3 in its final phase (fig. 254). The narrow entrance hall (room 14) led to a small courtyard with a staircase. There was a cesspool/cistern in the courtyard, at the edge of the staircase. A cooler and a gypsum torso of a male figure with a ram were found in room 14. The courtyard of C7G2 was connected to the courtyard of the adjacent House C7G3 by a door on its south-western corner. Room 7 located to the north of the courtyard and communicated with the other three rooms, 5, 9, and 12 was the main room of the house. A door placed at the back of this room provided passage to House C7G.
7.2.3.12 House C7G3
House C7G3 was entirely built of mud-brick just like House C7D located to its east. Access to the house was provided from the street on its west, through the entrance hall 17 that was adjacent to the entrance hall of House C7G2. The house had a relatively spacious courtyard but no staircase was recorded. Room 21 to the south of the courtyard was the main room that had two annex rooms, 23 and 20, in origin. Room 20, nevertheless, was turned into a shop at some time, but it continued to communicate with rooms 19 and 21. Three projecting walls on the east wall of the courtyard, in front of room 22, and a linear feature on the south wall, to the east of the entrance of the main room, were shown on the plan but they were not mentioned in the reports. Room 24 of House C7D was once part of that house and communicated with room 22 as the blocked door in between attested.
7.2.4 Agora/Sector G
The Hellenistic agora of Dura-Europos covered an area of eight regular insulae in the center of the city that measured about 150.00 m x 160.00 m in total. The Seleucid Agora was originally a large open space bordered by shops on three sides (fig. 231). From the beginning of the Parthian period, the area was gradually covered by
240
dwellings, most of which were associated with shops; despite transforming into a residential sector it remained as the commercial center of the city throughout its history (fig. 232).
The agora section, or sector G in Yale designation, was first excavated in 1923 by Franz Cumont and in 1924 by the British soldiers stationed at the site (Cumont, 1926: 242-250; 476). A large part of the area was exposed by the Yale University-French Academy mission during the fifth, seventh, and tenth excavation seasons of excavations that took place in 1931-32, 1933-34, and 1936-37 respectively. The results of the excavations were published in 1944, in Preliminary Report 9.1, a volume dedicated only to this sector (Rostovtzeff et al., 1944).
The archaeological evidence attested that the conversion of the regularly planned Seleucid Agora to a rather irregularly developed and crowded commercial area, or to an “oriental bazaar” as Frank Brown identified, was a long process that continued until the end of the city (Brown, 1944b: 28-68). According to Brown this transformation started by the integration and remodeling of several shops to form modest dwellings (1944b: 43). He also argued that the temporary structures in the open area of the agora might have become permanent and enlarged in time. The buildings in the sector, either residential or commercial, witnessed several phases of reconstructions and/or restorations. By the final phase of the city, around 256 AD, the area was occupied by eight insulae (G1-G8) that did not conform to the city grid in size and layout. The excavation reports recorded 56 dwellings and 182 shops within the agora section (Brown, 1944b: 65). Nevertheless, a number of buildings that were initially identified as dwellings by the excavators were redefined as commercial establishments like brothels and bars in later studies (Baird, 2006: 249).83
83 Unit C in insula G5 was initially identified as a house and then as a brothel (Brown, 1944c, Baird, 2006: 327). The wall inscription mentioning a group of prostitutes and entertainers who had travelled to and from Zeugma came from this unit.
241
7.2.4.1 Insula G1
Insula G1 was the largest building block on the eastern side of the agora. It was almost entirely unearthed, except for a small area behind the shops in the south-western corner (fig. 255). The southern façade of the insula was aligned by shops that were entered directly from the street. There were also shops on the east and west sides, as well as around the central open court in the middle of the insula. A colonnade was built in front of the shops located in eastern half of the southern façade in the Roman period. Seven housing units were identified in insula G1 (Houses A-G) in the excavation reports. Among them units D and E, found on the south-western corner of the block, were only partially excavated; the unit G seems to have been a commercial establishment since it did not have the usual house arrangements or features. Houses G1A and G1B, the two larger houses uncovered in insula G1, will be examined in more detail (fig. 256).
7.2.4.1.1 House G1A
House G1A was the largest and most elaborate domestic unit in insula G1 (fig. 257). It was dated to the early 2nd c. AD and was built on virgin soil, in the open area of the Hellenistic Agora (Brown, 1944c: 136). The earlier buildings to the south of the house were demolished to the ground and an open inner space within the insula was created.
House G1A was accessed from the street to the north through an L-shaped entrance hall (room 25) in its north-western corner. The entrance hall was divided into two sections by an arch and gave way to the central courtyard. Both the entrance hall and the courtyard of the house were paved with brick. The courtyard was equipped with a central cesspool/cistern that was covered with a square pierced gypsum slab and a cooler was placed in front of the door of main room A3. To the north of the courtyard were a central stairwell and two side rooms that were entered through arched openings from the court (A26 and A27). In the first landing of the staircase, a wide bay located about 0.65 m above the street level opened to the exterior. According to Brown, this
242
opening could have been used for loading and unloading heavy goods from carts or pack animals (1944c: 140).
From the south side of the courtyard was entered the main room of the house (room A3) via a wide opening with double doors and two stone steps. The floor of the room was of beaten earth and there was no sign of a masonry bench but it yielded a pilaster bust of a woman probably used to support a wooden bench (Hopkins, 1934: 50). Room A3 communicated with four different rooms (rooms A4, A2, A21, and A20) on its three sides. Rooms A4 and A20 were relatively larger and were entered via double doors. The floors of all the four rooms were of beaten earth, too. A large dolium and a pointed amphora were found on the floors of rooms A20 and A4 respectively.
To the east of the courtyard was a rectangular room with an axially located door. The function of this room could not have been identified. A door opened on its west wall in the latest phase of occupation provided passage to the neighboring House G1B.
On the east wall of the main court were two doorways with double doors that gave way to two different groups of rooms. The first group was composed of rooms A28, A36, and A38. This suite was interpreted as the “women’s quarter” by Hopkins and the rooms were identified respectively as the probable “kitchen”, “servant’s quarter”, and the “storeroom” by Brown (Hopkins, 1934: 49; Brown, 1944c: 141). Since this suite of rooms yielded significant remains, like a bronze pin, a stone and a bronze altar, a bronze pendant, a few armor plates, and a basalt platter it is more likely that it functioned as the living quarters of the family. The staircase along the south wall of room 36 indicated the existence of an upper level or the use of a terrace roof in this part of the house.
The second group of rooms on the east side of the courtyard comprised rooms A14 and A32. A14 was identified as stable since there was a 1.32 m high rubblework manger at with a 0.38 m deep trough (Brown, 1944c: 141). The floor of the room was of beaten earth and was roughly flagged with gypsum. The interpretation of room 14
243
as a stable, however, was challenged by the finds it yielded. On the wall above the doorway from the courtyard, a bas-relief head of a divinity was found. Among the other finds unearthed were a bronze clasp pin and a three-handled green-glazed vessel. Room A32, located behind room A14, seems to have provided passage to the service quarters located at the eastern edge of the house.
House G1A had a secondary courtyard (A39) that had indirect access from the main court through rooms A14 and A32, and also from the principal room A3 via rooms A20, A17, and A12. It was also reached from the open space to the south of the house through rooms A19 and A12. The location, size, and irregular layout of court 39 indicate its use as a service courtyard. Its floor was paved with brick and had a cesspool/cistern. An alcove-like space on the north-eastern corner of the court that was differentiated with an unpaved floor yielded a painted amphora. In a later phase of occupation, the eastern part of the court was separated by a thin wall and room 34 was created. The northern half of this room was also paved with brick and the floor of the southern half that was raised about 0.08 m was of plaster mortar.
Room A12 to the south of the second court was identified as the “women’s diwan” by Brown who also labeled court 39 as “women’s court” (1944c: 141-142). Nevertheless, he did not base his suggestion of a gender-based spatial differentiation on any archaeological evidence. Room A12 having single or double doors on three sides, nevertheless, seems more likely to have been a service and/or passage space. It gave way to rooms A17 and A19, the latter of which had an entrance from the open space to the south and contained two coolers. Three rooms, S1, S11, and S 120, whose construction was contemporary with the house, were identified as ‘warehouses’ and associated with the house, though they had no direct communication (Brown, 1944c: 142). They opened to the open public space to the south and could well have functioned as shops, owned by the house owner or someone else. Among them, room S11 yielded a large dolium and a cooler.
244
7.2.4.1.2 House G1B
The construction of House G1B was dated to a slightly later period than that of house G1A since its walls abutted the western façade wall of the latter (Brown, 1944c: 142). The house was built on virgin soil, too.
House G1B was more modest in size and presented a more typical plan layout than house G1A (fig. 258). It was accessed from the street to its west, via entrance hall 31 that was located across an empty space in front of the Chreophylakeion in block G3. The level of the plastered mortar floor in the entrance hall was lower than the street level and it was reached by two steps. There were two more steps at the end of the entrance hall under the arched passage leading to the inner courtyard. The floor of the courtyard was of beaten earth and a 0.27 m high narrow bench of rubblework ran along its west wall and the western part of its south wall. Two columns with Doric capitals formed a short portico on the eastern side of the courtyard. The southern corner of the portico was suggested to have been a stable by Brown because of a 1.38 m high trough inserted between the southern column of the portico and the wall across (1944c: 144). To the west of the courtyard, a stairwell with a three-flight staircase led to an upper level. Beneath the first flight in the north-west corner of the stairwell was a cistern.
The main room of House G1B (room B8) was located to the south of the courtyard and was entered via a large door and two stone steps. Its floor was of beaten earth and a wide and low bench surrounded all four sides. Room B8 was annexed by rooms B23-B24 and B7-B11 on its west and east sides respectively. The floor of room 23 was lower than that of room B8 and there was a step in the doorway. In room 23 a bronze bell and a bronze rosette were found. Room 24 was a kind of corridor that provided passage from room 23 to the courtyard. The communicating rooms B7 and B11 had several niches in their walls.
House G1B originally had a second court (room B37) on its northern edge to which the spacious room B29 opened. The cesspool/cistern of the house was placed in this
245
court instead of the main court. In a later phase of occupation, the court was roofed and was divided into three sections. A37 became a room, probably for domestic activities. A circular depression in the middle of the room and a low rubble parapet with a u-shaped depression in front of it were interpreted as the places of loom and wool dyeing equipment (Hopkins, 1934: 53; Brown, 1944c: 146). Along the eastern wall of the room a staircase of seven steps reached to a gallery or loft. To the south of room B37, a small area with the cesspool/cistern was walled up and a small corridor providing passage between the main court and room B29 was formed.
On the north-western corner of the house, close to the entrance door, were two interconnected rooms, B33 and B35. From room B33 came fragments of two pointed amphorae buried into the ground, a large basalt grinding stone, and from room B35 a cooler.
Room B21 found on the east side of the courtyard was entered through an arched passage without any doors. On its east wall was a niche and against the east wall stood a 0.52 m high bench. On the beaten earth floor of the room was a cooler. There was another cooler in front of the door of room B21 in the courtyard. In the latest phase of occupation room B21 communicated with room A13 of house G1A.
7.2.4.2 Insula G3
Insula G3 was situated in the north-eastern corner of the agora (fig. 259). It began to develop from the earliest period of Dura-Europos and the unearthed remains represented subsequent construction phases. The Chreophylakeion or the archive building of the city, shown with the letter A in the plan, was located on the south-western corner of that insula. In the final period of the city, there were ten houses and two independent shops with two or three rooms in the insula, besides the archive building.
246
Remains of the neighboring houses G3B, G3C and G3G in insula G3 demonstrated well the transformation of that part of the Seleucid commercial area into a residential district with several phases of constructions and restorations (figs. 260, 261). House G3G and the northern part of the latest House G3B (the part comprising rooms B6-11) were contemporary and occupied the existing area of the Seleucid shops in the early Parthian period. Houses G3C and the southern part of House G3B (the part comprising rooms B1-5) were also more or less contemporary and built on virgin soil in a later period, in the 120s or 130s according to Brown (1944c: 91).
7.2.4.2.1 House G3G
House G3G initially occupied the area of three existing Seleucid period shops, each with two rooms; G1-G6, G2-G7, and G3-G4 (figs. 270, 271). In the initial phase of the residential occupation, the walls of the shops remained in-situ and the rooms of the house were fitted into the existing spaces with very little structural changes (Brown, 1944c: 90). These changes included the opening of a new door between rooms G6 and G3 to connect the shops to the house, and the blocking of the street entrance of the former shop G2. In this phase, the entrance door of House G3G was on the north façade and opened directly to court G1 that had a cesspool/cistern dug into the ground. The court gave way to main room G2 to the west via a wide door and to room G6 to the south. Room G2 was interconnected with room G7 whereas room G6 led to shops G3 and G4.
The construction of house G3B (part comprising 1-5) to the south of the insula G3 blocked the southern façade of the earlier houses. Accordingly, the street entrance of shop G4 was walled up which necessitated a rearrangement in that area of the house. The Seleucid era partition walls between shops G4, B6, and B10 were demolished and the area of the three shops was divided into four rooms by new partitions. In this reconstruction, room G5 took a 1.00 m wide space from shop B6. At the same time,
247
the existing doorway between rooms G3 and G4 was walled up and a new doorway was opened on the same wall.
House G3G remained almost unchanged until shortly before the end of the city when the northern portion of the house was extensively remodeled probably due to the cumulative rise in the level of the adjacent street to the north (Brown, 1944c: 92). In the latest phase of occupation, the floor level of court G1 and rooms G2 and G7 were raised about 1.00 m. The court was filled with earth and debris and cellars were built beneath rooms G2 and G7. The original partition wall between G2 and G7 was replaced with a new one that was built about 1.40 m to the south of the earlier and at the expense of room G2. Similarly, the partition walls between rooms G2-G1 and G7-G6 were demolished and rebuilt on the existing foundations. The floor level of room G6 remained unchanged since it gave way to the cellars via an arched opening. The courtyard with a new floor level was rearranged with the addition of a covered porch, and the construction of a new staircase, during which the mouth of the cesspool/cistern was also raised.
In its latest phase of occupation House G3G was entered from the northern street via double doors and a single step. The entrance of the house was screened with a partition wall and a single column that formed a very narrow interior porch at the northern edge of court G1. The floor of the porch, as well as the courtyard was of beaten earth. There were two coolers in the porch and two more in the courtyard. Along the east wall of the court ascended a staircase that was raised on three arches and rubblework piers. On one of the walls of the court was a bas-relief of Heracles (Hopkins, 1934: 67). A wide door to the west of the court opened to the main room of the house which contained wide rubblework benches of 0.14 m high along the west and south walls. The floor of room G2 and that of its annex room G7 to the south were paved with plaster mortar whereas all the other rooms of the house were of beaten earth. From a doorway, with double doors on the south side of the courtyard room G6 was accessed via a few steps. This room gave way both to the cellars beneath rooms G2 and G7 and
248
also to rooms G3 and G4-G5, that is, to the south of the house that was probably used as the living quarter of the family.
7.2.4.2.2 House G3B
House G3B (6-11) initially occupied the area of three Seleucid period shops, just like neighboring House G3G. To transform the shops into a dwelling the existing spaces were redefined and reorganized (figs. 270, 271). The entrance door of shop B6 was walled up, a new doorway between shops B7 and B11 was opened, two shop walls in B7 were demolished to build the small room B8.
Contemporary with the construction of Houses G3B (1-5) and G3C, a series of restorations took place in the northern part of House G3B (6-11) as was the case in house G3G. In this period, the street entrance of shop B10 was blocked and the partition walls between rooms B6 and B10 were reconstructed. In addition, the door between rooms B10 and B11 was blocked and room B10 was annexed to the main room B6 by opening a new door in between. During the same restoration process, the wall between rooms B7 and B11 was demolished and a new wall was built, about 1.00 m to the east of the earlier one.
The southern part of House G3B was built in the open public space in front of Block G3, sometime after the transformation of the northern shops into dwellings. This unit initially had no communication with the northern part of House B. The two houses were combined to form a single dwelling, presumably at the end of the 2nd or the beginning of the 3rd c. AD, as suggested by Brown (1944c: 92).
In its latest phase of occupation, House G3B had two distinct parts that communicated with each other only through a doorway between rooms B5 and B6. In this phase the house had a single entrance from the southern street since the existing door on the north facade of the house was blocked and the earlier entrance hall B9 became the annex of room B8. The house was entered through a doorway with double doors and
249
two steps that descended to the small entrance hall B1. The entrance hall opened to the courtyard via an arched opening. The floor of the entrance hall as well as that of the courtyard and all the other rooms of the house were of beaten earth. There was a cesspool/cistern in the middle of courtyard B2 and a cooler in front of its northern wall. A staircase which rested against the east wall was raised on three arches. There were terracotta ovens under the second and third arches. Court B2 gave way to rooms B3 and B4 to the west. On the south-eastern corner of room B4 was a large dolium. Room B5 functioned as a passage between the two parts of the house. It was reached either from the courtyard via a very narrow and low arched doorway built under the staircase or via a door from room B4. Room B5 opened to room B6 through an arched passage. Room 6 was a spacious room that was entered from court B7 through a wide door and a few steps. It was interconnected with room B10 that had shallow niches in its north and south walls in the place of the earlier doors. The northern part of court B7 was covered with a roof that was supported by a single column. Two coolers were found in this semi-open space. A staircase was placed on the southern side of the courtyard. A stepped masonry table with a terracotta basin on its top, was placed in the south-west corner of the courtyard.
7.2.4.2.3 House G3C
House G3C was built on virgin soil to the south of insula G3 and was adjoined the wall of the south façade of the existing houses (figs. 270, 271). It was accessed from the southern street and entered via a double door with two steps in front. It had a long, corridor-like entrance hall (C1) that led to the south-eastern corner of the central courtyard. The floor of court C1 and all the other spaces of the house were of beaten earth. There was no cesspool/cistern or cooler in the courtyard but a plaster basin was buried into the ground against the west wall. A staircase was built over two arches that were supported by piers to the north of the court. Behind the staircase, was a single column that supported the roof of semi-open space B6. The main room of the house was room G3 and it was placed to the south of the court. The house had two other
250
small rooms, C5 and C7, which the former communicated both with the courtyard and room C3.
7.2.5 Insula B2
Insula B2 was distinct in many respects among the building blocks unearthed so far in Dura-Europos. It covered a sloped terrain inside the ravine at the eastern edge of the city where it adjoined the south wall of the citadel (figs. 262, 263). The orientation and the size of the insula did not fit into the grid system but presented an irregular plan layout. It was also one of the few excavated blocks of the city where residential units were associated with craft activities.
Insula B2 was excavated in 1932-33, during the fourth season of the Yale University-French Academy mission, with the aim to determine the location of the eastern gate that was believed to be located on the continuation of the main street and to find the houses of the Seleucid period (Hopkins, 1979: 122-123; Allara, 2002: 60). The insula was fully uncovered, under the directorship of Mesnil du Buisson, but the results of the excavations remained unpublished. Between 1989 and 1992, the insula became the subject of the (posthumously published) doctoral thesis of Anny Allara, who overviewed the records of the earlier excavations in the Yale archives and reexamined the remains in the site under the auspices of MFSED expedition (Allara, 2002).
The earliest activity detected in insula B2 was the exploitation of a gypsum quarry that was active in the Hellenistic period, and presumably provided stone for the construction of the nearby citadel (Allara, 2002: 153; 1997: 181; 1992: 106). It was in the late Hellenistic or early Parthian period, when the quarry already went out of use, the area was terraced and the first constructions appeared. In the final days of the city,
251
insula B2 covered an area of about 3,500 m2, and was occupied by two open squares with furnaces and a dozen dwellings.84
On the eastern and south-western sides of insula B2 were two open squares, 23 and 10, with pottery furnaces. Square 23 was at the highest point of the insula and was dominated by one of the towers of the citadel. It was enclosed by walls and accessed from alley 17 via a door on its south wall. By the latest phase of occupation the furnace at the center of the area went out of use and the open square became surrounded by a series of rooms that were identified as shops by the excavators (Allara, 2002: 87). Square 10 was similar to Square 23 in its function and layout. It was enclosed by the high walls of the surrounding dwellings and entered through a gate on its north-west corner. There were two furnaces in square 10, one of which was associated with house D. In the final period, nevertheless, house D was abandoned and the furnace went out of use. Square 10, like Square 23, became surrounded with shops in its latest phase of occupation. A fourth furnace was discovered in the northern part of insula B2 but the considerable degradation in that area did not permit further work.
Archaeological evidence attested that insula B2 was progressively covered by dwellings and witnessed several phases of constructions and restorations, the chronology of which could not be ascertained. As stated by Allara, Houses A, B and C were the earliest dwellings and the center of the insula remained as a large open space for a while, to allow passage to the citadel (1997: 183). In the course of time, probably after the citadel was abandoned, the free spaces in the insula were gradually occupied by more modest dwellings. Two inner alleys, 17 and 18, gave way to some houses. Houses A and C came into prominence among the other units with their size and colonnaded courtyards, and will be examined in more detail below. Houses B and G were smaller, had less imposing courtyards. Houses E, F, H and I were the smallest
84 Its original size is unknown since a portion of the eastern side had collapsed into the river. According to Allara, the collapse might have happened in a late phase of occupation, probably during the earthquake of 160 AD, or after the abandonment of the city (1997: 184).
252
units containing few rooms while the poor state of preservation of houses D and K made it difficult to restore their plan.
7.2.5.1 House B2C
House B2C was the largest property of insula B2, measuring about 630 m2 in its latest phase of occupation (figs. 264, 265). According to Allara, House B2F as well as rooms E6 of House B2E and D5 of House B2D previously belonged to House B2C which originally measured about 920 m2 (2002: 76). The house was located in the south-east corner of the insula and was accessed from the street to the south. Its entrance hall gave way to the central courtyard through an L-shaped path.
The spacious courtyard was surrounded by colonnades on three sides with seven columns in total. On the fourth side, there were two more columns more or less corresponding to the axis of those on the opposite side, thus creating a visual impression of a peristyle to a certain extent. The two columns, nevertheless, differed physically since they rested on a low wall and were located at a higher level than the others. They defined a small semi-open space, room C5, on the northern edge of the courtyard. This space was attached to the stairwell on its south. The space underneath the staircase, room C6, was used as a storeroom. The courtyard was provided with a cesspool/cistern in its central area.
House B2C had two large rooms, C12 and C15 that were furnished with low benches. C12 was the largest room of the house and entered through a wide door and a few steps on the southern side of the courtyard. From room C12 there was access to two groups of small rooms. A door on the south wall of the room gave way to the interconnected rooms C1 and C11, both of which were heavily damaged due to the erosion of the area. A second door on the north-western corner of room C12 opened to another suite that comprised rooms C4 and C2.
253
The eastern side of the courtyard of House B2C was dominated by a second large room, (C15) that was equipped with benches on four sides and a brazier. This large room was annexed with rooms C13 and C17 on both sides. Room C13 was accessible from the courtyard as well.
A sort of corridor extending like a continuation of the northern portico of the courtyard led to room C18 on the northern edge of the house. This was probably an open space and according to Allara an annex of the courtyard (2002: 76). There were two niches in the north wall.
House B2C did not differ much from the unearthed houses of Dura-Europos in terms of plan scheme and functional program but it presented a rather irregular layout, probably as a result of adapting the structure to the sloped topography as well as the several reconstructions it had undergone. Traces of earlier constructions were determined especially in the courtyard, in room C12, and in room C18.
7.2.5.2 House B2A
House B2A was located on the north-eastern part of insula B2, on a higher level than the other houses. It covered a surface area of about 500 m2 (figs. 266, 267). The house was accessed from the inner alley 18 and was entered through rooms A9 and A10 that were similar in size and shape. Room 9 gave way to room A10 through a fairly narrow door and room A10 led to the courtyard. Along the west wall of room A10 ran a sort of bench stood in an oblique position with respect to the wall, probably indicating a reuse of an earlier structure.
The central courtyard of House B2A was roughly quadrangular in shape and had a colonnade of four columns, of which only three had remained. As recorded by Allara, this courtyard was identified as “the atrium of the frequent type in Pompeii” by Mesnil du Buisson who also defined room A4 as tablinum in the unpublished excavation reports (2002: 109). Considering the location of this space within the plan layout, its
254
form and way of communication with the surrounding spaces, it is difficult to resemble court A1 to an atrium or tablinum found in the Pompeian houses. In the center of the courtyard was a cesspool/cistern with a rectangular roof-like top that was different than the usual covers which had a hole in the middle. In the south-west corner of the courtyard was a staircase leading to an upper level, probably to the roof.
The largest room of house B2A was room A12 that was located on the northern side of the courtyard and faced, unusually, the south. It was entered from a wide door that was not situated with reference to the span of the columns of the portico. There were two niches on both sides of the entrance door and traces of plaster were observed on its south wall. Room A12 was annexed by rooms A13 and A18 on both sides.
The eastern and western sides of the courtyard were occupied by rooms A2 and A4 that were similar in size. Room A4 to the west communicated with room A12 through room A18. Similarly, room A2 once communicated with room A12 through room A13 but the door in between the two rooms was later blocked. In its latest phase, room 13 was directly entered from the courtyard through an arched opening without a door.
On the southern side of courtyard A1 was room 23 that had two distinct parts marked with different floor levels. The eastern half (A23a) was on the same level as the courtyard and opened to it via a portico with only one column that was smaller than those of the colonnade. In the latest phase of occupation, the eastern half of that portico was walled up. The western half of the room (A23b) was on a lower level and corresponded roughly to the floor level of House B. A door on the south side of court A1, in front of the stairs seems to have provided communication with House A and House B. It gave way to a totally disappeared room that could be the upper level or the roof of room B12, or the courtyard of House B that was about 2.5 m. lower than court A1. The existence of a staircase in that part of the house was attested by the traces of beams on the wall.
255
7.2.6 The House of Lysias (Insula D1)
The House of Lysias is one of the largest and relatively well-preserved residential buildings unearthed so far in Dura-Europos. While it was almost entirely cleared during the eighth (1934-1935) and ninth seasons (1935-1936) of the Yale mission the excavation reports have remained unpublished (fig. 268). In 2006, during the MFSED mission, the building became the subject of a doctoral thesis written by Ségoléne de Pontbriand, who restudied the remains in the site and the unpublished documentation of the earlier expedition that are kept in the Yale archive (Pontbriand, 2012, 2016).
The House of Lysias was situated in the south-eastern part of the city, in close proximity to the Redoubt Palace/the Strategeion and the Temple of Zeus Megistos. The name of the house was given after the discovery of several graffiti on the walls of the building that recorded the name of Lysias that belonged to one of the richest families of Dura-Europos and held the hereditary title of strategos kai epistates, ‘general and overseer’ of the city (Rostovtzeff, 1938: 48). One of the graffiti scratched on the jambs of a window mentioned epistates Lysias who had died during an expedition in 159 AD and was replaced by his son Lysanias. The House of Lysias, accordingly, has been interpreted as the palatial residence of the Parthian governor of the city, while the Strategeion or the Redoubt Palace as his official building where he exercised his power (Pontbriand, 2012: 90).
Several phases of reconstruction was observed in the house but no chronology was provided by the excavators. Pontbriand suggested the end of the Greek or the beginning of the Parthian period for its construction (2012: 91). According to her the house took its final layout during the Parthian reign and was not modified much in the course of the Roman period.
In its latest phase of occupation, the House of Lysias occupied the entire insula D1 covering an area of about 2,400 m2. Insula D1 was bordered by streets on its north, east, and west sides. It is seen that the street (Street 7) that should have theoretically
256
passed along the south façade of insula D1 did not continue after the western border of neighboring insula D3. The area beyond the south façade of the house was not excavated but it is likely that there was also a road along that border that separated insulae D1 and D2.
The main access to the house was from the northern street (fig. 269). The entrance door placed on the eastern side of the north façade, faced the Temple of Zeus Megistos, and was accentuated with a column on each side. The house had a second access from the western street, through a door in room 49. Two earlier doors that opened to the western and eastern streets from rooms 4a and 44 respectively, were blocked in a later phase(s) of occupation.
The House of Lysias was composed of several separate and communicating units each of which was organized around an open courtyard and had different functions. Pontbriand distinguished five main sections within the house, based on the organization of spaces and circulation patterns (figs. 270, 271).
The most monumental part of the house had large spaces organized around the main court 1. The main entrance gave way to that court through two entrance halls, room 13a and room 13b that were placed perpendicular to each other to form an L-shaped path. Two benches were placed in room 13a, presumably for the use of waiting guests. Rooms 20 and 17 that opened to entrance halls 13a and 13b respectively seem to have been the guardrooms from where the entrance and the visitor traffic were controlled. Court 1 was the largest space of the house, measuring about 16.00 m x 13.40 m. It was paved with fired bricks and surrounded by benches on the north, east and south sides (figs. 272, 273, 274). To the south of the court, a monumental entrance with triple doors and several steps led to two spacious rooms (2 and 3) that obviously formed the main reception suite of the house. Reception room 2 was fairly large, measuring about 7.20 m x 12.20 m for which room 3 served as an ante-room. Rooms 4a (4)-4b (41) and 36 that communicated with rooms 2 and 3, and also perhaps with the relatively large rooms 34 and 25 that opened directly to court 1, were the annex rooms that served the
257
reception suite. The blocked door in room 4a attested that this room once served as an entrance hall that provided direct access to the reception room from the western street.
Reception room 2 opened to a unit of seven spaces to its south through two doors pierced on its back wall (sector A3 in Pontbriand’s plan). The central room in that group, room 14, was identified as an open-air courtyard based on the existence of a cistern/cesspool. Similarly, the arrangement for a water feature in room 5 led it to be identified as a latrine or kitchen (Pontbriand, 2012: 87-88). The function of the other rooms could not be determined but the location of this unit in between the reception spaces and the presence of a second entrance made it likely that this part of the house was the service quarter, which could well have served both the public section to the north and the private wing to the south.
The northern side of court 1 was occupied by a distinctive rectangular unit, in which seven spaces were arranged in two rows (sector A4 in Pontbriand’s plan). The communication between this unit and the courtyard was provided through a single door built in the middle of the north wall of the court. The two rooms in that area, room 8 and room 11, had a linear masonry structure with a hollowed top, and which resembled feeders or watering installations for animals (fig. 275). Accordingly, these rooms and their annexes were identified as stable, used to accommodate transportation animals, like horses and donkeys (Rostovtzeff, 1938: 48; 1952; 73-74; Pontbriand, 2012: 88). The access route to the stables according to the published scheme, nevertheless, is rather an unusual one and has not been discussed so far in the published documentation. In the currently published plans, the only access to the stables was from court 1, as mentioned above. According to this plan, the animals should have entered the house from the main entrance and should have passed through the two entrance halls and the main courtyard to reach the stables, a highly odd and unexpected situation. Pontbriand mentioned about the traces of significant architectural modifications especially in that part of the house, which may indicate that the approach to the stables was not like this in the original plan. In the current state of our
258
knowledge, thus, it is not possible to identify the chronology and the functioning of the stables properly.
On the east and west sides of court 1 were placed two stairwells, E1 and E2 that enclosed fairly large staircases with three preserved flights of steps. These staircases were differentiated from the more modest ones in the secondary courtyards that probably led to the terrace roofs. The size and layout of staircases in court 1 suggested the existence of an upper level, with distinguished spaces, above the public part of the House of Lysias. The column drums found fallen in the courtyard supported this hypothesis. The excavators of the Yale mission suggested the existence of a fairly large covered space above rooms 8a and 8b that opened to a terrace through a portico (Pontbriand, 2012: 90). Similarly, Pontbriand suggested an upper gallery above room 3, based on the considerable thickness of its walls and the column drums found fallen along the southern part of court 1 (figs. 276, 277).
To the east of the public sector of the House of Lysias, was a separate and isolated unit of seven rooms (Sector B in Pontbriand’s plan). This unit had access from the south-east corner of court 1 through room 18 and was organized around the open and brick-paved court 24. A relatively large room (room 23) occupied the entire south side of court 24 and opened to it through a wide door of 1.70 m. On the opposite side, rooms 22 and 26 communicated directly with the courtyard and room 26 gave way to two successive rooms on its north (rooms 7 and 25). A single flight staircase that rested against the eastern wall of the court probably led to the roof. This unit that was isolated from but communicated with the monumental public section of the house was probably the private wing of the house that was reserved for the family of the owner and/or for the reception of more intimate guests.
A blocked door on the back wall of room 23, once gave way to a different unit of the house, seems to have been reserved for service spaces, especially for production and storage (Sector C in Pontbriand’s plan). In the latest phase of occupation, this unit was accessed only through a door on the east wall of room 3. This door opened to the long
259
corridor 21 that led to the service court 30. To the south of the courtyard were placed three interconnected rooms with a similar form and dimension. In each of them, there was a longitudinal thin wall that presumably served to support a shelving system to store jars. In the small room 29, on the south-west corner of the service court, was placed a basin-like feature which was supposed to be used for oil or wine production (Pontbriand, 2016: 121).
The southern section of the House of Lysias seems to have been originally occupied by two individual, adjoined dwellings (Sectors D and E in Pontbriand’s plan). They were attached to Lysias’ residence in the latest phase of occupation of the insula (Pontbriand, 2016: 121). These earlier dwellings had separate entrances, from the western and the eastern streets respectively, the latter of which was later blocked. Both were organized around open courtyards with main rooms (52 and 45b) located on the south side following the common plan layout of Dura houses. The courtyards were equipped with simple staircases and cisterns/cesspools dug into the ground. Court 39 was paved with baked brick like courts 1 and 24, while court 42 had an earth floor. The only communication with this part and the northern units of the house was provided through a door on the northern wall of court 39 that opened to room 28. Room 38, situated in the north-west corner of court 39 was obviously attached to the southern unit at a later time, as attested by its placement to the north of the east-west wall that separated the southern units from the northern one and by the blocked door on its north. Room 46 of the eastern unit of that part of the house was identified as a bathroom and room 40 was considered to be a worship area, based on the existence of four altars (Pontbriand, 2016: 121).85 Some installations related to water were found in rooms 47 and 55 which led to their identification as possible latrines/bathrooms or kitchens (Pontbriand, 2016: 121).
85 Pontbriand mentioned the existence of a private shrine in court 1 too, but she neither identified its location nor described it (2016: 121).
260
7.2.7 The House of the Large Atrium/The House of the Cistern86
The House of the Large Atrium or the House of the Cistern was located in the south-eastern region of the city, in the close vicinity of the Redoubt Palace/the Strategeion, the Temple of Zeus Megistos, the Temple of Artemis and the Temple of Atargatis. It was built in insula D5 that was the second insula to the east of the House of Lysias. This area was excavated in 1930-1931 by Pillet, with the hope of finding a circular edifice since there was a large circular depression in the middle of the insula (Pillet, 1933: 27; Hopkins, 1979: 174). The excavations, nevertheless, put forth that it was a deep cesspool/cistern cut into the rock in the middle of a large courtyard that caused the depression. The whole insula was cleared in the same season and a distinctive and wealthy habitation was revealed. A very brief description of the remains in the insula was given in the excavation reports (Pillet, 1933: 27-32). In 2006 during the expedition of MFSED, the insula was cleared once again and restudied by Baird (2006: 170-225).
The chronology of insula D5 could not be defined with the current state of information but some suggestions could be made. The existence of earlier wall foundations below the floor level attested that the House of the Large Atrium was built over some earlier structures. The reuse of Hellenistic blocks in the walls of the house may suggest the Parthian period for its construction. The coins found in and around the block indicated that the occupation in this region continued until the final days of the city (Bellinger, 1933: 261-262).
Insula D5 was occupied by a number of distinct units that communicated with each other at least in some occupation phases (fig. 278). Changes in the circulation patterns within the insula were attested by many blocked doors and new openings between the units but the sequence of these changes could not be ascertained in the absence of stratigraphic studies. According to Hopkins insula D5 was originally a standard plot
86 In this study the name the House of the Large Atrium is used since it has been more widely used in the publications.
261
with eight houses and the House of the Large Atrium was one of these, which was enlarged and integrated the other units (1979: 69-70). This hypothesis can be supported by the existence of different entrance doors for each unit, though some of them were later blocked. Pillet and Baird, nevertheless, argued that the insula was initially occupied by a single large house that was divided into several units in the course of its habitation (Pillet, 1933: 28; Baird, 2006: 174). What can be suggested with certainty is that the insula was occupied by four distinct units in its latest phase of occupation (fig. 279).
In Pillet’s published plan insula D5 was separated into a series of houses that were demarcated by letters given to the courtyards. Pearson, who restudied the insula in the fifth season or after, used a different numbering scheme in his unpublished plan. In this study the plan of Baird who revised Pearson’s plan, is used since it is more legible, more detailed and the numbering system is more consistent with the one used for the other insulae of the city. In this plan seven units were designated as A, C, D, E, F, F1 and G. Unit F was the largest and more elaborate unit of the insula. It had a large colonnaded courtyard that was initially labeled as ‘atrium’ by Pillet and that led the building to be named as the House of the Large Atrium (1933: 27).87 This name, on the other hand, referred to unit F in some studies whereas it defined the whole insula in others. To prevent this confusion only the letter codes will be used in the following part of this section.
Unit D5F was accessed from the street to the east, through a long corridor or more probably an inner alley, placed at the center of the insula (F19). The alley led to entrance hall 18 that opened to the eastern end of the north portico of the courtyard. Two blocked doors on the western façade of the house, one opening to room F10 built
87 In the preliminary report of the fourth season of the excavations, in the sub-section reserved for the House of the Large Atrium Pillet describes the houses of Dura as “… present a uniform Greek plan: a square or rectangular atrium which ventilates and surrounded by rooms which have no opening on the Street.” (1933: 28). In this description he used the term atrium for the central open spaces while in his descriptions of the other units in insula D5, except unit D with the colonnaded courtyard, and of the other houses in the same volume he used the term court and not the atrium.
262
underneath the stairs and the other to room F13, provided entrance to the house in the earlier phases.
The largest space of unit F, as well as of the whole insula, was courtyard F1 that measured 13.25 m x 15.45 m. The courtyard had colonnades on its north, south, and west sides and masonry benches that rested against them. The four columns on the north and south sides were 0.74 m in diameter and they were not placed symmetrically. The dimensions of the columns on the west side were not identified but they seem to be smaller in diameter on Baird’s plan. There was a basin and a cesspool/cistern cut into the rock below the courtyard that caused the depression in the area. By the final period, the columns were engaged into a wall that surrounded the gallery on four sides. The connection between the eastern and the southern portico was also cut by a wall. A spacious staircase with three flights of 1.35 m wide on the north-western corner of the court gave way to an upper level (fig. 280).
To the south of the courtyard was placed room F4, the principal room of the house that came into prominence with its large size, monumental entrance, and architectural layout (fig. 281). The room measured about 11.00 m x 6.70 m and opened to the courtyard through a triple entrance, with a wide axially placed door at the center and two narrower ones on both sides. The axis of the central door more or less coincided with that of the central intercolumnation of the portico but this does not seem to be intentional since the distance between the columns was less than the width of the door. The thickness of the northern facade of the room was more than twice the thickness of the other walls that gave the entrance a more monumental character. On the back wall of the room were built two tall niches with shelves (fig. 282). Room F4 was annexed with a spacious room to its west (room F6) that also communicated with a suite of three rooms on the west side of the courtyard (rooms, F7, F8, and F9). According to Baur, a number of painted terracotta tiles that were found during the excavations came from “the main room” or room F4, whereas according to Pillet they came from “a badly ruined room on the south-west side” of the house (Baur, 1933: 42; Pillet, 1933: 31). The tiles measured 0.41 m x 0.41 m x 0.05 m and were decorated with male and
263
female faces, fruits including pomegranates and apples, animals including fishes and goats and with geometric patterns (fig. 283). Baur and Pillet suggested that the tiles were used on the walls whereas Baird argued that they should have come from the ceiling as was the case in some other buildings in the site (2006: 172). Fragments of a cast plaster frieze found fallen in the building could also have belonged to room F4 (fig. 284).
In an earlier phase of occupation, reception room F4 was accessed from the southern street too, through the large entrance hall G2 and smaller anteroom F5. A second connection to these two units was given from a door placed between annex room F6 and room G3. By the final period the doors between rooms G2-F5 and G3-F6, as well as the two doors between court G1 and room G4 were blocked and a small, separate housing unit, unit G, was created in this part of the insula.
The south-east side of insula D5 was occupied by units C and D that formed a single dwelling in the latest phase of occupation. Unit D was noteworthy with its colonnaded courtyard and the three spacious rooms located around it. It had access from the eastern street via entrance hall D12. The entrance hall gave way directly to the northern portico of the central court. Courtyard D1 had a colonnade on its two sides with four columns in total (fig. 285). The columns were thinner than the ones in court F1, with a diameter of 0.58 m. Before the blocking of the door on the west wall of room D6, unit D had communicated directly with reception room F4, and then it probably served the public reception section of the complex.
Unit C, too, communicated with unit F before the door in C5 was blocked. By the latest phase of occupation, the doors between rooms C5-D6 and C3-D11 were also blocked and the passage between the two units was provided only through a narrow passage between C5 and D7. The entrance scheme of this unit was similar to that of unit D, through an entrance hall (room C8) that gave direct access to the courtyard. A large room (D4) opened to the south side of the courtyard via a wide door. Beneath this door was a cellar. Rooms C2 and C3 to the east of the court which once opened to the street
264
were identified as the kitchen by Pillet, based on the existence of a masonry bench and jars found in-situ (1933: 29). A spacious staircase like the one in court F1 was installed on the western side of court C1. Pillet mentioned the existence of a latrine with an in-situ stone seat in relation to the staircase (1933: 30). Though its placement was not clear in his report it was most probably located below the upper flights.
The northern part of insula D5 seems to have been reserved for the family. The layout of spaces and the finds like jewelry, funeral busts, and vases suggested a private use for this section, as opposed to the southern part of the insula which had a more public character. Three units, A, E, and F1, occupied this section. By the latest period before the destruction of the city, units A and E communicated with each other, and unit F1 was linked to unit F.
Unit F1 was an independent house with a separate entrance from the street, before the door in room F110 was blocked. The passage between units F and F1 was provided from room F17 on the north side of the colonnaded courtyard F1. Unit F1 was organized around a courtyard like the other units, but it presented some uncommon architectural features. Room F111 was built within the courtyard without resting against any pre-existing wall, probably in a later period. The southern side of the unit was occupied by the most spacious room of the house (room F15) with annex rooms on both sides (rooms F14, F16, and F17), neither of which had any direct communication with the courtyard. According to Pillet this large room with a column base on its north-west corner was an open court to which the kitchen (room F14) was annexed. Remains of earlier walls in the kitchen were reused as benches. Room F13 to the east of the unit that opened to the street was identified as a shop (Baird, 2006: 218).
Unit A presented the common plan layout of Dura houses with its central courtyard with a simple staircase, the principal room with low benches, and its annex rooms located to the south of the courtyard. The entrance to the unit was provided from the northern street through entrance hall A7 and communicated with unit E via a door on the west wall of room A6.
265
Unit D5E originally had four doors, three opening to the northern street and one to the western street. By its latest phase of occupation, two of these doors were blocked and the unit was entered only from the northern street through rooms E18 and E16. According to Baird, these two rooms as well as rooms E13, E14, E15, and E17 on the northern façade of the insula were a range of shops at one time (2006: 200). House D5E had a distinct layout with two adjacent courtyards, E1 and E7, the latter of which was paved with stone. On the north-east corner of court E1 was a staircase under which two small ovens were installed. Room E5 opened to the courtyard through a wide opening without a door. This room once provided passage to unit F by a central door on its south wall.
A large number of coolers in insula D5 were found during the excavations. They were mostly found in-situ, especially in the entrance halls (A7, E16, E11, F18, F110, G2) and the courtyards (C1, D1, E1, E7, F11, G1). Among the archaeological finds discovered during the excavations were a small hoard of Roman coins, most of which were of the 2nd and 3rd centuries, a small relief of Heracles, a broken statuette of Venus, and a number of terracotta lamps (Pillet, 1933: 30-32).
7.2.8 House E4/The House of the Parthian
House E4 or the House of the Parthian, built in insula E4, was one of the largest single houses excavated so far in Dura-Europos. It was unearthed by Frank Brown in 1932-1933, during the sixth season of the French Academy-Yale University expedition. The house was named as such since its first phase was dated to the Parthian period, to the third quarter of the 1st c. AD (Brown, 1936: 31-32). Remains of earlier walls below the floor level of the house indicated that it was built over an earlier structure that was razed to the foundation. After 210 AD, when the northern section of the city was turned into a Roman military camp, the house remained within the military quarter in a significant position between the Roman bath and the theatre. Subsequently, House E4 was remodeled to provide accommodation for the Roman soldiers (fig. 286). Archaeological evidence attested to two main phases for the military occupation. In
266
the first phase, the house seems to have served as the residence of a commander or a high official of the army and in the second it accommodated a large number of soldiers.
The house occupied the southern half of insula E4, covering an area of about 1,330 m2. The northern half of the insula was not excavated. According to Baird, this area might have been occupied by an open space or palaestra which would explain the entrance to the bath to the north, in insula E3, and the opening on the north wall of room 13 of House E4 (2006: 231). The presence of the stables on the northern façade of the house could support this idea.
The house was square in plan, measuring 36.50 m on each side. It originally had four entrances: one from the west through entrance hall 21, one from the east through entrance hall 26, and two from the north through entrance halls 7 and 24. In a later phase(s) the doors on the western and northern façades were blocked. Instead, a new door on the west façade was opened and room 5 was turned into an entrance hall. This door was accentuated by projecting walls on both sides and probably also became the main entrance of the house. In addition, a new wide doorway that was opened on the north wall of room 13 gave way to an open space.
House E4 was organized around two courtyards (14 and 22) each of which defined a separate sector probably with different functions. The larger courtyard 14 and the surrounding rooms seem to have constituted the public part of the house with reception spaces (fig. 287). The rooms in that part were arranged symmetrically on three sides of the court. To the south was located the largest room of the house, room 3. Its entrance, placed at the axis of the court was accentuated with a wide door preceded by two steps and two columns on both sides. Unlike the main rooms in the other large houses brought to light so far in Dura-Europos room 3 was not provided with masonry benches. Room 3 was annexed by two groups of interconnected rooms (rooms 1, 2, 4 and rooms 16, 25, 29). The one that contained rooms 1, 2, and 4 had some importance as suggested by their large size and the molded plaster cornices found on their walls. Room 4 had also direct access from the western street through entrance hall 21, before
267
the doors were blocked. In rooms 16 and 25, as well as in room 34, parts of earlier walls were left in-situ and were used as benches, as was the case in the above-mentioned room F14 of the House of the Large Atrium.
To the west of court 14 was large room 15 that obviously served as a kitchen. A series of ovens and three large dolia were found in-situ in this room. A fourth dolium was reused as a basin and set into the floor next to the door. A drain connected the basin to the cesspool/cistern in the middle of the court. In a later period, a large cooking area was placed to the north-west corner of the kitchen indicating the need to feed more people in the final period of the house. Room 11 presumably served the kitchen since the passage in between the two lacked a door and a graffito in room 11 depicted food purchasing.
On the east side of the courtyard was room 23 that was distinct with its layout and decoration. There were several niches on its east and west walls. It received lavish wall paintings most probably after the ownership of the house passed to the army as understood by the surviving fragments depicting a Roman officer making a sacrifice and an elaborate calendar. In the final phase of the building, the paintings were covered with a coat of plaster and the room was converted into a storeroom as attested by the fifty-two amphorae and four large dolia found in-situ.
A fairly spacious staircase next to room 23 on the east side of court 14 gave way to the upper level. House E4 is the only domestic building unearthed so far in Dura-Europos, in which the existence of an upper floor was attested by archaeological evidence. Fragments of mosaic and painted plaster were found fallen above the floor level of room 24 indicating a collapsed upper story. House E4 is also a unique example in which the use of mosaic floor pavements in the domestic context is attested in Dura-Europos. Besides the fragments of mosaics and painted plasters, a large quantity of window glass found in court 22 were thought to have come from the upper rooms that overlooked the court. The buttresses on the northern side of the house which were not the original members of the house but were added later to support the northern façade
268
wall led the excavators to think that the upper story could have been a Roman addition to the house (Baird, 2006: 230).
The public part of house E4 that consisted of court 14 and the surrounding rooms, was aligned by a second row of spaces on its north and east sides. This outer row seems to have comprised the domestic and the service spaces of the dwelling. Originally the eastern wing of the house was more isolated and was accessed from the eastern street through entrance hall 26 and was organized around the second court 22. In a later phase, presumably after the military occupation, the circulation pattern between the two parts of the house was redefined.
Court 22 measuring about 48 m2 was much smaller in size than the main court. On its east side was placed a staircase with three flights beneath which was installed a stove indicating that cooking activity took place in this part of the house too. There was also a cesspool/cistern dug into the ground. Originally, the two spacious rooms, 30 and 36, opened to court 22 but they were isolated from the court in the final period. Room 30 to the north was integrated into a unit of other rooms on the northern wing of the house while room 36 was separated from the court by blocking the door in between.
Room 36 and the four interconnected rooms to its south, rooms 33, 34, 38, and 39, seem to have constituted a separate unit within the house during the military occupation. This unit was accessed from the main court or from room 29 through a small, corridor-like room (32). Room 33 which opened to room 25 previously, was enclosed and turned into a bath with the installation of a hypocaust system. Two coats of plaster were found on the walls of the bath, the second of which had a painting imitating marble. In the latest phase of occupation, the bath went out of use as did the latrine in room 28. In room 38, along the western wall a series of rubble and plaster constructions forming large troughs or containers were installed. The function of these features could not be ascertained but an industrial installation like an olive press was proposed by Baird (2014: 134). Like the bathroom, this installation had been plastered over and was no longer in use in the final phase of the house. In the same phase, room
269
36 was being used as a storage room as indicated by the five amphorae and a pithos found in-situ.
The northern wing of House E4 was occupied by a row of rooms that were separated from the main court by a long corridor (room 20). On the south wall of the corridor, a text honoring the commanding officer of a soldier was scratched. Rooms 13 and 30 in this part were paved with stone flagging in the final period that was contemporary with the installation of a trough on the north side of room 30. Accordingly, these two rooms and room 24 in between seem to have been used as stables in the latest phase of occupation. A fairly wide door on the north wall of room 13, opening to an open space and a graffito of two horses scratched on the door between rooms 30 and 31 supported this hypothesis.
House E4 yielded a considerable number of finds during the excavations. These included several military equipment like scales of armor, a copper alloy plate from a dagger, and three missiles. Among the finds were also a die and knuckle bones, hooks, needles, bracelets, and finger rings.
7.2.9 The Christian House/ The House of the Christian Church
The so-called Christian House or the House of the Christian Church was discovered by Pillet in 1931 and excavated in the following seasons by the Yale University-French Academy mission. The building was converted into a gathering place for Christians in its latest phase of occupation. Since it was one of the most ancient Christian churches excavated so far in the Roman world it is one of the best-studied and published buildings in Dura-Europos. It was described in detail in the preliminary excavation reports and in the final report 8.2 that was dedicated to this single building (Hopkins, 1934: 238-25; Kraeling, 1967).
The Christian House was located at the western edge of the city, in the block to the south of the Palmyra Gate, behind tower 17. Its location near the defensive wall
270
provided it a certain degree of protection and its western walls were standing almost at their full height when it was excavated. The house was built on the north-western corner of insula M8 (fig. 288). The form and size of this insula differed from the regular ones in the grid plan of the city. It was trapezoidal in form, following the alignment of the city wall just like the other insulae situated along the wall. There were some empty plots in the insula which were thought to have been used by soldiers who were responsible to control, secure and/or maintain the fortifications during the Roman period (Baird, 2006: 492; 2014: 144).
Excavations revealed the remains of an earlier building under the house with a totally different orientation. This earlier structure probably belonged to a period prior to the extension of the orthogonal plan towards this part of the city in the Roman period (Kraeling, 1967: 32-34). The construction date of the Christian House over the earlier building came from a graffiti on the west wall of the room which indicated the year 545 of the Seleucid era, i.e. 232-3 AD. What this date stood for was not clear, but, according to Kraeling and Lassus, it most probably pointed out the date of construction of the house (Kraeling, 1967: 34-39, Lassus, 1969, 131-132). They argued that the transformation of the house into a church was likely to have been around 240 AD.
The Christian House was similar to many other dwellings in Dura-Europos with its relatively modest size, plan layout and construction techniques (fig. 289). It measured about 340 m2 and was organized around a central courtyard. The access to the house was from the northern street. The entrance door on the eastern edge of the north facade gave way to a fairly large entrance hall, room 8, measuring about 5.70 m by 2.50 m. The entrance hall opened to the courtyard through a 1.75 m wide opening that was surmounted by an arch and via two steps.
The central courtyard measured about 62 m2. Its floor was of beaten earth with plaster and its level was slightly lower than the surrounding rooms which were accessed via one or two steps from the courtyard. The eastern side of the courtyard was separated by two columns on a stylobate behind which was created a portico measuring about
271
9.00 m x 2.70 m. In the north-west corner of the court, a bottle shaped cesspool/cistern of 3.20 m depth was carved into the ground. The walls of the cesspool/cistern were plastered whereas the bottom was not. Its mouth was covered with a square gypsum slab of 0.85 m with a small circular aperture at the center.
The southern side of the courtyard constituted the most important part of the house, as it included the principal room and its annexes. A monumental door of about 1.60 m wide and 3.00 m high, with a stone molded frame, at the center of the south wall of the courtyard opened to principal room 4B. The floor level of that room was about 0.40 m higher than that of the courtyard and a pair of 0.50 m wide steps were built in front of its entrance. There was a niche of 0.30 m wide and an indeterminate height, at the lintel level just to the right of the entrance of the main room on the south wall of the courtyard. Considering its out of reach location from the door sill Kraeling suggested that the niche may have been used to place an object that had a prophylactic function or was connected to a domestic cult (1967: 12). Room 4B was the largest room of the house measuring 8.00 m x 5.15 m with a height of about 5.07 m. Its floor was paved with plaster mixed with pebbles. A plaster-covered rubble bench of about 1.00 m wide and 0.17 m high ran all around the room. Just to the west of the door of the room, encroaching upon the door sill, plaster ledges outlined a shallow hole of 0.95 m x 0.70 m which apparently served for the placement of braziers to heat the room in winters (Kraeling, 1967: 15). The walls of the room were plastered and were surrounded by a plaster frieze found about 2.00 m above the benches. The 0.17 m high frieze was decorated with rows of Satyr masks, symbols of Pans pipes, dolphins, and seashells. A pair of graffiti depicting a mounted archer and mounted lancer was scratched on the south wall (fig. 290).
The main room 4B communicated with two annex rooms to its east and west. Room 3 on its east was a relatively small room measuring 5.15 m x 2.55 m. Its floor level was about 0.10 m below the floor level of the main room and had a beaten earth floor.
272
Room 4A to the west of the main room measured 5.15 m x 4.25 m. Its floor level was 0.15 m lower than that of the main room and its height was 5.22 m, corresponding to the roof level of room 4B. The floor was of beaten earth and the walls were plastered. There was an alphabetic graffito on its south wall and the above-mentioned graffito with the date of 232-233 AD on its west wall. Along the west and south walls were a row of equally distanced peg holes at a height of 1.67 m from the floor level. There was a 1.15 m wide and 1.75 m high niche built 0.45 m above the floor on the north wall. According to Kraeling, the niche and the holes on the walls suggested the use of that room for storage, especially to store for floor covering materials like rugs, mats, pelts, and bolsters or more precious items like articles of clothing and containers of various sorts (1967: 14). Besides its storage function, room 4A should have served for other purposes, too, since it was lighted up by two windows placed on the west wall, at a height of 2.92 m from the floor level (fig. 291). The height of the windows was about 1.10 m and they were tapered towards the top, having a width of 0.65 m at the bottom and of 0.35 m at the top. Room 4A communicated with room 5 too, via a door of 1.50 m wide and 2.00 m height that was fully preserved.
The rooms on the western side of the courtyard had a similar arrangement with those on the southern side. A large central room was annexed by two smaller rooms on both sides. This kind of arrangement provided circulation between all rooms of the house, but the entrance hall, without using the courtyard. Room 5, at the center, was the second largest room of the house measuring about 4.20 m x 7.40 m. Its height was the same as rooms 4A and 4B, exceeding 5.00 m. The entrance to the room from the courtyard was accentuated, just like that of room 4B, with a 1.50 m wide door, molded stone frames and a 2.50 m long, 0.45 m wide and 0.23 m high step. Its floor was of beaten earth and the walls were not plastered. No windows were found on the west wall facing the rampart street but it is likely that there were windows on the destroyed upper portion of the east wall towards the courtyard. Room 5 communicated with room 4A to the south and with room 6 to the north via two doors. Room 6 was probably a service room as suggested by its irregular shape, poor ventilation and lighting, accessibility from the courtyard, and the placement of the cesspool/cistern just in front
273
of its door. It had two doors, one providing access from room 5 and the other from the courtyard. The room was provided with a 0.62 m wide and 0.26 m high bench on its east wall, a 0.88 m wide and 0.60 m high niche on the south wall near the entrance and an undetermined feature projecting from the wall, just below the niche which could be a step or a table.
To the north of the courtyard between room 6 and the entrance hall a stairwell enclosing a staircase with three flights was installed. No archaeological evidence suggested the existence of an upper floor in the Christian house so the stairs probably led to the terrace roof. The small space under the flights was entered by an arched passage from the courtyard and probably used as a service room. There was also a cellar under the entrance hall which was found filled with earth and rubble since the floor of the entrance hall was demolished. The height of the cellar was about 1.50 m but its entrance could not be spotted. In the plan of the building, an entrance through a few steps from the portico was suggested.
Around 240 AD the domestic building witnessed a change in function. It was transformed into a religious building with some architectural changes (figs. 292, 293). The partition wall between rooms 4A and 4B was removed and a large hall was created to serve as a meeting place. Room 6 was turned into a baptistery by the addition of a basin that was surmounted by a vaulted baldachin. The baptistery received wall paintings and was provided with an upper room. The courtyard was paved with tiles that covered the cesspool/cistern and put it out of use. Benches were built around the courtyard.
7.2.10 The Roman Palace/The Palace of Dux Ripae
In the north-eastern edge of the settlement, within the military section of Dura-Europos, a large palatial building was situated. It was excavated by Mesnil du Buisson and Comstock during the 1935-1936 season of the joint mission of Yale University and French Academy. In the following years, studies on the textual and archaeological
274
finds and the preparation of the architectural drawings of the building continued under the direction of the field director Frank Brown and M. I. Rostovtzeff. Due to the outbreak of the war in the 1940s, the results of the excavations could only be published in 1952 (Rostovtzeff et al., 1952). In 1988, further work on the building and its territory was conducted by Susan Downey under the auspices of MFSED (Downey, 1993).
The Roman Palace is the only residential building brought to light so far in the ancient city of Dura-Europos that was not adopted from a pre-existing structure but a totally new construction of the Roman period. A dipinto painted on the wall plaster of one of the rooms containing the name and filiation of Elagabalus, who ruled between 218 and 222 AD, gave a terminus ante quem for its construction. Fragments of several other dipinti found fallen in the rooms of the building referred to the name of Domitius Pompeianus and one mentioned him as the commander of the river bank (Rostovtzeff and Welles, 1952: 27-35). Consequently, Rostovtzeff identified this person as the Dux Ripae, the commander of the garrisons of the Middle Euphrates in the first half of the 3rd c. AD and defined the building as “the Palace of Dux Ripae” (1952: 93, 95-96). Nevertheless, the date of the formation of the office of Dux Ripae has been called into question in later studies and the building began to be called the Roman Palace (Edwell, 2008: 128-135; Baird, 2014: 148-151).
The Roman Palace was located on the edge of the cliff over the Euphrates, within the limits of the military camp to the north of the city. It was not fitted to the existing urban grid but was oriented towards the river, taking its reference from the city wall to which it rested against. The foundations of the earlier walls below and around the building indicated that the orientation of the earlier buildings in the area which were totally destroyed in order to establish the palace, followed the grid.
The Roman Palace is the largest single building excavated so far in Dura-Europos with a total ground surface area of approximately 5,500 m2 (figs. 294, 295). It apparently included both public and residential functions that were organized around and separated by two large peristyle courtyards. Court 58, on the south-eastern side,
275
provided the main entrance to the palace and probably served for public functions. It was the largest space of the building, almost square in form with sides measuring nearly 31.00 m and was surrounded by a colonnade with ten columns on four sides. The main entrance to the building was designed like a deep porch that was axially placed on the south-western façade of court 58 and projected from it. The width of the entrance was 3.40 m between the two Doric half columns attached to the interior walls of the porch. On the north-west side of the court was a lateral entrance similar to but narrower than the main one. On the axis of this second entrance, behind the south-eastern portico was an exedra measuring 7.10 m x 4.44 m. The exedra was emphasized by the wider distance between the two central columns of the colonnade in front and the arch they carried. The central area of court 58, as well as that of court 1, was paved with natural gravel while the porticos of both courtyards were paved with solidly packed gravel bounded with plaster.
The passage between forecourt 58 and inner court 1 was provided through room 35 that was axially located with respect to the main gate. This passage room was equipped with a platform of 0.85 m wide and 0.21 m high on one side and with a bench of 1.40 m wide and 0.43 m high on the other side. Room 35 had two interconnected annex rooms, 33 and 34. To its south-east was room 36 that also opened to court 58 and was provided with benches on three sides. Two more benches were installed against the walls of the portico on both sides of room 35. Apparently, these rooms that were either interconnected with the passage space or in close proximity served as guardrooms used to control the passage and/or the waiting areas for the visitors.
Court 1 formed the core of the Roman Palace around which several suits of rooms, either with public or private functions were organized. It measured about 25.00 m x 23.00 m and was surrounded by a colonnade with eight columns on each side.
To the west of court 1 was a stable complex that provided entrance to the palace directly from the alley on the north-western side of court 58. This alley ended on the façade of the building, just to the north-west of the enclosed courtyard. A large door
276
of 2.22 m wide on the facade wall provided entrance, presumably for the vehicles, into a large space (room 30) that measured about 66 m2 and had intermediate supports to carry the roof. To the north-west it communicated with the projected room 57 that was apparently a stable that could accommodate five to six horses as indicated by the remains of the mangers (Detweiler, 1952: 13). To the south-east, room 30 opened to room 31 that served as a passage space between the stable and the inner court. The small room 32, too, had the remains of a structure similar to the mangers of stable 57 and it probably sheltered small animals.
A symmetrical unit of five rooms occupied the north-western side of court 1. At the center of this unit was the largest room (28) that measured about 7.50 m x 5.32 m. As attested by the thickness of its walls and the height of its foundations, room 28 was apparently higher than the surrounding rooms. It opened to the peristyle courtyard via a 1.50 m wide door and its entrance was emphasized by the wider intercolumniation of the colonnade in front of it, as well as a circular foundation just behind, which was presumably a statue base. The floor of the room was covered with plaster and it had a decorative ceiling with a richly painted false vault of plaster. Room 28 was annexed by two symmetrical units, rooms 27 and 29ˈ and their anterooms 26 and 29. These latter rooms provided passage between court 1, room 28 and the annex rooms. Remains of benches indicated that they were also used as waiting rooms. The location, size, layout and decoration of this unit of five interconnected rooms indicated that it served as a reception suite.
Next to the reception suite, three rooms formed a service unit that communicated with the bath building that adjoined the palace to its north-west. The graffiti scratched on the walls of rooms 24 and 25 of this unit recorded the accounts of barley and cheese distribution. Room 11 that provided communication with the palace and the bath was also used as a storeroom as the fourteen circular pits in one of which was preserved the remains of a set of jars indicated. Since the bath building was not totally unearthed there is not much information on its chronology and the connection with the palace.
277
On the south-eastern side of the peristyle courtyard, was situated a smaller court to which opened a row of rooms. Court 43 was a large rectangle measuring 19.00 m x 5.00 m and was accessed by two doors, one at the center and the other at the edge of the south-eastern portico. The packed gravel flooring of this portico, behind the three central columns, was turned into flagstone pavement which was also used for the flooring of court 43. A row of rooms (38, 41, 45-47) with similar size flanked court 43 on its south-east. This section of the house appeared to have been a service area, probably the slave’s quarter, as Rostovtzeff pointed out (1952: 76).
At the north-east end of court 43, a small passage gave way to a suite of rooms comprising a bath and a latrine. This suite comprised corridor 23, bathroom 49, and room 48 that was adjoined it in a later period, as well as larger rooms 15 and 16 to the north-east. Room 15 provided entrance to this suite from corridor 14, which led to the living quarters of the master of the house and his family.
The north-eastern part of the Roman Palace was differentiated and physically separated from the rest of the building. The group of rooms on this section of the palace turned their back to the peristyle courtyard and opened to a terrace built on the edge of the building overlooking the Euphrates. The isolation of this part of the palace and the elaborate decoration of the rooms suggested that it constituted the private living quarters.
The communication between the central peristyle courtyard and the private section of the Roman Palace was provided by two long narrow corridors (10 and 14) that were located on both ends of the north-eastern portico. The corridors gave way to an arcade opening to the terrace that laid along the whole façade of the palace. The arcade was about 35.00 m long and had a plastered floor. The piers were 1.30 m x 0.80 m and were placed at an interval of 2.45 m. The arcade opened to two symmetrically placed exedrae on both ends. Exedra 61 was highly decorated with painted walls and a false vaulted ceiling with geometric motifs and imitation of marble incrustation. Since a large part of terrace 63 had fallen down to the river its exact dimensions and layout
278
remains unclear. The width of the terrace was restored as 7.55 m by Detweiler whereas according to Downey it was probably much narrower and measured about 4.00 m (Detweiler, 1952: 18; Downey, 1993: 183). The terrace was paved with rammed gravel.
Corridors 10 and 14 divided the north-eastern part of the house into three sections, each with a group of several rooms. The symmetrically planned middle section came into prominence due to the large apsidal hall placed at the center (room 2). This hall was situated on the axis of court 1 though the two spaces did not have a direct visual or physical relation. Room 2 measured 5.92 m x 5.32 m with an apse of 4.44 m in diameter at its end (fig. 296). The thickness of its walls and the dimensions of its main door attested that it was higher than the other rooms in the area. It was entered from arcade 62 through a door of 2.30 m wide and its entrance was also emphasized by the wider distance between the two central columns of the arcade. The floor of the room was paved with plaster and the floor of the apse was 0.025 m higher than the rest. Fragments of the false vault that were deprived of any kind of decoration were found fallen down on the floor of the room. There were three niches in the walls of the apse, one located at the center of the back wall and two in sidewalls. Remains of food such as peach stones, eggshells, bones of birds, and fish were found in the room and attested that it served, at least occasionally, as a dining room.
Room 2 was flanked by two groups of rooms on both sides with which it communicated. To its north-west, rooms 5, 8, and 9 formed a suite of highly decorated rooms. Room 8 provided entrance from arcade 62 and served as a passageway between rooms 2, 5, and 9. The door between arcade 62 and room 9 was partially blocked and turned into a window in a later phase of occupation. All three rooms had plastered floors and painted false vault ceilings with geometric decoration. The vault of room 8 nevertheless had a simpler pattern than those of rooms 5 and 9.
The south-eastern side of room 2 was occupied by two different suits that were formed by rooms 6-7 and 3-4. The apsidal room communicated only with the former one.
279
Room 7, similar to room 6, opened to the arcade and gave way to room 6. The earlier door of room 6 was also turned into a window, most probably at the same time as that of room 9. The false vaulted ceiling of room 7 was decorated with a simpler pattern of squares while that of room 6 was more richly painted. On the south-east wall of room 7, a dipinto recorded the names of a group of actors in the service of Domitius Pompeianus. Accordingly, Rostovtzeff suggested that rooms 6 and 7 were used to accommodate these actors (1952: 79). Rooms 3 and 4 located behind rooms 6 and 7 were only reached from corridor 14. The fragments of the ceiling of room 3 had decorative paintings too (fig. 297).
On the north-west end of the terrace was a separate suite of two rooms. Room 12 was entered from arcade 62 and gave way to room 13 behind. Both rooms had decorated false vaulted ceilings and room 12 had decorated wall panels with geometric paintings and borders imitating marble.
The third group of rooms within the private section of the Roman palace was located on the south-eastern end of the terrace. Located at the end of the arcade, room 60 served as a vestibule providing access to this section through room 18. An earlier door that provided entrance from the arcade to room 17 was blocked and turned into a niche. Rooms 19 and 20 were the largest rooms of this unit. Room 20 had an apsidal alcove that was 2.60 m in diameter and 1.75 m deep, on its southern corner. All four rooms were paved with plaster and covered with painted false vaults. The decoration of the vault of room 18 was less ornate like rooms 7 and 8 which had similar functions. The walls of rooms 18, 19 and 20, as well as the alcove on the other hand were decorated with painted panels imitating marble incrustation.
Room 60 opened to an elongated space of at least 12.70 m long to its south-east, through a 1.34 m opening on its back wall. Since a large portion of this part of the building was fallen into the ravine, a wholly convincing restoration of this area could not have been possible. The plaster flooring and the painted walls, which were used in more important rooms of the palace, nevertheless, indicated that it was most probably
280
a covered room. At the end of the preserved part of this space on the south-western wall was an apse of 3.35 m in diameter (fig. 298). Within the apse remains of two superimposed platforms were found. Detweiler suggested a possible tri-apsidal hall at the back of room 64 in his reconstruction (fig. 299) Downey, nevertheless, assumed that room 64 was much longer and measured about 18.68 m with the apse as its central feature facing the river (fig. 300). She also pointed out the change in the construction technique of the north-eastern walls of rooms 17 and 18 as well as the apses 21 and 22, and thus argued that they were part of a later phase of construction (1993: 188).
The Roman Palace differed from the other houses of Dura-Europos in terms of its general plan layout and organization of rooms It comprised three separate parts: court 58 with two monumental gates and probably with a military function; court 1 and the surrounding rooms reserved for semi-public use, and the north-eastern section opening to a terrace with private units. Rostovtzeff interpreted the residential portion of the palace as the combination of a peristyle house and a porticus villa (1952: 81-99). Several features including the Roman foot as a unit of measurement, axially located apsidal halls with annex rooms, and false vaulted ceilings attest an affinity with Roman architectural traditions.
7.3 Overview of Domestic Architecture in Dura-Europos
The rich archaeological remains and finds in Dura-Europos have demonstrated that there was a conceptual and structural continuity in the domestic context of the city throughout its long period of occupation. It is difficult to discern with certainty the Hellenistic, Parthian, and Roman phases of dwellings from the material remains. The excavated sample of houses, most of which were originally built in the Parthian period, represents the latest phase of occupation in the city before it was abandoned in the Roman period. Accordingly, the Roman period houses differed in their architectural layout and decoration from the contemporary examples found in other Roman cities in the region. The textual evidence in this regard, can provide a better understanding and interpretation of the archaeological remains brought to light at Dura-Europos. Among
281
numerous parchments and papyri recovered in the ancient city, a parchment coded ‘PDura 19’ is noteworthy, since it throws some light on both the architectural and the social aspects of the domestic milieu in the city. In the following part, ‘PDura 19’ will be studied first and then the data that came from the archaeological remains will be summarized and evaluated in reference to the information provided by this document.
7.3.1 Textual Evidence: PDura 19
The parchment coded ‘PDura 19’ was found in 1935, during the Yale excavations and was published in Final Report 5.1, which was dedicated to parchments and papyri (Fink et.al, 1959: 104-109). The document was recovered in the earth fill of Tower 22 that was piled up during the reinforcement of the city walls against the Sasanian threat. The parchment recorded the division of a property between four sons of a man named Polemocrates. The text was in Greek and dated to 88/89 AD, to the Parthian period of the city. It was relatively well preserved but fragmentary, and thus is partially restored in the following translation by Baird (2014: 51-52):
In the reign of the king of kings Arsaces, benefactor, just, manifest god, and friend of Greeks, in the year 336 as the king [of kings reckons, but 400 of the former era, on the ? day of the month ?, in Europos] in Parapotamia. There have made a voluntary distribution among themselves Demetrius so-called Nabusamus, Seleucus so-called [ - - - - - ] so-called Zabduas, the sons of Polemocrates, son of Demetrius and grandson of Polemocrates, of Europos; (they have distributed) the [houses] belonging [to them in - - - - of which they have stated that the one was] allotted to Polemocrates, [their father] just named, in [a division] made by him with his brother Apollophanes [through the local court in the] year 364 of the former reckoning, on the [?] of the month Audinaeus, while the other was bought by him from Nicanor, son of Alexander and grandson of Mnesippus [- - - in a transaction effected] through the same court [in the year ? of the former reckoning,] on the second of the month Dystrus, and they have acquired by allotment as follows: Demetrius [the oecus - - Nicanor - - - - the] storeroom in - [- Antiochus the house in - - ] and the upper rooms over it and over the common entrance and exit passageway [- - - Seleucus the house toward the] north and the upper room over it and [another] room which is partly collapsed - - erasure - - [and they will have] in common the court and the entrance and exit passageway and the balcony, which they will use in common. Nicanor shall block up the door of [the house which] fell to him – [that opening into the] andron
282
which fell to Demetrius, and he shall take to himself the doors and the trim and open for himself from the above storeroom another door [into the house allotted to him,] and Antiochus shall plaster up the door of his upper room opening into the upper room of Nicanor and he shall open for himself another door [on to the balcony (?),] and Seleucus [shall open] from his house a door into the common court; and all shall construct anew the [party wall] between this court and another court [- - - - - -;] and they shall furnish to each other access to this party wall and to the wall of the house which fell to Seleucus, placing upon the party wall [a flight of steps (?) - - - - - -] of the stoa through which they will ascend from his [?; and Demetrius shall give in addition to Seleucus toward the equalization of their shares [?drachmae] of silver [- - - - - - with the trim (?) and doors and all other belongings and appurtenances; and each of them shall be the owner of the household goods, in his own part. And all have agreed [- - - not to] bring a charge against each other now or in the future either concerning this division or concerning those things which each of them, separately, has acquired or may acquire or concerning [- - - - - - - but if anyone does not] abide by this agreement but raises such a charge, he will pay to those who abide by the agreement without argument or court decision a penalty of 1000 drachmae of silver, with the same amount going to the royal treasury, and [this distribution shall be valid even so; it is executed in double copy. Witness: Seleucus, Adaeus and Danymus, royal judges and memebers of the order of first and chiefly honored friends and bodyguards; [Pausanias, bailiff and the collector,] member of the order of bodyguards; Damonicus son of Marius, Democrates son of Damonicus, and Apollogenes son of Ptolemaeus.
PDura 19 throws light on both the architectural and the social aspects of the domestic contexts in Dura-Europos in the Parthian era, which does not seem to have changed much under Roman rule. It also demonstrates the hybrid culture of the city in that period. Though it was a Parthian document, the text was in Greek and the date was given in both Parthian and Seleucid calendars. The four brothers in question had both Greco-Macedonian names and Semitic nicknames, while their father, uncle, and grandfather bore only Greco-Macedonian names. The use of Greek terms to describe particular rooms within the house is also notable.
According to the document, the shared property consisted of two adjoining houses one of which was acquired by Polemocrates after an earlier property division between him and his brother Apollophanes, and the other was purchased also by him from a certain Nicanor. In 88/89 AD the property passed to the ownership of four brothers probably through inheritance following the death of their father, though this was not clearly
283
stated in the document. The house was apparently a two-storey structure with two adjoining courtyards and when the rooms were apportioned, each brother got a part around the courtyard on the ground floor and the space immediately above it on the first floor. In this allocation, Antiochus got the part of the house adjacent to and above the entrance, Demetrius got the room(s) called andron, Nicanor got a part including a storeroom and Seleucus got the part on the right side that comprised two rooms, one of which was ruined, and the space over them. They received the ownership of the household goods that were in use in their portions too. In the physical division of the house, the isolation of each unit was guaranteed by giving each brother some responsibilities, like blocking a door opening to the other person’s part and opening a new door from another point instead. The inequality in the lots of Demetrius and Seleucus was compensated by payment in cash. The space that was allotted to each of the four brothers was identified as their oikos. According to the document, each brother had equal rights of access to the entrance/exit, the courtyards, the staircase and to the ikria that was presumably a wooden mezzanine or balcony that provided circulation between the rooms of the upper floor (Saliou, 1992: 88; Baird, 2014: 53).88
7.3.2 Overview of the Archaeological Remains
The orientation, size and form of the housing units in Dura-Europos were mainly dictated by the building plots that were determined by the Hellenistic urban grid. It was suggested that the regular insulae were originally divided into eight equal plots each of which was occupied by an individual housing unit with a ground surface area of about 306.25 m2 (figs. 301, 302) (Benech, 2010; Hoepfner and Scwander, 1986: 215). None of the excavated blocks, nevertheless, could maintain the equal parceling system, due to the several phases of construction and restorations. In some sections of the city, where the urban grid was not fully applied, like in insulae G1 and G3 in the agora sector and insula B2 in the ravine, a more organic and irregular plan layout and development were applied.
88 Similar situation was attested in Ptolemaic Roman Egypt (Uytterhoeven, 2009).
284
The most common building material in the domestic and the public architecture of Dura-Europos, and in all of Mesopotamia in fact, was sun-baked mud brick. The houses were mostly constructed over stone rubble foundations that were consolidated with clay mortar and were plastered for waterproofing. The superstructures were built of mud-brick on top of the stone bases. In the construction of some later buildings of military use, fired brick was also used. The thresholds, jambs, and lintels of the doorways were carved, sometimes also molded, from gypsum blocks that were provided from the quarries in and around the city. The terrace roofs of the houses were constructed with wooden beams that were overlaid with woven mats made of reeds and were plastered (fig. 303).
In terms of size, a large quantity of the unearthed houses of Dura-Europos measured between 100 m2 and 400 m2. Larger houses, like the House of Lysias and the House of the Large Atrium, were basically formed by connecting the neighboring housing units within the same insula and they covered an area of about 2,400 m2 in their largest extent. In some cases, conversely, the larger houses were divided into smaller units, the ways and means of which were clarified in the papyrus PDura 19. Numerous modifications in the circulation patterns between the neighboring housing units as well as within the houses themselves during subsequent phases of occupation attested changes in spaces, room exchange between neighboring houses and change of property ownership.
The houses of Dura-Europos had an introverted character with very limited communication with the street. The facades were mostly left simple, without any external decoration or elaboration. The very few windows preserved on the facades were narrow and located above eye level. The entrance doors were usually located in one corner of the houses and gave way to narrow entrance halls that were enclosed by an interior wall, mostly by that of a stairwell, or a partition wall that formed an L-shaped path leading to the courtyard. As such, visibility from the street to the courtyard was blocked. As indicated in document PDura 19, a single house might have been owned and/or inhabited by more than one family and in the latter case the single
285
entrance of a modest house could have served different families that shared the same residential building. In larger houses, multiple street entrances allowed different groups of visitors or family members to enter the house separately. Passages between the neighboring houses that presumably belonged to the members of the same family could also have permitted circulation between houses without making it necessary to exit into the public space of the street. A considerable number of houses, especially those in the agora sector, were associated with shops that opened directly to the street.
The most characteristic and fundamental element of domestic architecture in Dura-Europos was the central courtyard around which the main rooms were organized. Larger houses had several courtyards that determined separate units, presumably with different functions. The courtyard provided air and light to the surrounding rooms and organized the circulation between them. It was also the functional core of the house where many domestic activities took place. The unearthed ovens or tannours that were mostly located beneath the staircases, and some movable equipment like mills and mortars attested that cooking and food processing activities took place in the courtyards (Baird, 2012: 236). Similarly, a number of altars, sculptures of deities, and a burning incense recovered in the excavated courtyards of different houses indicated that the religious activities also took place there. Based on the information provided by document PDura 19 the courtyards in some houses could have been shared by a number of families and could act as stages of small community life within the private context.
In terms of appearance, the courtyards in the houses at Dura-Europos reflected Mesopotamian influence rather than the Greco-Roman one.89 Among the excavated buildings, only two palaces: the Citadel Palace and the Roman Palace had peristyle courtyards. In the Redoubt Palace and in some of the relatively well-to-do houses, modest porticos with one, two or three columns were created on one or two sides of the courtyards. Courts of House B2-C and F1 of the House of the Large Atrium are
89 For an overview on the evolution of courtyard houses in Mesopotamia see Erarslan, 2021.
286
unique in this respect, since they had columns on three sides, though the colonnade of the latter was walled up in a later phase. The columns were constructed with rubble stone and plastered, and hence were not well preserved. The floors of the courtyards, as well as of the other rooms, were mostly of beaten earth or of plaster; in only some large houses like the House of Lysias and House G1-A the courtyards were paved with fired brick. Staircases and the cesspool/cisterns that will be mentioned further below, were the characteristic features of the courtyards.
In addition to the central courtyard, the most common element of the domestic architecture in Dura-Europos, was a spacious room that was differentiated from the other rooms with its size, location and layout. Annexed by side rooms, it was apparently the main room of the house and was usually located on the southern side of the courtyard facing north, providing a cool environment in the hot climate. The entrance of the room was on its long side and was monumentalized with wide doors and a few steps; in some more elaborate houses the entrance could receive triple doors or free standing pillars. The room, especially in more modest houses, was surrounded by benches of about 1.00 m wide and 0.10-0.20 m high that were built with plaster or plastered rubble. The benches were probably used as a base for carpets and/or cushions. In larger houses the main rooms were devoid of built-in benches and were probably furnished with more substantial furniture in wood or metal. In some houses the main rooms were decorated with molded plaster cornices, friezes and painted plasters. The floors were mostly of beaten earth or plastered. The only mosaic pavement in the domestic context came from House E4 had a Roman military use in its later phase of occupation. The fragments of the mosaic pavement had geometric patterns and belonged to an upper room whose function remains unknown.
The earliest example of this kind of reception rooms comes from the palatial context. The large rectangular rooms A and J in the Redoubt Palace, with their annex rooms, wide central doors and two-columned porticos built in front are distinguished in the plan layout. Nielsen identified these suits of rooms as ‘andron complexes’ that formed the official section of the palace (1994: 116). A room was called andron in PDura 19
287
document, presumably referred to the main room of the house in question.90 In the excavation reports and in later publications the main room in the Dura houses was thus identified either as ‘andron’ or as ‘liwan’ but most commonly as ‘diwan’, a word of Iranian origin that is used to denote the reception rooms in houses or palaces in Islamic architecture. (Welles, 1933: 80; Hopkins, 1934; Rostovtzeff, 1938; Brown, 1944c; Kraeling, 1967; Hoepfner and Schwander, 1994).
The main room, being the largest and most decorated space of the house, obviously functioned as the reception room. It was probably used for dining activities, too. A wall fresco found in the main room of house M7-W near the Palmyrene gate supported this idea. The fresco depicted a banquet scene in which three men with both Greek and Palmyrene names were reclining on cushions and were accompanied by two standing servants (fig. 304). In the smallest houses with two or three rooms, like House C7B, the room with benches should have been multi-functional and presumably was used for some domestic activities and even for sleeping.
In the Roman Palace, the architectural and decorative layout of the reception rooms differed from that of the earlier houses. Rooms 28 and 2, together with the symmetrically arranged annex rooms, were apparently formed the reception suits of the palace. They came into prominence with their size and height. The false vaulted ceilings with decorative paintings in both rooms and the apsidal form of room 2 are unique to this building. The entrance of the rooms was emphasized with wider doors as well as the wider intercolumniation of the portico in front of the door. A statue that presumably stood on the base found in-situ in front of the central colonnade indicated the concern for the view from room 28 towards the colonnaded courtyard. In the Roman Palace, the reception rooms were placed perpendicular to the courtyard and their entrance doors were placed on the short sides, as opposed to the main rooms of the other houses of Dura-Europos. This layout should have increased the sense of
90 Saliou argued that the andron mentioned in PDura19 documents should have designated the main room and the annex rooms together (1992: 94).
288
axiality and depth. What is striking in the Roman Palace is the lack of mosaic pavements which is very common in the Roman domestic context, as exemplified in Italy and provinces. The floors of the reception rooms in the palace were just plastered.
Though the two houses recorded in PDura 19 document were obviously two-story high, only a few houses in Dura-Europos yielded archaeological evidence for the existence of an upper story. In almost all of the houses, staircases were placed in the courtyards. Most of them were rather simple, resting against a wall of the courtyard and probably led to the flat roofs either to do maintenance or some domestic and daily activities that changed seasonally. In some houses there were wider and more comfortable staircases with three or four flights that were enclosed by walls. These were not likely to have been service stairs but probably led to upper floor spaces of some importance. The spaces at the terrace roof could have been semi-open spaces covered with removable lightweight shelters. Today, the terrace roofs in the same region are still used for domestic activities like drying fruits and vegetables and/or for sleeping on hot summer nights.
The water supply and sewage systems in the domestic context of Dura-Europos could not be fully understood from the current archaeological data. The plateau on which the city was founded is about 40.00 m above the Euphrates River and there was no water source on the plateau. An aqueduct and terracotta pipes uncovered in the northern part of the city were built in the Roman period and provided water only to the military buildings. Most of the houses had deep basins that were dug into the ground at the center of the courtyards. The mouth of the basins was covered by rectangular gypsum blocks with a circular hole in the middle. The proper function of those features could not be determined, except the cistern in the Citadel Palace, and they were interpreted either as latrines (Kraeling, 1967: 11; Hopkins, 1934: 31-32;), cesspools (Rostovtzeff, 1938: 49; Perkins, 1973: 22), septic tanks to collect water from the roof and courtyard from which water seeped into the ground (Hoepfner and Schwander, 1994: 235) or as cisterns to collect water (Pillet, 1933: 29; Baird, 2012: 236) in the excavation reports and later publications. The identification of those features as latrine or cesspool is not
289
plausible since they were placed in the middle of the courtyards that were the most frequented spaces of the house with a distributive function, and also the places where many domestic activities including cooking took place. In some cases, traces of waterproofing were noticed which suggested the use of these features as cisterns to collect and store water, but the practicality of covering their mouth with heavy stones is questionable (Allara and Saliou, 1997: 151-152). Taking into account that the courtyards of the houses in question were not covered with hard surfaces like stone and mosaic but left as beaten earth, and domestic activities like cooking and washing took place in them, it seems more probable that these underground were built mainly to drain rainwater and wastewater. Latrines and bathrooms were identified in some houses, but most securely identified examples occur in connection with military installations such as that in House E4 and the Roman Palace. Latrines were commonly placed under the staircases in the courtyards, especially in modest houses.
The most common utilities that were found often in-situ within the houses were large gypsum or plaster containers that were more than a meter high, with varying inner and outer diameters. They were mostly placed in the entrance halls and courtyards. They were indeed, not exclusive to the domestic context but were also found in shops, public buildings and, even the streets. These features were named “coolers” by the Yale excavators since they were thought to store water and hence to function as climatization installations that would contribute to cooling the environment. It was also suggested that the water they contained served guests to wash hands and feet (Hoepfner and Schwander, 1994: 280). The successive layers of plaster coats found in some of the containers led the excavators to interpret them as plaster mixing bowls, used during maintenance and repair works (Hopkins, 1934: 33).
The domestic architecture in Dura-Eurpos was distinct in many respects. The city apparently sustained its Parthian and Greek culture in the case of domestic architecture, which manifested from construction and spatial planning to space design, decoration and use. Neither the use of iwan, the semi-open alcove like space opening to the courtyards, nor the brick vaulting that were particular to Parthian buildings were
290
found in the Roman domestic context of Dura-Europos. No triclinium with the common T-shaped mosaic arrangement were identified. The lack of common decorational features of Roman domestic architecture, like mosaic pavements or decorative water features like nymphaeum and pool, even in the Roman Palace is noteworthy.
The distinct character of the domestic architecture of Dura-Europos in both the Parthian and Greco-Roman contexts can also be perceived from the problematic vocabulary used to define the houses in the excavation reports. The terms ‘atrium’, ‘tablinum’, ‘andron’, ‘oikos’, ‘iwan’, ‘diwan’, ‘haremlik-selamlık’ and ‘gynaiconitis’ have been interchangeably used by scholars in the same context for the identification of domestic spaces. In the excavation reports, no formal, spatial or contextual comparisons with the similar spaces in Hellenistic, Roman or Parthian period houses unearthed elsewhere were made. It is most probable that the terms were used out of context under the influence of current scholarship that was dominated with the typological studies on the houses of Pompeii and Herculaneum at that time, and the reports of more or less contemporary excavations in Olynthos.
291
CHAPTER 8
CONCLUSION:
LIVING AND DWELLING AS A ‘ROMAN’ IN THE EASTERN PROVINCE OF SYRIA
8.1 Living as a ‘Roman’ in the Eastern Province of Syria
The territory of the Roman province of Syria had seen continuous occupation since prehistoric periods with many cities that flourished in different cultural eras. The area was Hellenized circa 300 BC by the first Seleucid king Seleucus I Nicator as part of his colonization program. The Seleucid colonists had been settled in or near the existing Bronze and Iron Age settlements that were located at the main strategic points along the new boundaries and/or on the important trade routes. In 64 BC, the region was annexed by the Roman general Pompey and was proclaimed as a new province of the Roman Empire.
Romans did not build any new cities in the province of Syria but ruled the existing ones. All the five cities examined in this thesis-Antioch on the Orontes, Apamea on the Orontes, Zeugma, Palmyra, and Dura-Europos-were initially Seleucid colonial foundations that had flourished in a Semitic milieu. The economy of these cities was mainly based on trade and/or agriculture and all had military importance in both the Hellenistic and Roman periods. The five cities, nevertheless, differed in terms of their political, social, and cultural contexts and urban developments since they experienced different forms and extents of Greco-Roman power and influence.
The brief overview of the historical and political development of the five cities have revealed that there were not always sharp transitions in the order and functioning of the cities between Hellenistic, Roman, and even Parthian rules. Different institutions
292
and approaches amalgamated in the political and administrative contexts of the cities. The Greek institutions remained in function in the Roman administrative system while the local tribes or families were always influential in politics. The long-standing local and regional traditions and practices that matured and sustained throughout the deep-rooted history of the region became exposed to varying degrees of Greco-Roman influence, a mutually fostered process that brought about a social and cultural amalgamation in the cities of the eastern province of Syria.
The population of the cities was dominated by Greco-Macedonian colonists, and the Semitic Mesopotamians, the indigenous inhabitants of the region. The residents with Latin origin were relatively few in number, since the Roman expansion in the eastern territories, unlike the Hellenistic colonization, was not accompanied by any significant immigration of people from the mainland.91 There was also a fairly large Jewish community and some Arabs and Iranians among the citizens as attested by the inscriptions and the funerary steles. The elite were the first to receive the legal status of Roman citizenship, as a reward for their loyalty, following the annexation of the region in 64 BC. Then, in 212 AD all the free citizens of the province, regardless of their ethnic origins, received the same status.
What it meant to be a ‘Roman’, is a significant question that is raised up in the recent scholarship on the Roman provinces which is often associated to the contexts of Roman Imperialism and discourse of ‘Romanization’. Such approaches focus more on the concept of identity rather than the more conventional insights of ‘Romanization’ like the issue of cultural transformation in reference to a unitary Roman culture and the discussions based on the binary oppositions the center and periphery and/or civilian and barbarian.92 According to Woolf, to be a Roman was a discourse, rather than an
91 The only long-standing Roman colonist settlement of the East was Berytus, which was established by the veterans under Augustus, and was later split off as the colonia of Heliopolis (Millar, 1993: 527).
92 For critical overviews on the concepts and paradigm of ‘Romanization’ see Janniard and Traina, 2006; Woolf, 2014 and 2021; Van Oyen, 2015. For studies focusing on identity in the Roman provincial contexts see Woolf, 1994 and 1995; Jones, 1997; Mattingly, 2004; Pitts, 2007 and 2008; Revell, 2008 and 2015.
293
absolute status, based upon a shared idea of being Roman (1998: 7-16). Jas Elsner presents a similar approach and define ‘Romanization’ as “a process whereby Gauls, Libyans and Levantines might share the ideals of a single culture, despite all their manifest differences” (1998: 118). This thesis, too, focused on the personal and communal identities of the population of eastern cities under the Roman rule and argued the notion of ‘being Roman’ rather than ‘becoming Roman’ in the provincial domestic context. By throwing light on personal and communal self-representations it aimed tried to understand to what extent the members of this population perceived themselves as Roman and how they, consciously or subconsciously, expressed this identity. The multi-linguism in the written contexts and the nomenclature, multi-religious sphere of the cities, the variety in the funerary cults and arts, and the material culture examined so far revealed the complex nature of the social and cultural formations as well as the sense of identity of the individuals in the Roman cities of the eastern province of Syria. ‘Being Roman’ in this environment, thus, did not represent a single and constant identity; it was a multiple, temporal, and situational state and was represented in varying extents and multiple ways through different mediums from onomastic to art and architecture.
8.2 Dwelling as ‘Roman’ in the Eastern Province of Syria
The main determinants in the planning of the houses unearthed and studied so far in the Roman province of Syria were the urban plan and topography into which the buildings were integrated, and the spatial requirements of the domestic practices and rituals. The climatic conditions, availability of construction materials, the ongoing local traditions and the prevailing Greco-Roman influences in cultural and architectural contexts were also operative in the social and architectural formation of the domestic space and the design of the residential buildings. Nevertheless, neither the political, economic, and social contexts of the region as mentioned above nor the architectural practices, approaches and trends were constant in the entire geography and time period under consideration but were subject to changes and modifications in the course of the seven centuries of Roman rule.
294
In the following part of this section, the sample scrutinized so far and the data and interpretations put forth in the previous chapters will be overviewed in a chronological perspective to highlight the common conceptions and features, as well as site-specific issues and cultural assimilations in the domestic milieu of the Roman cities in the eastern province of Syria. Following the framework of the previous chapters, the residential buildings will initially be reviewed with reference to the urban contexts and the topography and then the main features of their design principles and architectural vocabulary will be highlighted.
8.2.1 Residential Buildings as Part of the Urban Fabric
The urban residences of the wealthy Roman elite were used to display power and wealth of the owner. With their size, splendor and/or location in the urban fabric these houses provided prestige and reputation to their owners and accentuated their public images (Wiseman, 1987; Uytterhoeven, 2020: 332-333; 2022). The importance of the visibility of the houses in the urban context and their ‘transparency’ for self-promotion was emphasized by many ancient authors. In 44 BC, in the De Oficiis Cicero wrote:
We have heard that Gnaeus Octavius- the first of that family to be elected consul- distinguished himself by building upon the Palatine an attractive and imposing house. Everybody went to see it and it was thought to have gained votes for the owner, a new man, in his canvass for the consulship… (De Officiis 1.138)93
Velleius Paterculus in his treatise on Roman History written in the early 1st c. AD mentions Livius Drussus who was a political figure in the city as:
When he was building his house on the Palatine on the site where now stands the house which once belonged to Cicero, and later to Censorinus, and which now belongs to Statilius Sisenna, the architect offered to build it in such a way that he would be free from the public gaze, safe from all espionage, and that no one could look down into it.
93 Translated by Miller, 1913
295
Livius replied, "If you possess the skill you must build my house in such a way that whatever I do shall be seen by all”. (Vell. Pat.2.14.3)94
Several large and luxurious mansions that were located in the prominent locations in the urban fabric and dated from Republican period to late antiquity also confirmed the ancient literature.95 Thanks to their location, close to the urban thoroughfares like colonnaded avenues and major streets or to public buildings like agora or theatre, these houses stood at the center of activity and were easily accessed (Ellis, 1997: 41). Some with their elevated positions could also attract the gaze and the attention of the passers-by and also had a good view of the urban scenery. Some of the houses in Syria examined in this study display similar concern for location and visibility within the urban fabric as well.
The urban layout of most of the Roman cities in Syria was characterized by the orthogonal grid system that was implemented during their foundation or in a later phase of urban development in the Hellenistic period. Among the five cities examined in this study, Antioch on the Orontes, Apamea on the Orontes, Apamea on the Euphrates and Dura-Europos were initially planned according to this principle. In these cities, the rectangular building plots or insulae that had different dimensions often with a ratio of about 2:1, defined the orientation and, to a certain extent, limits and layout of the buildings. The larger properties covered the whole insula or a large part of it whereas the modest ones shared the same insula with a number of other properties. In Apamea on the Orontes, for instance, the excavated insulae measured 110 m x 55 m and were occupied by one to three houses (fig. 79). The excavated domestic buildings in that city were vast residences that measured between 2,000 m2 to 4,500 m2. Dura-Europos was unique among the five cities examined in this study in terms of urban development since it did not have the essential components of the Greek and Roman
94 Translated by Shipley, 1924.
95 For the Terrace Houses at Ephesus see Österreichischen Akademie der Wissenschaften in the Forschungen in Ephesos series, for the Villa Above the Theatre in Ephesus see Thür 2002, 258-264; Baier 2013; for the Urban Mansion at Sagalassos see Uytterhoeven et al. 2009, 2014.
296
cities in the region like the colonnaded streets and the theatre -except the modest one intended for the use of the Roman army. In Dura-Europos, the insulae measured about 70 m x 35 m and the houses were relatively modest, most of which measured between 300 m2 and 400 m2 (fig. 243). The largest residences brought to light in Dura-Europos measured about 2,450 m2 covering the whole insula. The Roman Palace which was the latest dated unearthed building, on the other hand, did not fit into the urban grid and extended over a large area of about 5,500 m2. Palmyra and Zeugma differed from these cities since they did not develop according to a predetermined Hippodamian plan. In Zeugma the steep topography seems to have challenged the orthogonal grid plan and the city developed in an irregular urban layout. The residential buildings in Zeugma were laid out on artificial building terraces that were partly carved from the bedrock and partly backfilled. This was also the case in the higher districts of Antioch on the Orontes that extended on the slopes of Mount Silpius and in some districts of the port city of Seleucia Pieria located on the slopes of Musa Dağ. In those cities, the size and layout of the buildings were somewhat restricted by the stretch of artificial building terraces whereas the sloped topography enabled the houses to extend on different floor levels. In Zeugma, the largest residences measured about 500 m2 on average. Different than all these cities, Palmyra developed in reference to a different model of urban space organization. The residential quarter in the western part of the city was organized around streets that were not exactly parallel to each other but ran in the same direction with varying distances in between. In this district, the dwellings had to expand longitudinally between the streets. In the relatively later developed eastern quarter of Palmyra there seems to have been more spacious building plots that allowed a relatively free expansion of buildings. The size of the unearthed houses in Palmyra thus varied between 450 m2 and 2,000 m2.
The unearthed houses that occupied a considerable area within the urbanscape of the five cities studied in this thesis could approach the scale of some public buildings. Not every house with large dimensions, nevertheless, displayed the same grandeur and accent within the urban fabric. The houses in Apamea on the Orontes and Dura-Europos offer a good sample for such a comparison since they were relatively well-
297
preserved with some of their façade walls still standing up to a certain height prior to the excavations. The use of large stone blocks as construction material, by itself, seems to have provided a certain degree of grandeur to the residential buildings of Apamea on the Orontes, in comparison to the mud-brick used for the façade walls in Dura-Europos (fig. 82). The street doors of the houses in Apamea, too, displayed a monumental scheme with their large size and decoration that had parallels in several public buildings in the same city (fig. 83). In Dura-Europos, nevertheless, the vast residential buildings did not display such a grandeur within the urban fabric. The House of Lysias and the House of the Large Atrium, for instance, covered the whole insulae measuring about 2,450 m2. Yet, they were basically formed by the combination of several neighboring housing units and that is why the façade walls of these two extensive houses were not differentiated much from the facades of the other insulae that included several modest dwellings (figs. 244, 270, 278). The main entrance door of the House of Lysias displayed a rare case since it was accentuated with two columns on both sides (fig. 269). Two houses from Palmyra, the House F and the House to the Southeast of the Theatre, too, display different approaches towards the integration of the residential buildings into the urban fabric. House F, being a large dwelling of almost 2,000 m2, reflected a Mesopotamian influence rather than the Roman one which is further mentioned below (fig. 215). It did not reflect its extent or wealth to the outside via its façade. The House to the Southeast of the Theatre, too, was a spacious residential building that was located in the close vicinity of an important public building. The north façade wall of the house had a curvilinear layout following the line of the colonnaded street in front and the span between two columns of the portico corresponding to the street door of the house was wider than the others, accentuating its entrance (figs. 224, 225). As such, not only the private building was adapted to the urban space but the urban space and components were also adjusted according to the private building and, hence also marked it.
Whilst at least some of the private residences were designed to attract the gaze of those outside in the street, a certain privacy was intended in most cases by preventing any direct visual or physical access from the street into the core of the house. The entrance
298
doors and the vestibules could not be spotted for all the houses but the preserved ones displayed similar entrance schemes and circulation patterns. The vestibules were usually located on or close to one of the corners of the houses and they provided an L-shaped path towards the courtyards. This approach seems to have changed partly in late antiquity when a more axial and direct communication between the domestic and urban spaces were provided as was the case in the House to the South-East of the Theatre in Palmyra and the later phases of occupation in the Building with Triclinos and the House with the Trilobed Columns in Apamea on the Orontes. In the latter two houses, this type of rearrangement provided an axial sightline and physical access from the street, to the courtyard and the main reception room of the house (figs. 70, 110).96
Due to the poor state of preservation of the remains, there is very limited information for the window openings on the façade walls. In the relatively well-preserved houses of Dura-Europos, only a few windows were spotted. Those were very narrow openings above the eye level and were obviously intended to provide light and air to the inner space rather than to provide a visual communication with the outside. In the later built Roman Palace in the same city, nevertheless, a different type of communication with the outer world is observed. The north-eastern part of the Roman Palace that probably constituted the private living quarters of the family turned its back to the peristyle courtyard and was opened to a terrace overlooking the Euphrates built on the edge of the building. The rooms on that part of the palace opened to the terrace via windows or doors and seems to have enjoyed the view of the river. In Apamea on the Orontes, the vast reception room of the House of the Consoles opened to the street via tall windows while the reception rooms of the House of the Console Capitals, the House of the Deer, and the House of the Pilasters had windows and doors that opened to the backyards. It is also very likely that the totally collapsed upper floors of the houses in Apamea on the Orontes as well as in the other cities, had windows opening outside that provided not only light and air but also a view of the street. In some houses, the
96 For the axial arrangement in the Pompeian houses see Clarke, 1991: 17 and Wallace-Hadrill, 1994: 38-64.
299
flat glass fragments and the iron grills found in the debris, though not frequent, attested to the existence of the openings and/or glazed windows on both the ground and upper floors. The unearthed remains in Antioch on the Orontes, provided almost no data on the existence and location of doors and windows since walls were not preserved. Still, as Libanius pointed out, the residents of Antioch especially those living on the higher slopes of the city and in Daphne should have enjoyed the nature and the view through terraces, balconies, and windows (Or. XI. 200).
8.2.2 Architectural Layout and Vocabulary: Form, Function, and Meaning in the Domestic Space
The urban residential buildings unearthed so far in Roman Syria displayed an introverted architectural layout and were organized around inner courtyards. The courtyards were not designed as planted gardens in the Roman fashion but were paved or beaten earth open spaces, resembling the Greek or Mesopotamian houses. The two main functions of the courtyards were to provide light and air to the surrounding rooms and to regulate the circulation between them. Though the inner courtyards were the most common and enduring feature in the plan layout of the urban residences in the region for several centuries they did not display the same architectural and conceptual approaches.97 The design principles of the residential buildings, and accordingly, the form, function, and meaning of the domestic spaces, including the courtyards, varied in different contexts and were subject to changes and transformations from the Hellenistic period until the end of antiquity.
The houses excavated in the five cities under consideration and dated to the Hellenistic period are only a few in number and they have provided limited architectural data due to their poor state of preservation and/or later rearrangements. These buildings were characterized by the central courtyards with either a peristyle or a portico on one or two sides that are reminiscent of pastas or prostas in Greek houses. The porticos were
97 For a discussion of courtyard houses that focus on late antique Asia Minor see Özgenel, 2022.
300
arranged in front of the most distinguished rooms of the buildings that usually came into prominence with their size, layout, and axial location and that were presumably used as reception rooms. The porticos not only highlighted the entrance of these rooms from the courtyard but also formed a kind of ante-space for them. Two palatial buildings excavated in Dura-Europos dated to the 2nd c. BC and two urban residences partially uncovered in Zeugma whose initial construction was dated to the Seleucid or Commagenian rule in the city well illustrate these architectural approaches. The partially preserved southern section of the Citadel Palace in Dura-Europos displayed an architectural layout with several rooms organized around a large peristyle courtyard (fig. 236). The most spacious room of this section axially opened to the courtyard through a wide opening with a second row of the colonnade that was formed by three columns. The House of the Helmets in Zeugma displayed a similar layout on a more modest scale with its peristyle courtyard with three columns on all its four sides (figs. 123, 124). The north porch of the peristyle, which was wider than the other three was arranged in the form of a portico and probably gave way to a prominent room in the unexcavated northern part of the house. The Redoubt Palace in Dura-Europos, which was more or less contemporary with the Citadel Palace, had a central courtyard that was surrounded by rooms on its four sides (fig. 239). On the southern and western sides of the courtyard, porticos that were formed by two columns led to spacious rooms and in a way served as ante-spaces for these rooms. The House of the Fountain in Zeugma had a small courtyard arrangement but it too had a wide opening with two columns on its southern side that gave way to a spacious room in the form of a semi-open loggia (fig. 121).
The design principles of the excavated houses that were built and/or inhabited in the first two centuries of Roman rule in Syria did not differentiate much from their Hellenistic predecessors. The above-mentioned Hellenistic houses in Zeugma continued to be inhabited without much architectural change until at least the end of the 2nd c. AD when they underwent major modifications. Two latter houses in the same city whose initial construction date was estimated as the 1st c. AD displayed similar plan layouts with their antecedents. The House of Dionysos and Ariadne, just like the
301
House of the Helmets, was organized around a modest peristyle courtyard with three columns on its four sides whereas the House of the Bull had a two-columned portico on the north-eastern side of its inner court that gave way to a semi-open loggia, resembling the one in the House of the Fountain (figs. 133, 123). The rooms of these houses were placed around the courtyards with which the largest rooms had more direct and often axial communication. In Antioch on the Orontes, the earliest dated house was the so-called Atrium House that was presumably built in the Augustan period. The Atrium House was the only residential building excavated in the city that had a peristyle courtyard. (fig. 12). The Atrium House reaching about 1,500 m2 in size, and its accordingly large peristyle, were much more monumental in scale than the houses of Zeugma. The northern wing of the building with three rooms that were distinguished by their size, layout and good quality wall masonry, probably formed the ceremonial section of the residence. One of these three rooms, room 130, was likely to have been centered at the axis of the courtyard and was the main reception room of the house in the first phase of occupation and until the house underwent major restoration works in the 2nd c. AD. The excavated houses of Apamea on the Orontes that were built or rebuilt after the earthquake of 115 AD displayed similar plan layouts. The peristyle courtyards in those houses, often with the Rhodian layout, were more elaborate than those in Zeugma in size, reaching 1,350 m2 in some cases (figs. 90, 94, 100, 123, 133). The remains of the residential buildings spotted in the western urban quarter of Palmyra were dated to the first half of the 2nd c. AD. The partly hypothetically drawn plans of these relatively smaller houses reveal that they were organized around peristyle courtyards with Rhodian layout too (fig. 213). In all these houses at least one room behind the emphasized colonnade of the peristyle came into prominence with size, location, and mostly with its axial communication with the courtyard. These rooms are assumed to have been used for reception purposes where the owner of the house received and/or entertained his guests for different occasions.
From the second half of the 2nd c. AD, an increased amount of western influence that embraced some eastern traditions is observed in the urban residential architecture in provincial Syria. In that period new domestic practices and rituals and accordingly new
302
types of spaces, the most significant of which was the triclinium, were introduced into the domestic context. The existing spaces in earlier buildings were also redesigned and/or redecorated to provide a suitable setting for the current needs of domestic life. As an outcome of these rearrangements, the public spaces within the private sphere became much more differentiated and emphasized. The reception spaces became the visual target and the physical destination at the end of the pronounced circulation path of the house that started from the courtyard, or in some cases even from the street. The courtyards, too, were rearranged and more articulated with the introduction of new architectural and decorative vocabularies, like mosaic pavements, nympahea and shallow pools or basins. As such the courtyards were transformed into places of comfort and relaxation rather than spaces of circulation. The visual and physical interconnections between the courtyard and the main reception space of the houses became much more accentuated, as well.
The triclinium, a Roman form of space destined for reclined banqueting, was introduced in the urban domestic architecture of provincial Syria in the 2nd c. AD (Martz, 2015: 176). The earliest dated triclinium with the common T-shaped mosaic arrangement in the eastern Roman context was found in the Atrium House in Antioch on the Orontes (figs. 12, 13). The Atrium House underwent a major restoration work after its partial destruction presumably caused by the earthquakes of 94 AD and/or 115 AD. During this restoration work, the peristyle courtyard of the house was reduced in area with the addition of new rooms, and some parts of the house were redesigned. The northern wing of the building that comprised the presumed ceremonial section of the dwelling was rearranged and one of the three rooms in that part, room 82, was enlarged and paved with a sumptuous mosaic with five figural panels that were arranged in T shape, indicating the use of the room as a triclinium. The entrance of the room from the courtyard was highlighted by the corresponding two columns on the northern portico, which offered a wider opening than the others. In the later houses of Antioch on the Orontes, there was neither a peristyle type of colonnade. Instead the triclinia became the spatial focus with their illustrious size, layout, and decoration and the courtyards were treated to function as their annexes or extensions. There were often
303
colonnaded porticos on one or two sides of the courtyards to which the triclinia and the annex rooms opened. These porticos might have been used to compensate for the lack of peristyle in the courtyards of smaller houses.98 In the houses of Antioch on the Orontes and its suburbs, the triclinia and the courtyards, as well as some other smaller rooms that were probably the living units of the family, were decorated with sumptuous mosaics. The mosaic pavements did not only enhance the visual quality of the rooms but were utilized as an architectural tool to link the spaces by sightlines and accentuate the circulation patterns. In addition to colored mosaic ensembles, decorative nymphaea which magnified the spatial value of the triclinia and provided a theatrical effect for the diners, while also constituting a medium for displaying comfort and luxury, were utilized in almost every courtyard.
More or less contemporary with Antioch on the Orontes, more specifically during the late 2nd and the first half of the 3rd centuries AD, the houses of Zeugma underwent some significant architectural modifications. In parallel to the economic prosperity, increase in population and the urban growth in the city in this period, the houses were transformed in terms of function and appearance by means of reorganizing circulation patterns, (re)defining the hierarchy between spaces, and the introduction of mosaic pavements, wall paintings, and running water. The courtyard remained as the main and central element of the houses in Zeugma until the destruction of the city in the mid-3rd c. AD, but its function and impression changed progressively. The spatial transformation in some houses attested to the increasing significance of reception rooms to the expense of the peristyle courtyards in that period. In the House of Poseidon, for instance, the western portico of peristyle courtyard A6/P9 was removed by the construction of a row of rooms, one of which was probably a reception room (fig. 144). Similarly, in the House of Synaristosai, the peristyle courtyard of the earlier phase of occupation was replaced with the large triclinium P17 (fig. 185). At the same
98 Wallace-Hadrill drew attention to the gardens of Pompeian houses which had a portico on one side, typically on the side nearest the entrance and tablinum (1994: 84). He argued that these porticos might have framed the view from the atrium between a pair of columns and thereby give the impression of a full peristyle beyond.
304
time, the preserved peristyle courtyards as well as some smaller secondary courts received shallow basins with mosaic pavements, mostly with representations of marine themes, and in some cases with central dispositions like water jets and statues. Monumental and decorated fountains were usually added on one side of the shallow basins. The triclinia of the houses of Zeugma received the most sumptuous decoration with figural mosaic pavements, and wall paintings in this period. It is noteworthy that, none of the triclinia in the city that have been studied here, had an axial placement in reference to the peristyle courtyards or an axial opening towards them. They were often located on the façade and were preceded by inner courts with decorated water features, or by elaborate anterooms. Thanks to the sloped topography and malleable bedrock formation, carved rooms were often provided in the houses as well. Some of these rooms, too, came into prominence with their distinct location, layout, and decoration and seem to have provided a stage for reception ceremonies probably for more intimate guests and/or clients.
In Palmyra the House(s) of Achilles and Cassiopeia and the House to the South-east of the Theater, both of which date to the late 2nd or the early 3rd c AD, exhibited a more elaborate layout and decoration than the earlier dated houses excavated in the city. Both houses were organized around peristyle courtyards in which one of the porticos was wider than the other three (figs. 221, 224). Behind the more spacious portico was the largest and the axially located room and its annexes that probably constituted the reception suite. It is noteworthy that, in both of the houses the reception rooms were annexed with secondary courtyards that were relatively secluded and smaller but rather sumptuous with two or three columns on each side and, as was the case in court 17 of the House(s) of Achilles and Cassiopeia, with mosaic pavements representing scenes from Greco-Roman iconography.
House F in Palmyra displayed a different architectural design approach from its contemporaries. Its initial construction was dated to the second half of the 2nd c. AD, when the city was a Roman tributary, and the building continued to be inhabited throughout the Roman period without much architectural change. It was a spacious
305
residence of about 2,000 m2 and probably belonged to a well-to-do and crowded household. The plan layout of the house was characterized with a strict division of the building into two sections, each having its own entrance, which was probably done to separate clearly the more private wing reserved for the family members and the public section that comprised reception suits and service spaces (figs.215, 216). The five courtyards of the building were either paved with stone slabs or left unpaved and all but one were colonnaded. The colonnades, nevertheless, do not seem to have provided any direct visual or physical reference point for the location and layout of the rooms. Behind each portico was organized a suite of two or three rooms. Some of these rooms were distinguished with their slightly large size and walls that were decorated with friezes, painted plasters, and niches. Hence, House F with its divided plan scheme, relatively modest layout with small and irregular rooms, and lack of right angles in many walls seem to have responded to the accustomed local practices of dwelling and the associated architectural applications rather than adjusting its spatiality to a Greco-Roman lifestyle and architecture.
The unearthed houses in Dura-Europos, like House F in Palmyra, reflected a more local Mesopotamian influence rather than the Greco-Roman one in terms of appearance and function. They were constructed during the Parthian rule and survived in their more or less initial form under Roman rule for at least a century, until the mid-3rd c. AD when the city was abandoned. The most characteristic and fundamental element of domestic architecture in Dura-Europos was the central courtyard around which the other rooms were organized. Larger houses had several courtyards that determined separate units, sometimes with their own entrances (figs. 270, 278). The courtyards of these houses were often left unpaved with no columns. The ovens, some movable equipment like mills and mortars and the underground cisterns/cesspools found in these areas indicated that the courtyards served as the functional core of the house where domestic tasks from cooking, food processing and storage to washing took place. The flat roofs, which were accessed via simple staircases from the courtyards were probably used also for domestic activities, as is the case in the Mesopotamian region today. As attested by the papyrus PDura 19 even the modest
306
houses in Dura-Europos could have been shared by several families and the courtyards served as their common space. In some spacious houses of Dura-Europos, like the House of the Large Atrium, the House of Lysias, and Houses A and C in insula B2, the courtyards were colonnaded and/or paved with brick. None of these were peristyle courtyards but they had columns on one, two or at most three sides. The columns were constructed in rubble stone and plaster and were not always set out in a regular or symmetrical order (figs. 264, 266). Nor, a direct visual or physical connection of the colonnades with the reception rooms, such as axiality or symmetry, seem to have been aimed for in the architectural design. The reception rooms in the houses of Dura-Europos came into prominence with their size, location, and layout but they differed from the contemporary examples in the other cities in some respects. Their entrance was located on the long side and was monumentalized with wide doors and a few steps, in a few more elaborate houses with thick façade walls, triple doors, or free-standing pillars (fig. 273). The interior space of these elevated rooms were surrounded by benches built with plaster or plastered rubble which was about 1.00 m wide and 0.10-0.20 m high and their floors were mostly left as beaten earth or plastered . In some houses, reception rooms were decorated with a molded plaster cornice, friezes, painted plasters, and profiles applied on the door frames which sustained some of the decorative elements of Greek architecture.
The houses that were built or continued to be occupied after the 3rd c. AD once more attested a transformation in the domestic contexts of the Roman east, in parallel to the changing dynamics of the period. The era that roughly covered the period between the 3th and 8th centuries and is commonly described as late antiquity was an epoch of dynamic change and transformation for the Roman Empire, regarding the interrelated contexts of social, political, urban and architectural spheres. Administrative changes, increasing outer threats and military reorganizations, changing economic conditions, the reorientation of public and private revenues and the rise of Christianity were the major factors that had significant impact on the fate of the Roman cities in both the eastern and the western empire. Despite a series of catastrophic events including several earthquakes, an outbreak of bubonic plague, big fires, several revolts, and
307
Sasanian invasions in the 4th, 5th, and 6th centuries, many eastern cities managed to survive and continued to prosper to a certain extent, especially thanks to the intervention of late antique emperors. Urban architecture, nevertheless, lost its previous importance as a stage for elite competition, especially after the 5th c. AD though the power became more associated with individuals and the rivalry between the autocratic elites grew greater (Ellis, 2000: 134). As a consequence, in late antiquity, the houses as seats of power and prestige came to play a more central role in the social and political life of the Roman elite, and accordingly, they were architecturally transformed to provide the suitable setting for the changing domestic patterns, rituals and tastes of the period (Scott, 1997: 59; Uytterhoeven, 2022).
Among the sample examined in this study, the houses in Apamea on the Orontes and the relatively poorly preserved houses excavated in Antioch on the Orontes provided the most suggestive data on the domestic contexts of late antiquity. This sample primarily attested to the changing nature of reception ceremonies and spaces as well as the introduction of new forms, materials, and themes in the architectural layout and decoration of the houses from the 3rd c. AD onwards.
The most significant aspect observed in the architectural layout of the houses from the 3rd c. AD onwards was the gradual disappearance of the triclinium. Archaeological evidence attested that the traditional layout of a triclinium with three rectilinear couches and a central table was initially replaced with a semicircular couch known as stibadium and a semi-circular dining table called sigma (figs. 103, 106). Accordingly, the common layout of the triclinium with T or U-shaped-arranged mosaic pavements gradually disappeared. The apsidal dining room in the House of the Buffet Supper in Antioch on the Orontes, the sigma tables in the large rectangular room A and the apsidal room F in the House of Deer in Apamea on the Orontes, and the remains of food such as peach stones, eggshells, bones of birds and fish found in apsidal room 2 in the Roman Palace in Dura-Europos indicated that these rooms were used, at least occasionally, for dining (figs. 105, 31, 296). The mosaic floor pavement of the dining room in the House of the Buffet Supper was dated to the 3rd c. AD and its semi-circular
308
pattern depicting a very large sigma table set with refined dishes followed the apsidal form of the room. This mosaic pavement remained in-situ when the apse of the rooms was omitted and the room was enlarged at the end of the 3rd c. AD. Similarly, the 4th or 5th c. AD dated mosaic pavement of room F in the House of Deer presented a purely geometric decoration that was coherent with the apsidal form. The rectangular room A, which was the largest and obviously the most prominent room of the same house, was paved with a geometric mosaic that was enriched by bird motifs and was dated to the 6th c. AD. The only triclinium with the standard-setting of a T-shaped mosaic floor in Apamea on the Orontes was that of the House with the Bilobed Column that was dated to the 3rd c. AD. Nevertheless, the two column bases and a pillar that were placed directly on that mosaic floor indicated that the layout and even the function of the room could have been changed in a later phase of occupation (fig. 109).
The mosaic pavements of the reception rooms in two later houses, the House of Ge and the Seasons and the Constantinian Villa, that were partially exposed in Daphne presented a very different layout than the earlier triclinia. In the House of Ge and the Seasons the main room was an apsidal one with a nymphaeum in an earlier phase the date of which could not be determined (fig. 56). In the late 5th century AD the room was enlarged and received a new mosaic pavement that had a central layout with five medallions arranged within a geometric pattern (fig. 55). The central motif representing Ge was laid to be viewed from the entrance while the other four representing seasons were oriented to the center. The so-called Constantinian Villa partially excavated in the same city was dated to the second quarter of the 4th c. AD. The rich and complex mosaic floor of the main room of that building was divided into two panels, one square and one rectangular, reminiscent of the arrangement of some triclinia found in other houses though it differed in its organization (figs. 53, 54). The square panel placed in the front part of the room was divided into four triangles with figural scenes that were separated by four standing figures representing four seasons at the diagonals. It was surrounded by a series of small trapezoidal figural panels with hunting scenes. At the center of this portion of the mosaic floor was placed an octagonal fountain. The rectangular panel at the back of the room had alternating
309
geometric motifs and medallions with busts and figures. The layout of this mosaic panel could recall the T-shaped arrangement of the mosaic panels in the triclinia of the former houses as well as the two-partite arrangement of the reception rooms that will be mentioned below.
More or less contemporary with the disappearance of the common layout of the triclinium a new form of reception space, indicating a transformation in the format of the reception ceremonies, emerged in the urban domestic architecture of the Roman East. The usually largest and most elaborate rooms in many houses that were built or reconstructed in late antiquity displayed a two-partite space organization. In this layout the main space was terminated with an alcove or an apse that was differentiated with its architectural and decorative vocabulary. The earliest example of this kind of arrangement was that of rooms 16-17 in the C300 complex in Daphne (figs. 29, 34). Dated to the late 3rd c. AD, more or less contemporary with the restoration of the triclinium of the House of the Buffet Supper in the same insula, the unearthed part of this complex was likely to have constituted a smaller scale reception suite. In that part of the house, a mosaic-paved courtyard gave way to square room 21 through a portico with two columns. This room measuring about 12 m2 was paved with a good quality mosaic in which a geometric panel was framed by a figurative border depicting birds. It opened fully to a smaller and elevated space that had a lavish geometric mosaic representing cubes in perspective. A column base found in-situ against the south wall indicated a monumental passage between these two spaces. In the 5th c. AD, shortly after the House of the Buffet Supper and C300 complex were abandoned, the same insula was occupied by a new dwelling named Building B or the House with Floral Pavements. This house was organized around a large colonnaded courtyard, on the north-east side of which was placed the largest room of the house that measured about 11.50 m by 15.00 m (fig. 35). This room displayed a similar layout with earlier room 16-17 of C300 in a more monumental scale. The main space of the room (room 12) opened to an alcove (room 13) at its end via a wide arch. The floor of the main space was paved with marble and the floor of the alcove was paved with opus sectile. Across
310
that room on the opposite side of the courtyard was a monumental nymphaeum that was centered on the same axis with the room.
The two-partite spatial layout was common in the vast reception rooms of the residential buildings in Apamea on the Orontes, too. The houses of Apamea, like some public monuments in the city, underwent large-scale and remarkable refurbishment works in the 6th c. AD, probably after the earthquakes of 526 and 528 AD. During these comprehensive works not only the destroyed parts of the buildings were restored but also the domestic spaces were modified in their layout and decoration. As attested by the remains of the excavated houses, the reception spaces in the domestic context were more monumentalized with the addition of alcoves or apses to the end of the rooms. The alcoves and apses were often raised with a few steps and visually separated by means of a wide arch between two columns. Room A-Aˈ, the most distinguished room of the House of the Consoles, constitutes a good sample for this type of arrangement (fig. 81). It was the largest space of the entire structure and was located centrally on the axis of the peristyle to which it opened via three bays. Nearly one-third of this vast room (A') was separated from the rest (A) with a semicircular arch resting on columns. The division between the two sections was also marked on the ground by a low step that raised the back part. As such, room A' became a kind of alcove that was lit by tall windows opening to the street. During its restoration, the splendor of room A-A' was also increased with the renewal of the floor and wall decorations. In this phase, the earlier mosaic pavement of the room was covered with a sumptuous opus sectile pavement which presented a much richer repertoire in alcove A' (fig. 85). The walls of the room were also decorated with alternating opus sectile and painted panels (fig. 86). In addition, room A received a central rectangular fountain probably in the same period. Room E-F in the House of the Console Capitals and Room A in the House of the Pilasters exhibited a similar layout with room A-A' with their elaborate dimensions, two-partite organization, and ornamentation (figs. 91, 94).
In the well-to-do houses of late antiquity, it was also common to accentuate the spatial and visual quality of the distinguished rooms with the addition of apses. Especially
311
from the 3rd c. AD onwards, as attested by archaeological evidence, the reception rooms, just like the ones reserved for banqueting as mentioned above, were built or redesigned in apsidal forms. The apses were often differentiated from the main space with their more elaborate decoration and slightly raised floor levels and were presumably intended to create a similar spatial effect with that of the rectangular alcoves. The Building with Triclinos, the initial construction of which was dated to the 4th c. AD, was one of the most spacious and elaborate houses in Apamea on the Orontes and its plan layout was distinguished with the existence of several apsidal rooms placed around the peristyle courtyard (figs. 70). The largest room of the building, room AB, measuring about 160 m2, was an apsidal room that had a monumental entrance with triple doors and several annex rooms. Room AB was richly decorated with a spectacular mosaic pavement and crustae wall decoration and it was most probably the main reception room of the house in the initial phase of occupation (fig. 77). In the 6th c. AD, presumably after the violent fire that destroyed the northern section of the house, an extensive restoration was undertaken that led to a shift in the spatial focus of the structure from the northern section to the eastern one and also a modification in the main circulation axis accordingly. During this restoration work apsidal room A that was located at a central position to the north of the peristyle courtyard and was slightly smaller than room AB came into prominence. Probably at the same time a new street door on the same axis with Room A across the courtyard was opened. The floor of the main space of apsidal room A was paved with a splendid mosaic depicting a hunting scene and the slightly raised floor of its apse was paved with a simpler geometric mosaic. The third distinct room of the same house was room B that had larger dimensions and an apsidal layout, too.
A new type of apsidal room appeared in the domestic architecture of Apamea on the Orontes towards the very end of antiquity, probably sometime after the modifications of the 5th and 6th centuries but before the latest state of occupation. In some of the spacious residences like the House of the Consoles, the House of the Pilasters, and the Building with Triclinos the vast peristyle courtyards were rearranged by blocking the intercolumniations in one of the galleries of the peristyle and by adding an apse to one
312
end of this space (figs. 80, 87, 88, 70). As such elongated apsidal rooms that were sometimes adorned with mosaic pavements were created. In the House with Floral Pavements in Antioch on the Orontes, whose construction was dated to 5th c. AD, room 7 built on one side of the central courtyard displayed this scheme with its elongated apsidal form (fig. 35). The door of such rooms, was often located in close proximity to the main entrance door of the house, presumably for the entry of the visitors without penetrating into the core of the house, whereas a second door at the center or near the apse was most probably used by the master of the house. This type of elongated apsidal rooms usually in the peripheral locations were common in the late antique urban houses excavated in other provinces such as Asia Minor and Cyrenaica. Usually called audience halls, these rooms most probably served as public spaces within the private sphere where the owner of the house received his clients in a spatial setting that recalled the basilical space organization of the churches the number of which was increasing in late antiquity.
The elaborate reception rooms of late antiquity that were spatially accentuated and divided into two by the addition of alcoves or apses recall both eastern and western influences. The raised and often lavishly decorated apses and alcoves, highlighted at the end of the large rectangular rooms, reminiscent of the iwan in Islamic architecture may indicate an eastern influence. At the same time, they remind the similarly functioned tablina that was a common space in the early Roman period houses unearthed in the Campania region. The often slightly elevated tablina in those houses adjoined the atrium as a stage, some of which were closed only on two sides. In the more elaborate setting of the reception spaces of late antiquity, the master of the house could receive his guests in a more formal, ceremonial, and autocratic manner by distinguishing and separating him from his audience in the highlighted spatial arrangement of the apse or alcove.
The eastern influence is more obviously noticed in the architectural vocabulary of the Yakto Complex in Daphne, where at least two exedrae/iwans opened to a large colonnaded courtyard (fig. 58). One of these, exedra 8, originally belonged to the first
313
building phase and continued to be used in the 5th c. AD. It measured about 5.50 m by 3.70 m and as indicated by the remains its walls and/or the semi-dome of was covered by a sumptuous mosaic with a golden background. The second exedra (47) did not belong to the disposition of the first building but was added to the structure in the 5th c. AD, to the detriment of some earlier spaces. This later exedra was larger than the first one and measured approximately 6.00 m wide and 4.30 m deep. Its floor level was elevated than the portico it opened to. Several fragments of a mosaic decoration similar to those found in exedra 8 were revealed in this room too. The two cruciform rooms, rooms 10 and 52, in the Yakto complex also recall a foreign influence since the use of that form was unique among the sample studied so far. In the northern section of the complex, nevertheless, the rooms that were organized around a second colonnaded court yielded sumptuous mosaic pavements representing scenes from the Greco-Roman iconography.
In addition to the changing architectural vocabulary and the spatial quality, the decoration schemes in the domestic architecture of late antiquity were also modified. From the 4th c. onwards, probably linked to the diminishing role of the classical pagan religion in the face of Christianity, the mythological repertory was no more the main source of the figural compositions in the mosaic panels. Instead, new themes and designs, like repetitive geometric or vegetal patterns, animal figures, and a themes like seasons and the scenes of hunting became widespread in the decoration of the houses. The mosaics depicting scenes of hunting, an activity associated with the lifestyle of the elite, was a common theme in the rich houses in Asia Minor and Africa, too. Ellis and Dunbabin linked this theme to an increasing concentration of the power of the individual and a desire to promote the self-image of a wealthy Roman by displaying visibly the possession of estates, hounds, horses, and servants and, even more important the means of leisure necessary to engage in such activities, as well as, the heroic courage required for confronting the wild animals (Ellis, 2000: 133; Ellis, 1997: 51; Dunbabin, 2003: 141, 146).
314
In the 5th and 6th centuries AD, marble that was already used as a common architectural and decorative material in the monumental public architecture of the Roman cities became widespread in the decoration of the houses. In the earlier centuries, probably because of the high costs of the material, marble revetments were often imitated in the wall paintings of domestic architecture, as was a common practice in the excavated houses at Zeugma. From the 5th c. AD onwards, nevertheless, in parallel to the relative prosperity in the eastern cities, especially in Antioch on the Orontes and Apamea on the Orontes, the use of marble became widespread in the decoration of the houses, especially in the form of opus sectile floor pavements and wall revetments. During the large-scale restoration works of the 6th c. AD in Apamea on the Orontes, the mosaic floors of some distinct rooms were covered with marble or opus sectile pavements as was also the case in the portico of the Grand Colonnade and in some churches in the same city. In Antioch on the Orontes, too, the earlier mosaic pavements in the House of the Floral Pavements and the Yakto Complex in the suburb of Daphne opus sectile pavements were either replaced or used together with the earlier mosaic pavements in the 5th and 6th centuries AD.
Towards the end of antiquity, the weakening of Roman power and influence and the subsequent Arab conquests gave rise to radical changes in the social and economic dynamics and accordingly had a significant impact on the urban and architectural spheres in the eastern periphery of the empire. In this period, the cities of the eastern province of Syria witnessed a radical transformation towards ruralization in both the urban and the domestic contexts. Archaeological evidence attested that this transformation began in the first decades of the 7th c. AD, in the capital city Apamea on the Orontes, whereas it started much earlier, during the 4th or 5th centuries AD, in smaller cities like Zeugma and Palmyra. The nature and extent of the archaeological data provided so far by the excavations in Antioch on the Orontes are not sufficient to follow subsequent occupational phases but the city presumably underwent the same transformation process more or less contemporary with Apamea on the Orontes. Dura-Europos, on the other hand, was totally abandoned after the Sasanian sack in the mid-3rd c. AD and was never reoccupied.
315
In Zeugma some of the buildings that were ruined during the Sasanian sack in the mid-3rd c. AD were slowly buried under colluvium coming from the slopes of Belkıs Tepe or encroached by domestic structures, as was the case in the theatre probably at the end of the 3rd c. or the beginning of the 4th c. AD. By the 5th c. AD, the newly built houses were often smaller and simpler for which, spolia was used or the earlier houses were incorporated. The archaeological documentation suggests that there was a more pastoral life in Zeugma in late antiquity. Tandoor ovens built in a very simple style, used for bread making were found in almost all the houses in this period. Spindle whorls and ivory needles used for spinning wool were usually found in contexts also dating to late antiquity. The lavishly decorated rock-cut rooms of some 2nd and early 3rd century dated houses, in addition, were turned into stables. In Palmyra, the encroachment of the existing public monuments with poor residential buildings and workshops started as early as the end of the 4th or the beginning of the 5th c. AD as the evidence that came from the Temple of Baalshamin, the Temple of Nabu and the annexes of the agora suggested (Intagliata, 2018: 37-41). The elaborate urban residences of the well-to-do families, on the other hand, gradually turned into smaller, poor-quality and multi-functional units with rural settings. Larger houses, as in the case of House F, were split into smaller apartments to accommodate more families after the earthquake of late 6th c. AD. Upper floors were abandoned, small shops were added to the facades, larger spaces were subdivided into smaller units and some spaces turned into stables and into workshops for production and storage activities. The excavations conducted so far in Apamea on the Orontes provided the most sufficient data for the transformation of urban residences in the 7th c. AD. The urban residences of the wealthy elite in the provincial capital city were divided into smaller units and gradually transformed into poor-quality habitats with a rural character. Courtyards continued to be the spatial and functional core of the houses that were shared by multiple units. The galleries of the peristyles as well as the larger rooms were blocked and divided into smaller rooms. The circulation patterns were redefined with the blocking of many doors and windows and sometimes with the opening of new ones. Many of the spaces received artisanal and commercial functions with the installation of presses, millstones, and cisterns. Wells were dug in random places while many of
316
the earlier reservoirs and fountains went out of use. Several troughs and hooks indicate the existence of animals kept within the dwelling units.
The small, poor quality, and densely occupied dwellings of the 7th c. AD in Apamea on the Orontes defined a setting for a very disparate lifestyle from that of the earlier spacious and splendid urban residences. Such modest dwellings seem to have been inhabited by a different social class living a rural life. It is very likely that, because of the increasing threats of the Arabs and the weakening of the imperial power in the region, the wealthy aristocrats in the cities of Syria retreated to the capital city of Constantinople or the provinces of Asia Minor mainly for security and economic concerns. The new inhabitants, as Balty argued, could have been the refugees from the countryside who left their villages and settled in the residences of the aristocratic landowners (1984: 498-501). A similar migration had already happened in some of the cities of Syria in the mid-3rd c. AD, due to the Sasanian threats. In Zeugma, archaeological evidence has reflected the anxiety and concern for a possible attack in the domestic lieu in the first half of the 3rd c. AD. During that period, the houses were modified in terms of accessibility and functionality. Existing houses were subdivided by blocking doors, some lavishly decorated rooms were given over to storage and food processing, latrines and fountains in some houses went out of use and some walls with painted plaster were covered by graffiti. The inhabitants of Zeugma presumably managed to escape before the sack, probably with the hope to return back to their houses, as several finds including hoards of coins, abandoned furnishings and household objects indicate. Nevertheless, most of these houses were not reoccupied, at least for two centuries, after the sack. Similarly, Dura-Europos was totally abandoned in the mid-3rd c. AD and the inhabitants of the city never turned back to their homes.
From the early 8th c. AD onwards, in parallel to the political, economic and urban contexts of the Islamic empires, a progressive withdrawal from social and economic
317
activities in search of privacy was observed in the domestic context.99 The early Islamic period witnessed a growing separation between domestic and professional life. Literary sources and archaeological evidence attested that the courtyard had lost its social and economic functions and began to be used for utilitarian purposes like storage, cooking, washing and keeping animals.100 The houses, respectively, were planned to respond to the increased concern for the intimacy of the family. The living quarters of the family were designed on the upper floors whereas the ground floors reserved to utilitarian functions were surrounded with blind doors and small door openings.
The chronological and comparative overview of the domestic architecture of provincial Syria provides new perspectives and enriched framework to discuss the social and architectural contexts of the ‘Roman house’. The term ‘Roman’ simultaneously defined a time span, a geographical location, a personal or communal identity and a form of material culture (Revell, 2008: x). While the geography and time span were definite, the ‘Roman identity’ and the material culture of the ‘Roman’ societies did not form a single, fixed and stable entity, and hence were not clearly and sharply definable. The acceptance of a type or form of material culture from a specific geography and time period as ‘Roman’ is, therefore, misleading as attested by the earlier approaches that associated the ‘Roman house’ almost exclusively to the atrium based on the evidence coming from the Campania region. As put forward by Simon Ellis “the essence of the Roman world was its way of life more than anything else” (1988: 565). The houses of the Roman period, in this respect, are important contexts that throw light to the way of life of their inhabitants, as they constitute the social and spatial framework in which the daily routines and rituals were both structured and represented.
99 The Islamic period is out of scope of this study and thus the transformation in the domestic context after late antiquity is only briefly mentioned. For more information on the economic contexts of the period see Bessard, 2020; for the urban transformation in the early Islamic period see Kennedy, 1985; King and Cameron, 1994.
100 See Whitehouse, 2002; Walmsley, 2007; Al-Mulla Hwaish, 2018.
318
The houses of the Roman elite were not only the locus of private life but also of social, political and business activities, a phenomenon discussed extensively.101 The reception of different groups of guests including the clients, peers and friends, for various occasions were among the main social activities that took place in the houses.102 Some of the ritualistic activitites, like the morning salutatio and evening banqueting, necessitated distinct spatial settings like tablinum, triclinium, stibadium or apsidal halls, within the domestic context. In the 1st c. BC, Vitruvius pointed out this phenomenon by emphasizing the necessity of providing public spaces appropriate for the professional requirements of the house owner and the importance of manifesting status and dignity through such spaces (Book VI-5). The domestic spaces of the Roman elite, thus, not only provided the suitable spatial settings for reception activities but also functioned as mediums and environments to display wealth and power to visitors and enhance social status. As Wallace-Hadrill argued what mattered most in the Roman houses was the capacity to receive visitors and impress them through luxury rather than the capacity to house a family (2015: 185).
The owners of the houses examined in this study are not known, except in few cases, but the size of the buildings, their location in the urban fabric and their architectural and decorative features give clues on the socio-economic status and lifestyle of the possessors. Most of the houses studied are large residences that display a certain degree of splendor and luxury. Even the relatively small houses, like the ones excavated in Antioch on the Orontes and Zeugma, came into prominence with their splendid architectural features and fine decoration. It is very likely that these houses were inhabited by relatively well-to-do families. Dura Europos is exceptional in this sense since the domestic context include small dwellings with few rooms some of which were associated with shops or craft activities. These dwellings obviously belonged to the economically less affluent.
101 See supra 66.
102 For different types of interactions and rituals in the Roman domestic context see Hudson, 2010; Dunbabain and Slater, 2011; Verboven, 2011.
319
Though the sample of the houses studied so far display a great variety in their architectural layout and decoration they had some common elements. In this regard, the concern of visibility through architectural grandeur and/or location within the urban fabric, the search for symmetry and axiality, at least in some parts of the plan layout, the elaboration of ‘public’ spaces within the domestic context through architectural layout and decoration, the integration of water elements to enhance the spatial and visual quality of spaces and the use of certain decoration themes and schemes have already been pointed out for many houses. These trends and approaches have already been attested in different cities of the Roman world, in both Italy and the provinces. Similarly, the presence of representative spaces like colonnaded courtyards, triclinia, stibadia, apsidal halls and in some cases also the bathrooms, which were arranged to provide the necessary social settings for domestic activities and rituals, as well as their lack, provide narratives about the life of inhabitants, and indicate, also, to what extent they adopted a Roman way of life. The social and architectural contexts of the houses in Roman Syria displayed varying degrees of ‘eastern’ and ‘western’ influences, and it was mainly through these houses, more than anything else, that the local elites manifested their adoption or ignorance of a ‘Roman identity’.
320
BIBLIOGRAPHY
Abadie-Reynal, C (2001a) “Zeugma Rapport Préliminaire de la Campagnes de Fouilles de 2000”, Anatolia Antiqua IX, pp. 243-305.
Abadie-Reynal, C. (2001b) “Apamée et Séleucie-Zeugma: des Villes Jumelles sur l’Euphrate?”, Histoire Urbaine 3, pp. 7-24.
Abadie-Reynal, C. (2006a) “Roman Domestic Architecture at Zeugma”, in Ergeç, R. (ed.) Uluslararası Geçmişten Geleceğe Zeugma Sempozyumu. Gaziantep: Gaziantep Üniversitesi Matbaası, pp. 1-6.
Abadie-Reynal, C. (2006b) “Zeugma Tiyatrosu Kazısı”, XXVII. Kazı Sonuçları Toplantısı 1, pp. 331-338.
Abadie-Reynal, C. (2008) “Eaux Decoratives, Eux Symboliques à Zeugma”, Syria 85, pp. 99-118.
Abadie-Reynal, C. (2010) “Acculturation et Habitat à Zeugma à l’Époque Romaine”, Bollettino di Archeologia on line I 2010/Volume speciale, pp. 15-24.
Abadie-Reynal, C. (2012a) “Fouilles de l’Habitat (2): la Maison des Synaristôsai” in Abadie-Reynal, C. (ed.) Zeugma 3: Fouilles de L’Habitat (2) La Maison des Synaristosai. Nouvelles Inscriptions. Lyon: Maison de l’Orient et de la Méditerranée- Jean Pouilloux, pp. 11-150.
Abadie-Reynal, C. (2012b) “Les Fouilles Françaises dans la Maison de Poséidon (ou Maison 1), la Maison de l’Euphrate (ou Maison 2) et la Maison 4. Années 1999-2000” in Abadie-Reynal, C. and Ergeç, R. (eds.) Zeugma 1: Fouilles de L’Habitat (1) La Mosaique de Pasiphae. İstanbul: Institute Français d’Etudes Anatoliennes- George Dumézil, pp. 183-237.
Abadie-Reynal, C. and Bucak, C. (2000) “Work at Apameia in 1998”, XXI. Kazı Sonuçları Toplantısı 2, pp. 271-282.
Abadie-Reynal, C. and Bucak, E. (2001) “Work at Apameia in 1999”, XXII. Kazı Sonuçları Toplantısı 2, pp. 233-240.
321
Abadie-Reynal, C. and Darmon, J. P. (2003) “La Maison et la Mosaique des Synaristosai” in Humphrey J. H. (ed.) Zeugma: Interim Reports, Rescue Excavations. JRA Supplementary Series 51, pp. 79-99.
Abadie-Reynal, C. and Ergeç, R. (1998) “Zeugma- Moyenne Vallee de l’Euphrate Rapport Preliminaire de la Campagne de Fouilles de 1997”, Anatolia Antiqua VI, pp. 379-406.
Abadie-Reynal, C. and Ergeç, R. (1999) “The Zeugma and Apameia Works 1997”, XX. Kazı Sonuçları Toplantısı 2, pp. 403-416.
Abadie-Reynal, C. et al. (1996) “Mission Archeologique de Zeugma. Rapport sur la Campagne de Prospection 1995”, Anatolia Antiqua IV, pp. 311-324.
Abadie-Reynal, C. et al. (1997) “Mission de Zeugma- Moyenne Vallee de l’Euphrate”, Anatolia Antiqua V, pp. 349-370.
Abadie-Reynal, C., Bucak, E. and Bulgan, F. (2000) “Zeugma- Moyenne Vallee de l’Euphrate Rapport Preliminaire de la Campagne de Fouilles de 1999”, Anatolia Antiqua VIII, pp. 279-337.
Abadie-Reynal, C., Ergeç, R. and Bucak E. (1999) “Zeugma- Moyenne Vallee de l’Euphrate Rapport Preliminaire de la Campagne de Fouilles de 1998”, Anatolia Antiqua VII, pp. 311-366.
Abbasoğlu, H. (2001) “The Founding of Perge and its Development in the Hellenistic and Roman Periods”, in Parrish, D. (ed.) Urbanism in Western Asia Minor: New Studies on Aphrodisias, Ephesus, Hierapolis, Pergamon, Perge and Xanthos. JRA Supplementary Series 45, pp. 172-188.
Acar, Ö. (2000) “Yıllardır Devam Eden Zeugma Trajedisi”, Arkeoloji ve Sanat 98, pp. 6-10.
Ainsworth, W. F. (1842) Travels and Researches in Asia Minor, Mesopotamia, Chaldea and Armenia, vol. I-II. London: John W Parker,West Strand.
Akurgal, E. (1993) Ancient Civilizations and Ruins of Turkey. Istanbul: Net.
322
Al As’ad K., Schmidt-Colinet, A. (2000) “Zur Urbanistik des Hellenistischen Palmyra. Ein Vorbericht”, Damaszener Mitteilungen 12, pp. 61-93.
Algaze, G. et al. (1991) “The Tigris Euphrates Archeological Reconnaissance Project: Preliminary Report of the 1989-1990 Seasons”, Anatolica 17, pp. 175-240.
Algaze, G., Breuninger R. and Knudstad J. (1994) “The Tigris Euphrates Archaeological Reconnaissance Project: Final report of the Birecik and Carchemish Dam Survey Areas”, Anatolica 20, pp. 1-96.
Allara, A. F. (1986) “Les maisons de Doura-Europos. Questions de Typologie”, Syria 63, pp. 39-60.
Allara, A. F. (1988) “Les Maisons de Doura-Europos. Les Données du Terrain”, Syria 65, pp. 323-342.
Allara, A. F. (1992) "L'îlot des Potiers et les Fours à Doura-Europos" Syria 69, pp. 101-120.
Allara, A. F. (1997) “Entre Archives et Terrain. L’Exemple d’un Îlot d’Habitation de Doura-Europos” in Castel, C. et al. (eds.) Les Maisons dans la Syrie antique du IIIe Millénaire aux débuts de l'Islam. Pratiques et représentations de l'espace domestique. Beirut: Institute Français du Proche Orient, pp. 173-186.
Allara, A. F. (2002) Problemi di Architettura Domestica a Dura-Europos sull’Eufrate. L’Isolato dei Vasali (B2). Napoli: Istituto Universitario Orientale di Napoli.
Allara, A. F. and Saliou, C. (1997) “Constitution d’une Repertoire de l’Architecture Domestique a Doura-Europos” in Leriche, P. and Gelin, M. (eds.) Doura Europos Etudes IV 1991-1993. Beirut: IFAPO, pp. 145-154.
Allison, P. (1993) “How Do We Identify the Use of Space in Roman Housing?” in Moorman, E. M. (ed.) Functional and Spatial Analysis of Wall Painting. Proceedings of the Fifth International Congress on Wall Painting. Leiden: Babesch, pp. 1-8.
Allison, P. (2001) “Using the Material and Written Sources: Turn of the Millenium Approaches to Roman Domestic Space”, American Journal of Archaeology 105, pp. 181-208.
323
Allison, P. (2004) Pompeian Households. An Analysis of the Material Culture. Los Angeles: Costen Institute of Archaeology.
Allison, P. (2007) “Engendering Roman Domestic Space”, British School at Athens Studies 15, pp. 343–350.
Al-Maqdissi (2000) “Annexe. Première Campagne 1965 et Deuxième Campagne 1967”, Syria 77, pp. 155-158.
Al-Mulla Hwaish, A. N. (2016) “Concept of Islamic House: A Case Study for Early Muslims Traditional House”, International Journal of Advances in Mechanical and Civil Engineering 3, pp. 18-25.
Alpi, F. (2007) “Maison Païenne et Chrétienne: Représentation de l’Habitat Domestique à Antioche chez Quelques Auteurs Tardo-Antiques (IVe- Vie s.)” in Galor, K. and Waliszewski, T. (eds.) From Antioch to Alexandria, Recent Studies in Domestic Architecture, Warsaw: Institute of Archaeology, pp. 37-49.
Alston, R. and Lieu, S. (eds.) (2007) Aspects of the Roman East. Vol. I. Turnhout: Brepols.
Andrade, N. J. (2013) Syrian Identity in the Greco-Roman World. New York: Cambridge University Press.
Andrae, W. and Lenzen, H. (1933) Die Partherstadt Assur. Leipzig: WVDOG.
Anguissola A. (2013) Privata Luxuria: Towards and Archaeology of Intimacy: Pompeii and Beyond. Münchner Studien zur Alten Welt. Münich: Herbert Utz Verlag.
Ault, B. A. and Nevett, L. C. (2005) Ancient Greek House and Households: Chronological, Regional and Social Diversity. Philadelphia: University of Pennsylvania Press.
Aylward, W. (2013a) “The Rescue Excavations at Zeugma in 2000” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute, pp. 1-54.
324
Aylward, W. (2013b) “A Monumental Building in Trench 15” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute, pp. 119-123.
Aylward, W. (2013c) “The Rescue Excavations at Zeugma in 2000” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute, pp. 1-54.
Aylward, W. (ed.) (2013) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute.
Baier, C. (2013) “Attolitur monte Pione. Neue Untersuchungen im Stadtviertel oberhalb des Theaters von Ephesos”,ÖJh 82, pp. 23-68.
Baird, J. A. (2006) Housing and Households at Dura-Europos: A Study in Identity on Rome’s Eastern Frontier. Unpublished Phd. Thesis. Leicester: University of Leicester.
Baird, J. A. (2012) “L’Habitat d’Époque Romaine à Europos- Doura: Replacer Les Objets en Contexte”, in Leriche, P., Coqueugniot, G. and Pontbriand, S. (eds.) Europos- Doura Varia I. Beyrouth: Ifpo, pp. 231- 239.
Baird, J. A. (2014) The Inner Lives of Ancient Houses, An Archaeology of Dura-Europos. Oxford: Oxford University Press.
Ball, W. (2000) Rome in the East. The Transformation of an Empire. London and New York: Routledge.
Balty, J. (1969) La Grande Mosaïque de Chasse Du Triclinos. Bruxelles: Centre Belge de Recherches Archéologiques à Apamée de Syrie.
Balty, J. (1977) Mosaiques Antiques de Syrie. Brussels: Centre Belge de Recherches Archéologiques à Apamée de Syrie.
Balty, J. (1984) “La Maison aux Consoles” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 59-78.
325
Balty, J. (1988) “Apamea in Syria in the Second and Third Centuries A.D.”, Journal of Roman Studies 78, pp. 91–104.
Balty, J. and Balty, J. Ch. (1969) “Le Cadre Topographique et Historique” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1965-1968, Actes du Colloque Tenu à Bruxelles les 29 et 30 Avril 1969. Brussels, pp. 29-51.
Balty, J. Ch. (1969) “L’Édifice dit ‘Au Triclinos’” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1965-1968, Actes du Colloque Tenu à Bruxelles les 29 et 30 Avril 1969. Brussels, pp. 105-116.
Balty, J. Ch. (1972) “Apamée, 1969-1971” in Balty, J. C. and Balty, J. (eds) Apamée de Syrie, Bilan des Recherches Archéologiques 1969-1971, Actes du Colloque Tenu à Bruxelles les 15,17 et 18 Avril 1972. Brussels, pp. 15-32.
Balty, J. Ch. (1984a) “Introduction” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 11-16.
Balty, J. Ch. (1984b) “Notes sur l’Habitat Romain, Byzantin et Arabe d’Apamée Rapport de Synthèse” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 471-503.
Balty, J. Ch. (1989) “Apamée au VIe Siècle. Témoignages Archéologiques de la Richesse d’une Ville”, in Abadie-Reynal, C. (ed.) Hommes et Richesses dans l’Empire Byzantine, IV-VIIe Siècle. Paris: Lethielleux, pp. 79-96.
Balty, J. Ch. (1991) “Apamée et la Syrie du Nord aux Époques Hellénistique et Romaine”, Revue du Monde Musulman et de la Méditerranée 62, pp. 15-26.
Balty, J. Ch. (1997) “Palais et Maisons d’Apamée” in Castel, C., Al-Maqdissi, M. and Villeneuve, F. (eds) (1997) Les Maisons dans la Syrie Antique du IIIe Millénaire aux Débuts de l’Islam, Pratiques et Représentations de l’Espace Domestique. Actes du Colloque International, Damas 27-30 Juin 1992. Beyrouth: IFAPO, pp. 283-295.
326
Balty, J. Ch. (2000) “Claudia Apamea. Données Nouvelles sur l’Histoire et la Topographie d’Apamée”, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 144e Année, No.1, pp. 459–481.
Baranski, M. (1994) “The Roman Army in Palmyra: A Case of Adaptation of a Pre-existing City”, in Dabrowa, E. (ed.) The Roman and the Byzantine Army in the East. Kraków : Drukarnia Uniwersytetu Jagiellońskiego, pp. 9-17.
Baranski, M. (1995) “The Great Colonnade of Palmyra Reconsidered”, ARAM 7, pp. 37-46.
Baratte, F. (1984) “La Maison des Chapiteaux a` Consoles” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 107-140.
Barbet, A. (ed.) (2005) Zeugma II, Peintures Murales Romaines. Varia Anatolica XVII. İstanbul: De Bocard/IFEA.
Barsanti, C. (2012) “The Fate of the Antioch Mosaic Pavements: Some Reflections”, JMR 5, pp. 25- 42.
Başgelen, N. (2000) Zeugma: From the Air. İstanbul: Archaeology and Art Publications.
Başgelen, N. and Ergeç, R. (2000) Belkıs/Zeugma, Halfeti, Rumkale. A Last Look at History. İstanbul: Archaeology and Art Publications.
Baur, P. V. C. (1933) “Architecture” in Baur, P. V. C., Rostovtzeff, M. I. and Bellinger, A. R. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Fourth Season October 1930- March 1931. New Haven: Yale University Press, pp. 42-55.
Becker, L. and Kondoleon, C. (2005) The Arts of Antioch, Art Historical and Scientific Approaches to Roman Mosaics and a Catalogue of the Worcester Art Museum Antioch Collection. Worcester, MA: Worcester Art Museum.
327
Bek, L. (1983) “Questiones Convivales, the Idea of the Triclinium and the Staging of Convivial Ceremony from Rome to Byzantium”, Analecta Romana Instituti Danici 12, Odense, pp. 81-107.
Bellinger, A. R. (1933) “Coins” in Baur, P. V. C., Rostovtzeff, M. I. and Bellinger, A. R. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Fourth Season October 1930- March 1931. New Haven: Yale University Press, pp. 259-282.
Bellinger, A. R. (ed.) (1949) The Excavations at Dura-Europos: Final Report, vol. VI- The Coins. New Haven: Yale University Press.
Benech, C. (2010) “The Use of ‘Space Syntax’ for the Study of City Planning and Household from Geophysical Maps: The Case of Dura-Europos (Syria)” in Städtisches Wohnen im östlichen Mittelmeerraum 4. Jh. v. Chr.–1. Jh. n. Chr., Oct 2007, Vienne, pp. 403- 416.
Berenfeld, L. M. (2009) “The Triconch House and the Predecessors of the Bishop’s Palace at Aphrodisias”, American Journal of Archaeology 113.2, pp. 203-229.
BerenfeldCampbell, S. (1996) “Signs of Prosperity in the Decoration of some 4th-5th Century Buildings at Aphrodisias” in Rouché, C. (ed.) Aphrodisias in Late Antiquity (Journal of Roman Archaeology Supplement 20), pp. 187-199.
Bernard, P. (1971) “La campagne de fouilles de 1970 à Aï Khanoum (Afghanistan)” Comptes rendus des séances de l'Académie des Inscriptions et Belles-Lettres, 115ᵉ année, N. 2, pp. 385-453.
Bernard, P. (1974) “Fouilles de Aï-Khanoum (Afghanistan), campagnes de 1972 et 1973”, Comptes rendus des séances de l'Académie des Inscriptions et Belles-Lettres, 118ᵉ année, N. 2, pp. 280-308.
Bernard, P. (1976) “Campagne des Fouilles de 1975 à Aï Khanoum (Afghanistan)”, Comptes rendus des séances de l’Académie des Inscriptions et Belles-Lettres, 120e Année, No.2, pp. 287-322.
Berry, J. (1997) “Household Artefacts: Towards a reInterpretation of Roman Domestic Space" in Laurence, R. and Wallace-Hadrill, A. (eds.) Domestic Space in the Roman World: Pompeii and Beyond. JRA Supplementary series 22, pp. 183-196.
328
Bessac, J. - C. (1988) “L’Analyse de Procédés de Construction de Ramparts de Pierre de Doura-Europos. Questions de Méthodologie”, Syria 65, pp. 297-313.
Bessard, F. (2020) Caliphs and Merchants. Cities and economies of the Power in the Near East (700-950). Oxford: Oxford University Press.
Breasted, J. H. (1922) “Peintures d’Epoque Romaine dans le désert de Syrie’, Syria 3, pp. 177-206.
Breasted, J. H. (1924) Oriental Forerunners of Byzantine Painting: First Century Wall Paintings from the Fortress of Dura on the Middle Euphrates. Chicago: Oriental Institute.
Brett, G. (1942) “The Mosaic of the Great Palace in Constantinople”, Journal of the Warburg and Courtauld Institutes 5, pp. 34-43.
Brown, F. E. (1936) “House in Block E4: History, Architecture and Decoration” in Rostovtzeff, M. I. et al. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Sixth Season of Work. October 1932- March 1933. New Haven: Yale University Press, pp. 4-34.
Brown, F. E. (1944a) “The Hellenistic Agora” in Rostovtzeff, M. I. et al. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Ninth Season 1935-36. Part 1: The Agora and Bazaar. New Heaven: Yale University Press, pp. 3-27.
Brown, F. E. (1944b) “The Bazaar Quarter” in Rostovtzeff, M. I. et al. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Ninth Season 1935-36. Part 1: The Agora and Bazaar. New Heaven: Yale University Press, pp. 28- 68.
Brown, F. E. (1944c) “Houses and Shops of the Final Period” in Rostovtzeff, M. I. et al. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Ninth Season 1935-36. Part 1: The Agora and Bazaar. New Heaven: Yale University Press, pp. 69- 158.
Brown, P. (1992) Power and Persuasion in Late Antiquity: Towards a Christian Empire. Madison: University of Wisconsin Press.
329
Bryce, T. (2014) Ancient Syria: A Three Thousand Year History. New York: Oxford University Press.
Burns, T. S. (2017) Origins of the Colonnaded Streets in the Cities of the Roman East. Oxford and New York: Oxford University Press.
Burns, T. S. and Eadie, J. W. (eds.) (2001) Urban Centers and Rural Contexts in Late Antiquity. Michigan: Michigan State University Press.
Butcher, K. (2003) Roman Syria and the Near East. London: British Museum Press.
Cabouret, B. (1999) “Sous les Portiques d’Antioche”, Syria 76, pp. 127-150.
Campbell, S. and Ergeç, R. (1998) “New Mosaics” in Kennedy, D. L. (ed.) The Twin Towns of Zeugma on the Euphrates/Rescue Work and Historical Studies. JRA Supplementary Series No.27, pp. 109-128.
Campbell, W. A. (1934) “A Byzantine Stadium” in Elderkin, G. W. (ed.) Antioch-on-the-Orontes I: The Excavations of 1932. Princeton: Princeton University Press, pp. 32-33.
Clarke, J. R. (1991) The Houses of Roman Italy 100 BC-AD 250: Ritual, Space and Decoration. Berkeley, Los Angeles, Oxford: University of California Press.
Clarke, J. R. (2003) Art in the Lives of Ordinary Romans: Visual Representation and Non-Elite Viewers in Italy, 100 B.C.-A.D. 315. Oxford: University of California Press.
Cohen, G. M. (2006) Hellenistic Settlements in Syria, the Red Sea Basin and North Africa. Berkeley, Los Angeles& London: University of California Press.
Çonkır, E. (2005) Architectural Elaboration of the ‘Public’ in the Domestic Architecture of Late Antique Anatolia: Changes and Transformations in the Private Setting. Unpublished MA Thesis. Ankara: METU.
Cooper, C. F. (2007) “Closely Watched Households: Visibility, Exposure and Private Power in the Roman Domus”, Past & Present 197, pp. 3-33.
330
Crouch, D. P. (1972) “A Note on the Population and Area of Palmyra”, MUSJ 47, pp. 241-250.
Crowther, C. R. (2003) “Inscriptions on Antiochus I of Commagene and Other Epigraphical Finds” in Humphrey J. H. (ed.) Zeugma: Interim Reports, Rescue Excavations. JRA Supplementary Series 51, pp. 57-67.
Crowther, C. R. and M. Facella (2003) “New Evidence for the Ruler Cult of Antiochus of Commagene from Zeugma” in Heedemann, G. and Winter, E. (eds.) Neue Forschungen zur Religionsgeschichte Kleinasiens. Asia Minor Studien 49. Bonn: Rudolf Habelt, pp. 41-80.
Cumont, F. (1917) Etudes Syriennes. Paris: Auguste Picard.
Cumont, F. (1926) Fouilles de Dura-Europos (1922-1923), vol. 1-2. Paris: Geuthner.
Cumont, F. (1934) “The Population of Syria”, Journal of Roman Studies 24, pp. 187-190.
D’Andria, F. (2002) “Hierapolis Antik Kenti 2001 Yılı Kazı ve Onarım Çalışmaları”, 24. Kazı Sonuçları Toplantısı 2, pp. 415-422.
De Giorgi, A. U. (2007) “The Formation of a Roman Landscape: the Case of Antioch”, JRA 20, pp. 283- 298.
De Giorgi, A. U. and Eger, A. A. (2021) Antioch: A History. London and New York: Routledge.
Delia, D. (1988) “The Population of Roman Alexandria”, Transactions of the American Philological Association 118, pp. 275-292.
Delplace, C. (2013) “Les Recherches de la Mission Archéologique Française à Palmyre”, in Gawlikowski, M. and Majcherek, G. (eds.) Studia Palmyrenskie XII. Fifty Years of Excavations in Palmyra 1959-2009. Warsaw: Warsaw University Press, pp. 37- 48.
Detweiler, A. H. (1952) “The Site and the Building” in Rostovtzeff, M. I. et al. (eds) The Excavations at Dura-Europos. Preliminary Report of the Ninth Season of
331
Work, 1935- 1936, Part 3: The Palace of the Dux Ripae and the Dolicheneum. New Haven: Yale University Press, pp. 1- 26.
Dewez, M. (1969a) “Le Cadre Geographique” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1965-1968, Actes du Colloque Tenu à Bruxelles les 29 et 30 Avril 1969. Brussels, pp.23-27.
Dewez, M. (1969b) “Découvertes Préhistorique” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1965-1968, Actes du Colloque Tenu à Bruxelles les 29 et 30 Avril 1969. Brussels, pp.53-60.
Dieudonné-Glad, N. (2006) “Locks and Padlocks of Zeugma” in Ergeç, R. (ed.) Uluslararası Geçmişten Geleceğe Zeugma Sempozyumu. Gaziantep: Gaziantep Üniversitesi Matbaası, pp. 43-48.
Dirven, L. (1999) The Palmyrenes of Dura-Europos: A Study of Religious Interaction in Roman Syria. Leiden, Boston & Köln: Brill.
Dixon, S. (1992) The Roman Family. Baltimore and London: The John Hopkins University Press.
Dobbins, J. J. (2000) “The Houses of Antioch” in Kondoleon, C. (ed.) Antioch: The Lost Ancient City. Princeton, New Jersey: Princeton University Press, pp. 51-62.
Donnay-Rocmans, C. and Donnay, G. (1984) “La Maison du Cerf” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 155-180.
Downey, G. (1958) “The Size of the Population of Antioch”, Transactions and Proceedings of the American Philological Association 89, pp. 84-91.
Downey, G. (1961) A History of Antioch in Syria. From Seleucus to the Arab Conquest. Princeton, New Jersey: Princeton University Press.
Downey, S. B. (1985) “Two Buildings at Dura-Europos and the Early History of the Iwan”, Mesopotamia XX, pp. 111-129.
332
Downey, S. B. (1986) “The Citadel Palace at Dura-Europos”, Syria 63, pp. 27-37.
Downey, S. B. (1988a) “Further Observations on the Citadel Palace of Dura-Europos”, Syria 65, pp. 343-347.
Downey, S. B. (1988b) Mesopotamian Religious Architecture: Alexander through the Parthians. Princeton: Princeton University Press.
Downey, S. B. (1992) “Frank Brown’s Notes on the Citadel Palace at Dura-Europos”, Syria 69, pp. 141-151.
Downey, S. B. (1993) “The Palace of the Dux Ripae at Dura-Europos and ‘Palatial’ Architecture of Late Antiquity”, in Eius Virtutis Studiosi: Classical and Postclassical Studies in Memory of Frank Edward Brown, Studies in the History of Art 43. Washington: National Gallery of Art, pp. 183- 198.
Downey, S. B. (2000) “The Transformation of Seleucid Dura-Europos” in Fentress, E. (ed.) Romanization and the City: Creation, Transformation and Failures. JRA Supplementary Series 38, pp. 155-172.
Dulière, C. (1968) La Mosaique des Amazones. Brussels: Centre Belge de Recherches Archéologiques à Apamée de Syrie.
Dulière, C. (1974) Mosaiques des Portiques de la Grande Colonnade. Brussels: Centre Belge de Recherches Archéologiques à Apamée de Syrie.
Dunbabain, K. M. D. (1991) “Triclinium and Stibadium”, in Slater, W. J. (ed.) Dining in a Classical Context. Ann Arbor: University of Michigan Press. pp. 121-148.
Dunbabin K. M. D. and Slater, W. J. (2011) “Roman Dining”, in Peachin, M. (ed.) The Oxford Handbook of Social Relations in the Roman World. Oxford: Oxford University Press, pp. 438-466.
Dunbabin, K. M. D. (1978) The Mosaics of North Africa. New York: Cambridge University Press.
Dunbabin, K. M. D. (1999) Mosaics of the Greek and Roman World. Cambridge: Cambridge University Press.
333
Dunbabin, K. M. D. (2003) The Roman Banquet: Images of Convivality. Cambridge: Cambridge University Press.
Dunbabin, K. M. D. (2013) “Mosaics” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute, pp. 149-167.
Duval, N. (1984) “Les Maisons d’Apamée et l’Architecture ‘Palatiale’ de l’Antiquité Tardive” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 447-470.
Dwyer, E. (1991) “The Pompeian House in Theory and Practice” in Gazda, E. K. (ed.) Roman Art in the Private Sphere: New Perspectives on the Architecture and Decor of the Domus, Villa and Insula. Michigan: University of Michigan Press, pp. 25-48.
Early, R. (2003) “Rescue Work by the Packard Humanities Institute: Interim Report, 2000” in Humphrey, J. H. (ed.) (2003) Zeugma: Interim Reports, Rescue Excavations. JRA Supplementary series 51, pp. 8-56.
Edwell, P. M. (2008) Between Rome and Persia: The Middle Euphrates, Mesopotamia and Palmyra under Roman Control. London and New York: Routledge.
Ellis, S. P. (1988) “The End of the Roman House” American Journal of Archaeology 92: 4, pp. 565-577.
Ellis, S. P. (1991) “Power, Architecture and Décor: How the Late Roman Aristocrat Appeared to His Guests” in Gazda, E. K. (ed.) Roman Art in the Private Sphere: New Perspectives on the Architecture and Decor of the Domus, Villa and Insula. Michigan: University of Michigan Press, pp. 117- 134.
Ellis, S. P. (1997) “Late Antique Dining: Architecture, Furnishings and Behaviour” in Laurence, R. and Wallace-Hadrill, A. (eds.) Domestic Space in the Roman World: Pompeii and Beyond. JRA Supplementary series 22, pp. 41-52.
Ellis, S. P. (2000) Roman Housing. London: Duckworth.
334
Ellis, S. P. (2004a) “Early Byzantine Housing” in Dark, K. (ed.) Secular Life and the Archaeology of Everyday Life in the Byzantine Empire. Oxford: Oxbow, pp. 37-51.
Ellis, S. P. (2004b) “The Seedier Side of Antioch” in Sandwell, I. and Huskinson, J. (eds.) Culture and Society in Later Antioch. Oxford: Oxbow Books, pp. 126-133.
Elsner, J. (1998) Imperial Rome and Christian Triumph. Oxford and New York: Oxford University Press.
Erarslan, A. (2021) “Evolution of the Courtyard House in the Pre-Islamic Period in Mesopotamia” in Ranguelav, B. et al. (eds.) Developments in Engineering and Architecture. Sofia: ST. Kliment Ohridski University Press. pp. 194-209.
Ergeç, R. (1993) “Belkıs/Zeugma Mozaik Kurtarma Kazısı 1992”, IV. Müze Kurtarma Kazıları Semineri, pp. 321-337.
Ergeç, R. (1995) “Belkıs Zeugmada Bir Roma Villası ve Taban Mozaikleri”, Arkeoloji ve Sanat 66, pp. 2-10.
Ergeç, R. (1998a) “Belkıs/Zeugma Roma Villası”, VIII. Müze Kurtarma Kazıları Semineri, pp. 407-418.
Ergeç, R. (1998b) “Rescue Excavations by the Gaziantep Museum” in Kennedy, D. L. (ed.) The Twin Towns of Zeugma on the Euphrates/Rescue Work and Historical Studies. JRA Supplementary Series No.27, Portsmouth, pp. 80-91.
Ergeç, R. (2000) “Belkıs/Zeugma 1992-1999/2000 Çalışmalar, Araştırmalar ve Kazılar”, Arkeoloji ve Sanat 98, pp. 20-29.
Ergeç, R. (2005) “Zeugma’nın Kronolojik Tarihi”, Ayıntap 4, pp. 19-22.
Ergeç, R. and M. Önal (1998) “Bekıs/Zeugma Roma Hamamı ve Kompleksi Kurtarma Kazısı”, VIII. Müze Kurtarma Kazıları Semineri, pp. 419-443.
335
Fink, R. O. et al. (1959) “The Texts” in Welles, C. B., Fink, R. O. and Gilliam, J. F. (eds.) The Excavations at Dura-Europos: Final Report, vol.5. I: The Parchments and Papyri. New Haven: Yale University Press, pp. 69- 417.
Finlayson, C. (2012) “New Excavations and a Reexamination of the Great Roman Theater at Apamea, Syria, Seasons 1-3 (2008-2010)” American Journal of Archaeology 116, no. 2, pp. 277-319.
Fisher, C. S. (1934) “Bath B, House A and the Roman Villa” in Elderkin, G. W. (ed.) Antioch on the Orontes 1: Excavations of 1932. Princeton: Princeton University Press, pp. 8-18.
Foss, C. (2000) “Late Antique Antioch” in Kondoleon, C. (ed.) Antioch: The Lost Ancient City. Princeton, New Jersey: Princeton University Press, pp. 22-27.
Frezouls, E. (1976) “A Propos de l’Architecture Domestique à Palmyre”, Ktema 1, pp. 29-52.
Gabriel, A. (1926) “Recherches Archéologiques a Palmyre”, Syria, T. 7, Fasc.1, pp. 71-92.
Gardner, J. F. and Wiedemann, T. (1991) The Roman Household: a Sourcebook. London and New York: Routledge.
Gates, C. (2011) Ancient Cities: The Archaeology of Urban Life in the Near East and Egypt, Greece and Rome. 2nd edition. London: Routledge.
Gawlikowski, M. (1974) “Les défenses de Palmyre”, Syria 51, pp. 231-242.
Gawlikowski, M. (1987) “Nomadism and Sedentarization in Palmyra”, in Frézouls, E. (ed.) Sociétés Urbaines, Sociétés Rurales dans l’Asie Mineure et la Syrie Hellenistiques et Romaines. Strasbourg: AECR, pp. 49-55.
Gawlikowski, M. (1991a) “Fouilles Récentes à Palmyre”, Comptes rendus des séances de l'Académie des Inscriptions et Belles- Lettres, 135e année 2, pp. 399-410.
Gawlikowski, M. (1991b) “Palmyra”, Polish Archaeology in the Mediterranean II- Reports 1989-1990, pp. 85-90.
336
Gawlikowski, M. (1993) “Rapports des missions actives à Palmyre 1990-1991”, Syria 70, Fasc. 3-4, pp. 561-576.
Gawlikowski, M. (1994) “Palmyra”, Polish Archaeology in the Mediterranean V- Reports 1993, pp. 133-143.
Gawlikowski, M. (1996) “Palmyra- Excavations 1995”, Polish Archaeology in the Mediterranean VII- Reports 1995, pp. 139-146.
Gawlikowski, M. (1997a) “L’Habitat à Palmyre de l’Antiquité au Moyen-Age” in Castel, C., Al-Maqdissi, M. and Villeneuve, F. (eds) Les Maisons dans la Syrie Antique du IIIe Millénaire aux Débuts de l’Islam, Pratiques et Représentations de l’Espace Domestique. Actes du Colloque International, Damas 27-30 Juin 1992. Beyrouth: IFAPO, pp. 161-166.
Gawlikowski, M. (1997b) “Palmyra-Excavations 1996”, Polish Archaeology in the Mediterranean VIII- Reports 1996, pp. 191-197.
Gawlikowski, M. (2007) “Beyond the Colonnades: Domestic Architecture in Palmyra”, in Galor, K. and Waliszewski, T. (eds.) From Antioch to Alexandria, Recent Studies in Domestic Architecture, Warsaw: Institute of Archaeology, pp. 79-94.
Gawlikowski, M. (2008) “Palmyra in Early Islamic Times” in Bartl, K. and Moaz, A. R. (eds.) Residences, Castles, Settlements. Transformation Processes from Late Antiquity to Early Islam in Bilad Al-Sham. Proceedings of the International Conference held at Damascus, 5-9 November 2006. Westf: Verlag Marie Lerdorf, pp. 89-96.
Gawlikowski, M. (2015) “Bel of Palmyra” in Blömer, M, Lichtenberger, A. and Raja, R. (eds.) Religous Identities in the Levant from Alexander to Muhammed. Turnhout: Brepols, pp. 247-254.
Gisler, J. R. and Huwiler, M. (1984) “La Maison aux Pilastres” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 79-106.
Goldman, B. (1999) "Pictorial Graffiti of Dura-Europos" Parthica 1, pp. 19-106.
337
Görkay, K. (2011) “Crossing the Euphrates in Antiquity”, Actual Archaeology Magazine 1, pp. 36-51.
Görkay, K. (2012a) “Zeugma in Light of New Research” in Konuk, K. (ed.) Stephanephoros de l’economie antique à l’Asie Mineure. Pessac: Ausonius, pp. 275-300.
Görkay, K. (2012b) “Zeugma Roma Dönemi Konutları”, Arkeoloji ve Sanat 141, pp. vii-xviii.
Görkay, K. (2015) Geçmişten Günümüze bir Geçit Zeugma. İstanbul: Türkiye İş Bankasi Kültür Yayınları.
Görkay, K. (2017) “Mosaic Programmes in Domestic Contexts at Zeugma”, JMR 10, pp. 183-211.
Görkay, K. (2020a) “New Research at Zeugma” in Görkay, K. (ed.) Zeugma. Between Two Worlds: The Houses and Tombs of Zeugma from Life to Eternity. İstanbul: Türkiye İş Bankası Kültür Yayınları, pp. 20-45.
Görkay, K. (2020b) “An Overview of Roman Houses in Zeugma” in Görkay, K. (ed.) Zeugma. Between Two Worlds: The Houses and Tombs of Zeugma from Life to Eternity. İstanbul: Türkiye İş Bankası Kültür Yayınları, pp. 46-69.
Görkay, K. (2020c) “Convivia and Spaces” in Görkay, K. (ed.) Zeugma. Between Two Worlds: The Houses and Tombs of Zeugma from Life to Eternity. İstanbul: Türkiye İş Bankası Kültür Yayınları, pp. 70-85.
Görkay, K. (2020d) “Living in Colour: Wall Decorations in the Houses of Zeugma” in Görkay, K. (ed.) Zeugma. Between Two Worlds: The Houses and Tombs of Zeugma from Life to Eternity. İstanbul: Türkiye İş Bankası Kültür Yayınları, pp. 138-159.
Görkay, K. (2020e) “Mosaics in the Roman Houses of Zeugma” in Görkay, K. (ed.) Zeugma. Between Two Worlds: The Houses and Tombs of Zeugma from Life to Eternity. İstanbul: Türkiye İş Bankası Kültür Yayınları, pp. 160-259.
338
Grahame, M. (1997) “Public and Private in the Roman House: The Spatial Order of the Casa del Fauno”, in Laurence, G. and Wallace-Hadrill, A. (eds.) Domestic Space in the Roman World: Pompeii and Beyond. Portsmouth, pp. 137-164.
Grahame, M. (1998) “Material Culture and Roman Identity: The Spatial Layout of Pompeian Houses and the Problem of Ethnicity” in Laurence, R. and Berry, J. (eds.) Cultural Identity in the Roman Empire. London and New York: Routledge, pp. 156-178.
Grainger, J. D. (1990) The Cities of Seleukid Syria. Oxford: Oxford University Press.
Grassi, M. T. (2008) “La romanità orientale e Palmira: nuove ricerche”, Bollettino Archeologia Online, pp. 1-11.
Grassi, M. T. (2011) “Une nouvelle Mission conjointe Syro-Italienne à Palmyre (Pal.M.A.I.S.): observations préliminaires sur les campagnes 2007-2009 dans le quartier sud-ouest”, Chronique Archéologique en Syrie V, pp. 189-196.
Grassi, M. T. and Al-As’ad W. (2013) “Pal.M.A.I.S. Researches and Excavations of a New Joint Syro-Italian Mission in the South West Quarter of Palmyra” in Gawlikowski, M. and Majcherek, G. (eds.) Studia Palmyrenskie XII. Fifty Years of Excavations in Palmyra 1959-2009. Warsaw: Warsaw University Press, pp. 115-128.
Grassi, M. T., Rocca, G. and Piacentini, D. (2015) “Les nouveautés épigraphiques de la Mission Archéologique Italo‐Syrienne de Palmyre”, LANX 20, pp. 1‐48
Gruber, E. and Dobbins, J. (2010) “Illuminating Historical Architecture: The House of the Drinking Contest at Antioch”, in Contreras, F., Farjas, M., and Melero, F. J. (eds.) Proceedings of the 38th Annual Conference on Computer Applications and Quantitative Methods in Archaeology, CAA2010, pp. 71-76.
Güney, Ö. (2012) Investigating the House Church in Dura-Europos: Production of Social Space. Unpublished Masters Thesis. Ankara: Middle East Technical University.
Hales, S. (2003) The Roman House and Social Identity. Cambridge: Cambridge University Press.
339
Hanson, J. W. (2011) "The Urban System of Roman Asia Minor and wider connectivity", in Bowman, A. K. and Wilson, A. I. (eds.) Settlement, Urbanization, and Population. Oxford Studies on the Roman Economy. Vol. 2. Oxford: Oxford University Press. pp. 229-275.
Hartmann, M. and Speidel, M. A. (2003) “The Roman Army at Zeugma: Results of New Research” in Humphrey, J. H. (ed.) (2003) Zeugma: Interim Reports, Rescue Excavations. JRA Supplementary series 51, pp. 100-126.
Hartmann, M. and Speidel, M. A. (2013) “Military Installations at Zeugma: An Overview of the Swiss Archaeological Investigations, 2001-2003.” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol III. Los Altos, California: The Packard Humanities Institute, pp. 381-392.
Hobson, B. (2009) Latrinae et Foricae. Toilets in the Roman World. London: Bristol Classical Press.
Hoepfner, W and Schwander E. L. (1986) Haus und Stadt im klassischen Griechenland. Munich: Deutscher Kunstverlag.
Hopkins, C. (1931) “Minor Finds and Building Details” in Baur, P. V. C. and Rostovtzeff, M. I. (eds.) The Excavations at Dura-Europos. Preliminary Report of Second Season of Work. October 1928-April-1929. New Haven: Yale University Press, pp. 53-74. Hopkins, C. (1934) “The Christian Church” in Rostovtzeff, M. I. (ed.) The Excavations of Dura-Europos: Preliminary Report of the Fifth Season 1931-32. New Heaven: Yale University Press, pp. 238- 254.
Hopkins, C. (1972) Topography and Architecture of Seleucia on the Tigris. Ann Arbor: University of Michigan.
Hopkins, C. (1979) The Discovery of Dura-Europos. New Haven: Yale University Press.
Hopkins, C. and Rowell, H. T. (1934) “The Praetorium” in Rostovtzeff, M. I. (ed.) The Excavations of Dura-Europos: Preliminary Report of the Fifth Season 1931-32. New Heaven: Yale University Press, pp. 201-237.
340
Hudson, N. F. (2010) “Changing Places: The Archaeology of the Roman ‘Convivium’”, AJA 114.4, pp. 663- 695.
Humphrey, C. (1988) “No Place like Home in Anthropology: the Neglect of Architecture”, Anthropology Today 4, pp. 16-18.
Immerwahr, H. (1944) “Dipinti from G5, C2” in Rostovtzeff, M. I. et al. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Ninth Season 1935-36. Part 1: The Agora and Bazaar. New Heaven: Yale University Press, pp. 203-265.
Intagliata, E. E. (2018) Palmyra after Zenobia AD 273-750: An Archaeological and Historical Reappraisal. Oxford: Oxbow Books.
Janniard, S. and Traina, G. (2006) “Sur le concept de «romanisation». Paradigmes historiographiques et perspectives de recherche” MEFRA 118, pp. 71-166.
Jashemski, W. F. et al. (eds.) (2018) Gardens of the Roman Empire. Cambridge: Cambridge University Press.
Jones, S. (1997) The Archaeology of Ethnicity. Constructing Identities in the Present and Past. New York and London: Routledge.
Jourdain, C. (1972) “Sondages dans l’Insula ‘Au Triclinos’, 1970 et 1971” in Balty, J. C. and Balty, J. (eds) Apamée de Syrie, Bilan des Recherches Archéologiques 1969-1971, Actes du Colloque Tenu à Bruxelles les 15,17 et 18 Avril 1972. Brussels, pp. 113-142.
Juchniewicz, K. (2013) “Late Roman Fortifications in Palmyra” in Gawlikowski, M. and Majcherek, G. (eds.) Studia Palmyrenskie XII. Fifty Years of Excavations in Palmyra 1959-2009. Warsaw: Warsaw University Press, pp. 193-202.
Kaizer, T. (ed.) (2022) A Companion to the Hellenistic and Roman Near East. Oxford& Cambridge: Blackwell.
Karabulut, H., Önal, M. and Dervişoğlu, N. (2011) Haleplibahçe Mozaikleri Şanlıurfa. İstanbul: Arkeoloji ve Sanat Yayınları.
341
Karampekos, S. (2020) “East of the Levant. Residential Architecture in Seleukeia on the Tigris: Neo-Babylonian Tradition and Hellenistic Innovation” in Held, W. (ed.) The Transition from the Achaemenid to the Hellenistic Period in the Levant, Cyprus, and Cilicia: Cultural Interruption or Continuity? Marburg: Eigenverlag des Archäologischen Seminars der Philipps, pp. 173-190.
Kenfield, S. (2014) “History of the Antioch Excavations” in Redford, S. (ed.) Antioch on the Orontes. İstanbul: Koç Üniversitesi Yayınları, pp. 37-77.
Kenfield, S. and Moss, C. (2014) “Catalogue” in Redford, S. (ed.) Antioch on the Orontes. İstanbul: Koç Üniversitesi Yayınları, pp. 129-254.
Kennedy, D. (1985) ‘From Polis to Madina: Urban Change in Late Antique and Early Islamic Syria’, Past and Present 106, pp. 3–27.
Kennedy, D. (1998a) “The Twin Towns and the Region” in Kennedy, D. L. (ed.) The Twin Towns of Zeugma on the Euphrates/Rescue Work and Historical Studies, JRA Supplementary Series No.27, Portsmouth, pp. 30-59.
Kennedy, D. (1998b) “)Ancient Sources for Zeugma” in Kennedy, D. (ed.) The Twin Towns of Zeugma on the Euphrates/Rescue Work and Historical Studies, JRA Supplementary Series No.27, Portsmouth, pp. 139-162.
Kennedy, D. (1998c) “Conclusions: Roman Zeugma” in Kennedy, D. L. (ed.) The Twin Towns of Zeugma on the Euphrates/Rescue Work and Historical Studies, JRA Supplementary Series No.27, Portsmouth, pp. 237-243.
Kennedy, D. (2006) “Demography, the Population of Syria and the Census of Q. Aemilius Secundus”, Levant 38, pp. 109-124.
Kennedy, D. and Graf, D. (1998) “Inscriptions on Stone, Ceramic and Mosaic”, in Kennedy (1998a) The Twin Towns of Zeugma on the Euphrates/Rescue Work and Historical Studies, JRA Supplementary Series 27. Portsmouth, pp. 92-108.
Kent, S. (ed.) (1993) Domestic Architecture and the Use of Space: An Interdisciplinary Cross-Cultural Study (New Directions in Archaeology). Cambridge: Cambridge University Press.
342
King, G. R. D. and Cameron, A. (eds.) (1994) The Byzantine and Early Islamic Near-East. Land Use and Settlement Patterns. Princeton: Darwin Press.
Kloeg, P. (2013) Antioch the Great: Population and Economy of Second Century Antioch. Unpublished Master Thesis. Leiden: Leiden University.
Knights, C. (1994) “The Spatiality of the Roman Domestic Setting: an Interpretation of Symbolic Content” in Pearson, M. P. and Richards, C. (eds.) Architecture and Order. Approaches to Social Space. London: Routledge, pp. 113–46.
Kondoleon, C. (1991) “Signs of Privilege and Pleasure: Roman Domestic Mosaics” in Gazda, E. K. (ed.) Roman Art in the Private Sphere: New Perspectives on the Architecture and Decor of the Domus, Villa and Insula. Michigan: University of Michigan Press, pp. 105-116.
Kondoleon, C. (2000) Antioch: The Lost Ancient City.Princeton: Princeton University Press.
Kosmin, P. (2011) “The Foundation and Early Life of Dura-Europos” in Brody, L. and Hoffman, G. (eds.) Dura-Europos: Crossroads of Antiquity. Chestnut Hill: McMullen Museum of Art, pp. 95-109.
Kowalski, S. P. (1997) “Late Roman Palmyra in Literature and Epigraphy”, Studia Palmyrenskie X, pp. 39-62.
Kraeling, C. H. (1967) The Excavations at Dura-Europos: Final Report, vol.VIII.II.- The Christian Building. New Haven: Yale University Press.
Krencker, D. (1932) “Bemerkungen zum Stadtplan” in Wiegand. T. (ed.) Palmyra. Ergebnisse der Expeditionen von 1902 und 1917. Berlin: Keller, pp. 36-37.
Lassus, J. (1934) “La Mosaique de Yakto” in Elderkin, G. W. (ed.) Antioch on the Orontes 1: Excavations of 1932. Princeton: Princeton University Press, pp. 114-156.
Lassus, J. (1935) “Dans les Rues d’Antioche”, Bulletin d'études orientales, T. 5, pp. 121-124.
343
Lassus, J. (1938) “Une villa de Plaisance” in Stillwell, R. (ed.) Antioch-on-the-Orontes II: The Excavations of 1933-1936. Princeton: Princeton University Press, pp. 95-147.
Lassus, J. (1969) “La Maison des Chretiens de Dura-Europos”, Revue Archéologique 1, pp. 129-140.
Lassus, J. (1972) Antioch-on-the-Orontes V: Les Portiques d’Antioche. Princeton: Princeton University Press.
Lassus, J. (1984) “Sur les Maisons d’Antioche” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29, 30 et 31 Mai 1980. Brussels, pp. 361-372.
Lavin, I. (1963) “The Hunting Mosaics of Antioch and Their Sources”, Dumbarton Oaks Papers XVII, pp. 179-286.
Leach, E. W. (1997) “Oecus on Ibycus: Investigating the Vocabulary of the Roman House” in Bon, E. S. and Jones, R. (eds.) Sequence and Space in Pompeii. Exeter: The Short Run Press, pp. 50-72.
Lefebvre, H. (1991) The Production of Space. (Translated by Donald Nicholson-Smith). Oxford& Cambridge: Blackwell.
Leriche P. and Al Mahmoud, A. (1988) "Bilan des Campagnes des 1986 et 1987 de la Mission Franco-Syrienne à Doura-Europos ", Syria 65, pp. 261-282.
Leriche, P. (1993) "Nouvelles données sur l'histoire architecturale et urbaine de Doura-Europos", in Invernizzi, A. and Salles, J.- F. (eds.), Arabia Antiqua: Hellenistic Centres Around Arabia. Rome: Istituto Italiano per il Medio ed Estremo Oriente, pp. 113-127.
Leriche, P. (1996) “Le Chreophylakeion de Doura-Europos et la Mise en Place du Plan Hippodamien de la Ville”, in Boussac, M.-F. and Invernizzi. A. (eds) Archives et Sceaux du Monde Hellénistique. Paris: Bulletin de Correspondance Hellenic Supplement, 29, pp. 157–69.
344
Leriche, P. (1997) “Pourquoi et Comment Europos a été Fondée à Dura?” in Brulé, P. and Oulhen, J. (eds.) Esclavage, Guerre, Économie en Grèce Ancienne. Rennes: Presses Universitaires de Rennes, pp. 191-210.
Leriche, P. (2003) “Europos-Doura Hellenistique”, Topoi Supplement 4, pp. 171- 191.
Leriche, P. (2004) “La Rue Principal et l’Urbanisme d’Europos-Doura. Etude Préliminaire”, Parthica 6, pp. 139-159.
Leriche, P. and Al Mahmoud, A. (1994) "Doura-Europos. Bilan des Recherches Récentes", CRAI 138-2, pp. 395-420.
Leriche, P. and MacKenzie, D. N. (1996) “Dura-Europos”, Encyclopædia Iranica, Vol. VII, Fasc. 6, pp. 589-594.
Levi, D. (1947) Antioch Mosaic Pavements I-II. Princeton: Princeton University Press.
Maas, M. (2000) “People and Identity in Roman Antioch” in Kondoleon, C. (ed.) Antioch: The Lost Ancient City. Princeton, New Jersey: Princeton University Press, pp. 13-22.
Malmberg, S. (2005) “Visualizing Hierarchy at Imperial Banquets”, in Mayer, W. and Trzcionka, S. (eds.) Feast, Fast or Famine: Food and Drink in Byzantium. Brisbane. pp. 11-24.
Malmberg, S. (2012) “Banquets, Byzantine”, in. Bagnall, R. et al. (eds.), Encyclopaedia of Ancient History. Oxford: Oxford University Press, pp. 1035-1036.
Manière-Lévêque, A. (2007) “The House of the Lycian Acropolis at Xanthos”, in Lavan, L., Özgene, L. and Sarantis, A. (eds.) Housing in Late Antiquity. From Palaces to Shops. Leiden, Boston: Brill. pp 475-494.
Manière-Lévêque, A. (2002) “La Maison de l’Acropole Lycienne à Xanthos”, Anatolia Antiqua 10, pp. 235-243.
Manière-Lévêque, A. (2006) “Habitat Protobyzantine à Xanthos: les Espaces de recéption de la grande résidence nord-est de l’acropole lycienne”, in Dörtlük,
345
K. et al. (eds.), Proceedings of The Third International Symposium on Lycia, Antalya 7.-10. November 2005. pp. 425-440.
Marro, C., Tibet, A and Bulgan, F. (2000) “Fouiles de Sauvatage de Horum Höyük (Province de Gaziantep): Quatrième Rapport Préliminaire”, Anatolia Antiqua VIII, pp. 257-278.
Martz, A. S. (2011) Maisons d'Antioche à l'Époque Romaine: Architecture et Mobilier Archéologique. (Available at https://hal.archives- ouvertes.fr/hal-00749696)
Martz, A. S. (2015) “Morphologie et signification des triclinia dans les maisons de Syrie du Nord” in Abadie-Reynal, C. and Yon, J. B. (eds.) Zeugma VI, La Syrie Romaine, Permanences et Transferts Culturels. Lyon, Maison de l’Orient et de la Méditerranée. pp.165-184.
Mattingly, D. J. (2004) “Being Roman: Expressing Identity in a Provincial Setting” Journal of Roman Archaeology 17, pp. 5-25.
Mayence, F. (1935) “La Quatrième Campagne de Fouilles à Apamée (Rapport provisoire)”, L'Antiquité Classique, Tome 4, fasc. 1, pp. 199-204.
Mayence, F. (1936) “La Ve Campagne de Fouilles à Apamée (Rapport provisoire)”, L'Antiquité Classique, Tome 5, fasc. 2, pp. 405-411.
Mayence, F. (1939) “La VIe Campagne de Fouilles à Apamée (Rapport provisoire)”, L'Antiquité Classique, Tome 8, fasc. 1, pp. 201-211.
Metheny, J. R. (1907) “Road Notes from Cilicia and North Syria”, JAOS 28. pp. 155-163.
Meyer, P. M. (1910) Griechische Papyri im Museum des Oberhessischen Geschichtsvereins zu Giessen. Leipzig: Teubner.
Mikocka, J. (2015) “Private architecture in Ptolemais (Libya): excavations and non-invasive surveys”, Novensia 26, pp. 83-102.
346
Mikocka, J. and Misiewicz, K. (2015) “Domestic Architecture in Ptolemais in Light of Non-invasive Surveys” in Jaworski, P. and Misiewicz, K. (eds.) Ptolemais II. Warsaw: Institute of Archaeology, University of Warsaw. 159-221.
Millar, F. (1993) The Roman Near East, 31BC-AD 337. Cambridge and Massachusetts: Harvard University Press.
Morvillez, E. (2004) “L'architecture domestique à Antioche dans l'Antiquité Tardive: Conservatisme ou Modernité?”, Antioche 2004, pp. 271-287.
Morvillez, E. (2007) “A Propos de l’Architecture Domestique d’Antioche, de Daphne et Seleucie” in Galor, K. and Waliszewski, T. (eds.) From Antioch to Alexandria, Recent Studies in Domestic Architecture, Warsaw: Institute of Archaeology, pp. 51-78.
Müller, C. O. (1839) Antiquitates Antiochenae. Göthingen: Libraria Dietrichiana.
Najbjerg, T. and Moss, C. (2014) “The History of Antioch on the Orontes in the Greco-Roman Period” in Redford, S. (ed.) Antioch on the Orontes. İstanbul: Koç Üniversitesi Yayınları, pp. 22-35.
Nava, S. (2015) “The marble decoration of the peristyle building in the SW quarter of Palmyra (Pal.M.A.I.S. Mission)”, ASMOSIA X, pp. 241-252.
Nevett, L. (1999) House and society in the Ancient Greek World. New York: Cambridge University Press.
Nevett, L. (2000) “A Real Estate ‘Market’ in Classical Greece? The Example of Town Housing”, BSA 95, pp. 329-343.
Nevett, L. (2010) Domestic Space in Classical Antiquity. New York: Cambridge University Press.
Nielsen, I. (1994) Hellenistic Palaces, Tradition and Renewal. Aarhus: Aarhus University Press.
Nielsen, I. (ed.) (1998) Meals in a Social Context: Aspects of the Communal Meal in the Hellenistic and Roman World. Aarhus: Aarhus University Press.
347
Nissin L. (2015) Sleeping Arrangements in the Houses of Herculaneum, in Tuori K. and Nissin L. (eds.) Public and Privatein the Roman House and Society. JRA Supplementary Series 101. Portsmouth, pp. 101-118.
Nissinen L. (2009) Cubicula Diurna, Nocturna-Revisiting Roman Cubicula and Sleeping Arrangements, Arctos, 43, pp. 85-107.
O'Conner, J. M. (2008) St. Paul's Ephesus: Texts and Archaeology. Collegeville: Liturgical Press.
Önal, M. (2005) “Belkıs/Zeugma 2003 Yılı Kazı Çalışmaları”, 14. Müze Çalışmaları ve Kurtarma Kazıları Sempozyumu, pp. 179-194.
Önal, M. (2013) Belkıs/Zeugma: Poseidon-Euphrates Evleri. İstanbul: Arkeoloji ve Sanat Yayınları.
Önal, M., Jean, E. and Barlak, H. (2012) “Fouilles du Musée de Gaziantep dans la Maison de Poséidon (ou Maison 1) et la Maison de l'Euphrate (ou Maisons 2 et 3). Années 1999-2000” in Abadie-Reynal, C. and Ergeç, R. (eds.) Zeugma 1: Fouilles de L’Habitat (1) La Mosaique de Pasiphae. Varia Anatolica XXVI. İstanbul: De Boccard/IFEA, pp. 65-112.
Özbek, A. and Özcan, I. (2015) Haleplibahçe Mozaikleri/Haleplibahçe Mosaics. Şanlıurfa: Karacadağ Kalkınma Ajansı.
Özgenel, L. (2007) “Public Use and Privacy in Late Antique Houses in Asia Minor: the Architecture of Spatial Control” in Lavan, L., Özgenel, L. and Sarantis, A. (eds.) Housing in Late Antiquity. From Palaces to Shops. Leiden, Boston: Brill. pp. 239-282.
Özgenel, L. (2018) “Between public and Private: rethinking Architecture and Use in Late Antique Houses in Asia Minor”, Abitare Nel Mediterraneo Tardoantico Atti del II Convegno Internazionale del Centro Interuniversitario di Studi sull’Edilizia abitativa tardoantica nel Mediterraneo (CISEM) (Bologna 2-5 marzo 2016), pp. 541-547.
Özgenel, L. (2022) “The Nature of the ‘Courtyard House’: Sampling Late Antique Houses in Asia Minor” in Uytterhoeven, I. and Ricci, A. (eds.) Palimpsest of the House. reassessing Roman, Late Antique, Byzantine, and Early Islamic Living Patterns. İstanbul: Koç University Press. pp. 165-195.
348
Packer, J. E. (1971) The Insulae of Imperial Ostia (Memoirs of the American Academy in Rome, XXXI). Rome: American Academy in Rome.
Palmieri, L. (2010) “Étude préliminaire sur les stucs trouvés dans le “Bâtiment à Péristyle” de Palmyre”, LANX 6, pp. 175-186.
Pamir, H. (2014) “Archaeological Research in Antioch on the Orontes and Its Vicinity: 2002-2012” in Redford, S. (ed.) Antioch on the Orontes. İstanbul: Koç Üniversitesi Yayınları, pp. 78-127.
Parton, H. (2013) “Milling and Weaving Equipment, including Hand-held Stone Tools, Mortars, Querns, and Stone Vessels, Loom Weights, and Spindle Whorls” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol III. Los Altos, California: The Packard Humanities Institute, pp. 295-344.
Pederson, P. (1992) “Excavations and Research in Halicarnassos 1991” XIV. Kazı Sonuçları Toplantısı 2, pp. 133-141.
Perkins, A. (1973) The Art of Dura-Europos. Oxford: Oxford University Press.
Pillet, M. (1931) “Results of Excavations” in Baur, P. V. C. and Rostovtzeff, M. I. (eds.) The Excavations at Dura-Europos. Preliminary Report of Second Season of Work. October 1930- March 1931. New Haven: Yale University Press, pp. 1-19.
Pillet, M. (1933) “General Report” in Baur, P. V. C., Rostovtzeff, M. I. and Bellinger, A. R. (eds.) The Excavations at Dura-Europos. Preliminary Report of Fourth Season of Work. October 1928-April-1929. New Haven: Yale University Press, pp. 1-41.
Pitts, M. (2007) “The Emperor’s New Clothes? The Utility of Identity in Roman Archaeology.” American Journal of Archaeology 111, pp. 693-713.
Pitts, M. (2008) “Globalizing the Local in Roman Britain. An Anthropological Approach to Social Change.” Journal of Anthropological Archaeology 27, pp. 493–506.
349
Poccardi, G. (1994) “Antioche de Syrie. Pour un Nouveau Plan Urbain de l'île de l'Oronte (Ville Neuve) du IIIe au Ve Siècle”, Mélanges de l'Ecole française de Rome. Antiquité, Tome 106, no.2, pp. 993-1023.
Polci, B. (2003) “Some Aspects of the Transformation of the Roman Domus Between Late Antiquity and the Early Middle Ages” in Lavan, L. and Bowden, W. (eds.) Theory and practice in Late Antique Archaeology. Boston: LAfeiden. pp. 79-109.
Pollard, N. (1996) “The Roman Army as ‘Total Institution’ in the Near East?” in Kennedy, D. (ed.) The Roman Army in the East. JRA Supplementary Series 18. Ann Arbor, MI, pp. 211-227.
Pollard, N. (2000) Soldiers, Cities, and Civilians in Roman Syria. Michigan: University of Michigan Press.
Pontbriand, S. (2012) “La Résidence de Lysias à Europos-Doura. Une Première Approche”, in Leriche, P., Coqueugniot, G. and Pontbriand, S. (eds.) Europos- Doura Varia I. Beyrouth: Ifpo, pp. 77- 92.
Pontbriand, S. (2016) “Europos- Dura at the Convergence of Hellenistic, Parthian and Roman Architectures: The Case of Governor Lysias’ Residence”, in Foietta, E. (eds.) Cultural and Material Contacts in the Ancient Near East. Proceedings of the International Workshop. 1-2 December 2014, Torino, pp. 118- 128.
Poulsen, B. (1997) “The City Personifications in the Late Roman Villa in Halicarnassos” in Isager, S. and Poulsen, B. (eds.) Patrons and Pavements in Late Antiquity. Odense: Odense University Press, pp. 9-23.
Raepsaet-Charlier, M. T. and Raepsaet-Charlier, G. (1984) “La Maison aux Colonnes Trilobées” in Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 181-201.
Rawson, B. (ed.) (1986) The Family in Ancient Rome. New York: Cornell University Press.
Revell, L. (2008) Roman Imperialism and Local Identities. New York: Cambridge University Press.
350
Revell, L. (2015) Ways of Being Roman: Discourses of Identity in the Roman West. Oxford: Oxbow Books.
Riggsby, A. M. (1997) “Public’ and ‘Private’ in the Roman Culture: the Case of the Cubiculum”, Journal of Roman Archaeology 10, pp. 37-56.
Ritter, C. (1843) Die Erdkunde im Verhältnis zur Naturund zur Geschicte des Menschen, Zehnter Teil, Drittes Buch, West Asien. Berlin.
Robinson, D. M. and Graham, J. W. (1938) The Hellenic House, The Excavations at Olynthus, vol. VIII. Baltimore: The Johns Hopkins University Press.
Robinson, O. R. (1992) Ancient Rome: City Planning and Administration. New York: Routledge.
Rose, C. B. (2013) “A New Relief of Antiochus I of Commagene and Other Stone Sculpture from Zeugma” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute, pp. 220-231.
Rostovtzeff, M. I, et al. (eds.) (1939) The Excavations at Dura-Europos conducted by Yale University and the French Academy of Inscriptions and Letters: Preliminary Report of the Seventh and Eighth Seasons of Work, 1933–1934 and 1934–1935. New Heaven: Yale University Press.
Rostovtzeff, M. I. (1932) Caravan Cities. Oxford: Oxford University Press.
Rostovtzeff, M. I. (1935) “Dura and the Problem of Parthian Art”, Yale Classical Studies 5, pp. 155-304.
Rostovtzeff, M. I. (1937) “Rapport sur les Fouilles de Doura-Europos, Campagne de 1936-1937”, Comptes Rendus. Academie des Inscriptions et Belles-Lettres, pp. 195-205.
Rostovtzeff, M. I. (1938) Dura-Europos and Its Art. Oxford: Oxford University Press.
Rostovtzeff, M. I. (1941) The Social and Economic History of the Hellenistic World, 3 vols. Oxford: Oxford University Press.
351
Rostovtzeff, M. I. (1952) “Interpretation” in Rostovtzeff, M. I. et al. (eds) The Excavations at Dura-Europos. Preliminary Report of the Ninth Season of Work, 1935- 1936, Part 3: The Palace of the Dux Ripae and the Dolicheneum. New Haven: Yale University Press, pp. 69- 96.
Rostovtzeff, M. I. and Little, A. (1933) “La Maison Fresques de Dura-Europos”, Memoires de l’Institute National de France, 43, pp. 167-190.
Rostovtzeff, M. I. and Welles, C. B. (1952) “Inscriptions” in Rostovtzeff, M. I. et al. (eds) The Excavations at Dura-Europos. Preliminary Report of the Ninth Season of Work, 1935- 1936, Part 3: The Palace of the Dux Ripae and the Dolicheneum. New Haven: Yale University Press, pp. 27- 39.
Rostovtzeff, M. I. et al. (1944) The Excavations of Dura-Europos: Preliminary Report of the Ninth Season 1935-36. Part 1: The Agora and Bazaar. New Heaven: Yale University Press.
Rostovtzeff, M. I. et al. (1952) The Excavations of Dura-Europos: Preliminary Report of the Ninth Season 1935-36. Part 3: The Palace of the Dux Ripae and the Dolicheneum. New Heaven: Yale University Press.
Rumscheid, F. (2010), “Fragen zur bürgerlich-hellenistischen Wohnkultur in Kleinasien”, in Ladstätter, S. and Scheibelreiter, V. (eds.) Städtisches Wohnen im östlichen Mittelmeerraum, 4. Jh. v. Chr.–1. Jh. n. Chr. Akten des internationalen Kolloquiums vom 24.–27. Oktober 2007.,Vienna, pp. 119-143.
Sachau, E. (1883) Reise in Syrien und Mesopotamien. Leipzig.
Saliby, N. (1996) “Maisons Palmyréniennes à l’Est du Temple de Nabu”, Annales Archéologiques Arabes Syriennes XLII, pp. 289-290.
Saliou, C. (1992) “Le Quatre Fils de Polemocratès (P. Dura 19). Texte et Archéologie”, Syria 69, pp. 65- 100.
Saliou, C. (2005) La Forme d’un Ilot de Doura-Europos… L’Ilot C7 Revisite” in Leriche, P., Gelin, M. and Dandrau, A. (eds.) Doura- Europos: études V 1994- 1997. Paris: P. Geuthner, pp. 65- 78.
352
Savino, E. (1999) “Palmira” in Savino, E. (ed.) Città di frontiera nell’impero Romano. Forme della romanizzazione da Augusto ai Severi. Bari: Edipuglia, pp. 47-93.
Schede, M. (1964) The Ruinen von Priene. Berlin: Walter de Gruyter.
Schmidt-Colinet, A., Al-As’ad, K. and Al-As’ad, W. (2013) “Thirty Years of Syro-German/Austrian Archaeological Research at Palmyra” in Gawlikowski, M. and Majcherek, G. (eds.) Studia Palmyrenskie XII. Fifty Years of Excavations in Palmyra 1959-2009. Warsaw: Warsaw University Press, pp. 299-318.
Scott, S. (1997) “The Power of Images in Late-Roman House” in Laurence, R. and Wallace-Hadrill, A. (eds.) Domestic Space in the Roman World: Pompeii and Beyond. JRA Supplementary Series 22. Portsmouth, pp. 53-67.
Segal, A. (1997) From Function to Monument: Urban Landscapes of Roman Palestine, Syria and Provincia Arabia. Exeter: The Short Run Press.
SegalMacDonald, W. L. (1982) The Architecture of the Roman Empire: An Urban Appraisal V.II. New Haven: Yale University Press.
Sertok, K. and Ergeç, R. (1999a) “A New Early Bronze Age Cemetery: Excavations near the Birecik Dam Southeastern Turkey Preliminary Report (1997-1998)”, Anatolica 25, pp. 87-107.
Sertok, K. and Ergeç, R. (1999b) “The Birecik Dam EBA Cemetery on the Middle Euphrates, Turkey. Preliminary Report of the 1997and 1998 Excavation Seasons”, Orient Express 3, pp. 86-89.
Seyrig, H. (1950) “Antiquités Syriennes”, Syria 27, pp. 229-252.
Smith, A. M. (2013) Roman Palmyra: Identity, Community and State Formation. New York: Oxford University Press.
Sodini, J. P. (2003) “Archaeology and Late Antique Social Structure” in Bowden, W. and Lavan, L. (eds.) Theory and Practice in Late Antique Archaeology. (Late Antique Archaeology 1). Boston: Brill Leiden, pp. 25-56.
353
Sodini, J. P. and Tate, G. (1984) “Maisons d’Epoque Romaine et Byzantine (IIe-VIe siècles) du Massif Calcaire de Syrie du Nord: Étude Typologique” Balty, J. (ed.) Apamée de Syrie, Bilan des Recherches Archéologiques 1973-1979, Aspects de l’Architecture Domestique d’Apamée, Actes du Colloque Tenu à Bruxelles les 29,30 et 31 Mai 1980. Brussels, pp. 377-430.
Sommer, M. (2004) “A Map of Meaning Approaching Cultural Identities at the Middle Euphrates (1st to 3rd Centuries AD)”, Egitto e Vicino Oriente 27, pp. 153-183.
Southern, P. (2008) Empress Zenobia. Palmyra’s Rebel Queen. London and New York: Continuum.
Spinola, G. (1996) “Note Sull'Evoluzione Planimetrica delle Domus della Cirenaica” Studi Miscellanei 29, pp. 281-292.
Stackelberg, K. T. (2017) The Roman Garden: Space, Sense, and Society. London: Routledge.
Stambaugh, J. E. (1988) The Ancient Roman City. Baltimore: Johns Hopkins University Press.
Starcky, J. and Gawlikowski, M. (1985) Palmyre. Paris: Librairie d'Amérique et d'Orient.
Stephens, R. J. (1937) "A Cuneiform Tablet from Dura-Europos", Revue d'Assyriologie 34, pp. 183-90.
Stillwell, R. (1941) “Outline of the Campaigns” in Stillwell, R. (ed.) Antioch-on-the-Orontes III: The Excavations of 1937-1939. Princeton: Princeton University Press, pp. 1-34.
Stillwell, R. (1961) “Houses of Antioch”, Dumbarton Oaks Papers, vol. 15, pp. 47-57.
Stoneman, R. (1992) Palmyra and Its Empire: Zenobia’s Revolt Against Rome. Ann Arbor: Michigan University Press.
Storey, G. R. (1997) “The Population of Ancient Rome”, Antiquity 71, pp. 966-978.
354
Stucchi, S. (1975) Architettura Cirenaica (Monografie di archeologia libica IX). Rome: L'Erma di Bretschneider.
Tate, G. (1992) Les Campagnes de la Syrie du Nord du IIe au VIIe siècle. Paris: Geuthner.
Tchalenko, G. (1953) Villages Antiques de la Syrie du Nord: Le Massif du Bélus à L'Époque Romaine I. Paris: Libraire Orientaliste.
Thébert, Y. (1987) “Private Life and Domestic Architecture in Roman Africa” in Veyne, P. (ed) A History of Private Life. Cambridge, Massachusetts and London: The Belknap Press of Harvard University Press, pp. 313-410.
Thébert, Y. (1993) “Private and Public Spaces: The Components of the Domus” in D’Ambra, E. (ed.) Roman Art in Context: An Anthology. Englewood Cliffs: Prentice Hall, pp. 208-237.
Thür, H. (2002) “Kontinuität und Diskontinuität im ephesischen Wohnbau der frühen Kaiserzeit”, in Berns, C. et al. (eds.) Patris und Imperium. Kulturelle und politische Identität in den Städten der römischen Provinzen Kleinasiens in der frühen Kaiserzeit, Kolloquium Köln, November 1998, BaBesch Suppl. 8, pp. 257-274.
Tobin, J. (2013) “The Houses: Domestic Architecture, Dated Deposits and Finds in Context” in Aylward, W. (ed.) Excavations at Zeugma, Conducted by Oxford Archaeology, vol I. Los Altos, California: The Packard Humanities Institute, pp. 71-118.
Trümper, M. (1998) Wohnen in Delos: Eine baugeschichtliche Untersuchung zum Wandel der Wohnkultur in hellenistischer Zeit (Internationale Archäologie). Rahden: Leidorf.
Tuori, K. and Nissin, L. (eds.) Public and Private in the Roman House and Society. JRA Supplementary Series 101. Portsmouth, pp. 177-186.
Uytterhoeven, I. (2009) Hawara in the Graeco-Roman Period. Life and Death in a Fayum Village. Leuven: Peeters Press.
355
Uytterhoeven, I. (2013) “Running Water. Aqueducts as Suppliers of Private Water Facilities in (Late) Roman Asia Minor”, in Wiplinger, G. (ed.) Historische Wasserleitungen. Gestern Heute Morgen. Tagungsband des Internationalen Frontinus-Symposiums. Wien, 19.-23. Oktober 2011, BaBesch Suppl. 24. Leuven, pp. 137-160.
Uytterhoeven, I. (2014) “Passages Full of Surprises. Circulation Patterns within the Late Antique Urban Elite Houses of Asia Minor”, in Kurapkat, D., Schneider, P.I. and Wulf-Rheidt, U. (eds.) Die Architektur des Weges. Gestaltete Bewegung im gebauten Raum. Internationales Kolloquium in Berlin vom 08. – 11. Februar 2012 veranstaltet vom Architekturreferat des DAI, Diskussionen zur Archäologischen Bauforschung 11, Regensburg. pp. 425-435.
Uytterhoeven, I. (2019) “A ‘Second Life: Transformation and Change in the Use of Space in the Late Antique Urban Elite Houses of Asia Minor, in Baldini, I. and Sfameni, C. (eds.), Abitare nel Mediterraneo Tardoantico. Atti del II Convegno Internazionale del Centro Interuniversitario di Studi sull’Edilizia abitativa tardoantica nel Mediterraneo (CISEM), Bologna 2-5 marzo 2016, Bari, pp. 273-280.
Uytterhoeven, I. (2020) “Investing in Luxury: Elite House Construction in Roman Imperial and Late Antique Pisidia and the Neighbouring Areas of Western Anatolia”, in Mörel, A. et al. (eds.) Production, Trade and Economy in Pisidia and Its Surrounding Areas. International Symposium Proceedings PISIDIAN STUDIES II, 31 October-03 November 2018, Isparta. pp. 331-361.
Uytterhoeven, I. (2022) “Splendid Luxury. Representation and Display in the Roman and Late Antique Urban Elite Houses of Asia Minor”, in Uytterhoeven, I. and Ricci, A. (eds.) Palimpsest of the House. reassessing Roman, Late Antique, Byzantine, and Early Islamic Living Patterns. İstanbul: Koç University Press. pp. 123-164.
Uytterhoeven, I. et al. (2009) “Urban housing in the Pisidian town of Sagalassos (Burdur, Turkey)”, in Ladstätter, S. and Scheibelreiter, V. (eds.) Städtisches Wohnen im östlichen Mittelmeerraum, 4. Jh. v. Chr. – 1. Jh. n. Chr., Akten des internationalen Kolloquiums vom24.–27. Oktober 2007 an der Österreichischen Akademie der Wissenschaften, DenkschrWien 397 /AF 18, Vienna. pp. 291-310.
Uytterhoeven, I. et al. (2014) “Late Antique Private Luxury. The Mosaic Floors of the Urban Mansion of Sagalassos (Ağlasun, Burdur)”, ÖJh 82, pp. 373-407.
356
Van Oyen, A. (2015) “Deconstructing and Reassembling the Romanization Debate Through the Lens of Postcolonial Theory: From Global to Local and Back?” Terra Incognita 6, pp. 205-226.
Vanesse, M. (2014) “Les Nymphées Tardifs de L’Architecture Domestique D’Apamée de Syrie: Mise en Scène de L’Eau et Expression d’un Status Symbol”, Meditterraneo Antico XVII-2, pp. 487-518. Venco Ricciardi, R. (1992) “Archaeological Research at Hatra, Preliminary Report on the 1989 Season”, Mesopotamia XXVII, pp. 189-198. Verboven, K. (2011) “Friendship among the Romans”, in Peachin, M. (ed.) The Oxford Handbook of Social Relations in the Roman World. Oxford: Oxford University Press, pp. 404-421.
Verhoogen, V. (1964) Apamée de Syrie aux Musées Royaux d’Art et d’Histoire. Brussels.
Veyne, P. (ed.) (1987) A History of Private Life: From Pagan Rome to Byzantium. Cambridge, Massachusetts and London: The Bleknap Press of Hravard University Press.
Von der Osten, H. H. (1930) Explorations in Hittite Asia Minor, 1929. Chicago: The Oriental Institute of the University of Chicago.
Von Gerkan A. (1939) “The Fortifications” in Rostovtzeff, M. I., Brown, F. E. and Welles, C. B. (eds.) The Excavations at Dura-Europos. Preliminary Reports of Seventh and Eighth Seasons of Work. 1933-1934 and 1934-1935. New Haven: Yale University Press, pp. 4-61.
Wagner, J. (1976) Seleuceia am Euphrat/Zeugma. Wiesbaden: Verlag.
Wallace-Hadrill, A. (1988) “The Social Structure of the Roman House” PBSR 56, pp. 43-97.
Wallace-Hadrill, A. (1989) “Patronage in Roman Society: From Republic to Empire” in Wallace-Hadrill, A. (ed.) Patronage in Ancient Society. London and New York: Routledge, pp. 63-88.
357
Wallace-Hadrill, A. (1991) “Houses and Households: Sampling Pompeii and Herculaneum” in Rawson, B. (ed.) Marriage, Divorce and Children in Ancient Rome, New York: Oxford University Press, pp. 191-227.
Wallace-Hadrill, A. (1994) Houses and Society in Pompeii and Herculaeneum. Princeton, New Jersey: Princeton University Press.
Wallace-Hadrill, A. (1997) “Rethinking the Roman Atrium House” in Laurence, A and Wallace-Hadrill, A. (eds.) Domestic Space in the Roman World: Pompeii and Beyond. JRA Supplementary Series 51, pp. 219-240.
Wallace-Hadrill, A. (2015) “What makes a Roman House a ‘Roman House’?” in Tuori, K. and Nissin, L. (eds.) Public and Private in the Roman House and Society. JRA Supplementary Series 101, pp. 177-186.
Walmsley, A. (2007) ‘Households at Pella, Jordan: Domestic Destruction Deposits of the Mid-8th C.’, in Lavan, L., Swift, E. and Putzeys, T. (eds.) Objects in Context, Objects in Use. Material Spatiality in Late Antiquity. Leiden: Brill. pp. 239–72.
Walmsley, A. (2012) Early Islamic Syria. An Archaeological Assessment. Debates in Archaeology. London, New York: Bloomsbury Publishing.
Ward-Perkins, J. B. et al. (1986) “Town Houses at Ptolemais, Cyrenaica: A Summary Report of Survey and Excavation Work in 1971, 1978–1979”, Libyan Studies 17, pp. 109-153.
Welles, C. B. (1933) “Graffiti” in Baur, P. V. C., Rostovtzeff, M. I. and Bellinger, A. R. (eds.) The Excavations of Dura-Europos: Preliminary Report of the Fourth Season October 1930-March 1931. New Haven: Yale University Press, pp. 79-177.
Welles, C. B. (1951) “The Population of Roman Dura” in Coleman-Norton, P. R. (ed.) Studies in Roman Economic and Social History in Honour of Allen Chester Johnson. Princeton, pp. 251-274.
Welles, C. B. (1956) “The Chronology of Dura-Europos”, EOS 48, pp. 464-469.
358
Welles, C. B., Fink, R. O. and Gilliam, J. F. (1959) The Excavations at Dura-Europos: Final Report, vol. I: The Parchments and Papyri. New Haven: Yale University Press.
Whitehouse, D. (2002) ‘The House of Siraf, Iran’, in Waines, D. (ed.) Patterns of Everyday Life. Aldershot: Ashgate. pp. 179–262.
Whittow, M. (2001) “Recent Research on the Late Antique City in Asia Minor: The Second Half of the 6th Century Revisited” in Bowden, W. and Lavan, L. (eds.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1). Boston: Brill Leiden, pp. 137-152.
Wickham, C. (2003) “Studying Long Term Change in the West ad 400-800” in Bowden, W. and Lavan, L. (eds.) Theory and Practice in Late Antique Archaeology (Late Antique Archaeology 1). Boston: Brill Leiden, pp. 385-403.
Wilber, D. N. (1938) “The Theatre at Daphne” in Stilwell, R. (ed.) Antioch-on-the-Orontes II: The Excavations of 1933-1936. Princeton: Princeton University Press, pp. 57-94.
Will, E. (1983) “Développement Urbain de Palmyre: Témoignages Épigraphiques Anciens et Nouveaux”, Syria 60, pp. 69-81.
Will, E. (1988) “La Population de Doura-Europos: une Evaluation”, Syria 65, pp. 315-321.
Wilson, A.I. (2012) “Water, Power and Culture in the Roman and Byzantine Worlds: An Introduction”, Water History 4.1, pp. 1-9.
Winter, F. (2006) Studies in Hellenistic Architecture. Toronto: University of Toronto Press.
Wiseman, T. P. (1987) “Conspicui Postes Tectaque Digna Deo: The Public Image of Aristocratic and Imperial Houses in the Late Republic and Early Empire”, L’Urbs: espace urbain et histoire (Ier siècle av. J.-C. – IIIe siècle ap. J.-C.). Actes du colloque international de Rome, 8-12 mai 1985, Publications de l’École française de Rome 98, pp. 393-413.
Woolf, G. (1994) “Becoming Roman, Staying Greek” Proceedings of the Cambridge Philological Society 40, pp. 116-143.
359
Woolf, G. (1995) “The formation of Roman Provincial Cultures” in Metzler, J. et al. (eds.) The Integration of the Early Roman empire. The role of culture and ideology. Luxembourg: Musée National d’Histoire et d’Art. pp. 9-18.
Woolf, G. (1998) Becoming Roman: the origins of provincial civilization in Gaul. Cambridge: Cambridge University Press.
Woolf, G. (2014) “Romanization 2.0 and its alternatives”, Archaeological Dialogues 21 pp. 45-50.
Woolf, G. (2021) “Taking the Long View. Romanization and Globalization in Perspective” in Belvedere and Bergemann (eds.) Imperium Romanum: Romanization between Colonization and Globalization. Palermo: Palermo University Press. pp. 19-32.
Wulf-Rheidt, U. (2011) “Die antiken Wohnhäuser von Pergamon”, in Grüßinger, R., Kästner, V. and Scholl, A. (eds.) Pergamon: Panorama der antiken Metropole. Begleitbuch zur Ausstellung: eine Ausstellung der Antikensammlung der Staatlichen Museen zu Berlin. Berlin, pp. 260-269.
Wulf-Rheidt, U. (2014) “The Ancient Residential Buildings of Pergamon/Antik Pergamon Evlerinde Yaşam”, in Pirson, F. and Scholl, A. (eds.) Pergamon. A Hellenistic Capital in Anatolia /Anadolu’da Hellenistik Bir Başkent. İstanbul: Yapı Kredi Yayınları. pp. 336-351.
Yaman, H. (2013) Zeugma Mezar Stelleri. Unpublished Phd. Thesis. Ankara: Ankara Üniversitesi.
Yon, J. B. (2002) Les Notables de Palmyre. Beyrouth: Presses de l’Ifpo.
Yon, J. B. (2008) “Bilinguisme et Trilinguisme à Palmyre” in Pouillou, J. (ed.) Bilinguisme Gréco-Latin et eÉpigraphie. Actes du Colloque Organisé à l'Université Lumière. Lyon: Maison de l'Orient et de la Méditerranée, pp. 195-211.
Zarmakoupi, M. (2010) “The Peristylium-Garden in Roman Luxury Villas: An Architectural and Cultural History”, Ladstätter, S. and Scheibelreiter, V. (eds.) Städtisches Wohnen im östlichen Mittelmeerraum, 4. Jh. v. Chr.– 1. Jh. n. Chr., Akten des internationalen Kolloquiums vom 24.–27. Oktober 2007, Vienna, pp. 621-631.
360
Zelazowski, J. (2012) “General Remarks on Polish Excavations in Ptolemais” in Zelazowski, J. (ed.) Ptolemais I. Warsaw: Institute of Archaeology, University of Warsaw. pp. 121-156.
Zimmer, T. (2010) “Repräsentatives Wohnen am Beispiel der Palastanlagen von Pergamon”, in Ladstätter, S. and Scheibelreiter, V. (eds.) Städtisches Wohnen im östlichen Mittelmeerraum, 4. Jh. v. Chr.– 1. Jh. n. Chr., Akten des internationalen Kolloquiums vom 24.–27. Oktober 2007 an der Österreichischen Akademie der Wissenschaften, DenkschrWien 397/AF 18, Vienna, pp. 155-166.
Zuchowska, M. (2000) “Quelques Remarques sur la Grande Colonnade à Palmyre”, Bulletin d'Etudes Orientales 52, pp. 187-193.
Zuchowska, M. (2008) “Wadi al Qubur and Its Interrelations with the Development of Urban Space of the City of Palmyra in the Hellenistic and Roman Periods”, in Kühne, H., Czichon, R. M. and Kreppner, F. J. (eds.) Proceedings of the Fourth International Congress of the Archaeology of the Ancient Near East. Vol. 1, pp. 141-153.
Zuchowska, M. (2011) “Space Organization and House Planning at Hellenistic and Roman Palmyra”, SWIATOWIT 9 (50)/A, pp. 141-153.
Ancient Sources
Ammianus Marcellinus (1940) Rerum Gestarum. Translated by J. C. Rolfe. Cambridge, MA: Harvard University Press.
Chrysostom, J. (2006) The Cult of the Saints: Select Homilies and Letters. Introduced, Translated and Annotated by Wendy Mayer. Crestwood, New York: St. Vladimir's Seminary Press.
Cicero, M. T. (1913) De Officiis. Translated by W. Miller. Cambridge and Massachusetts: Harvard University Press.
Libanius (1959) Antiochikos. Translated with commentary by G. Downey. Proceedings of the American Philosophical Society 103, pp. 652-686.
361
Malalas, J. (1986) The Chronicle of John Malalas. A Translation. Translated by E. Jeffreys. Melbourne: Brill.
Paterculus, C. V. (1924) Compendium of Roman Hıstory. Translated by Shipley, F. W. Cambridge and Massachusetts: Harvard University Press.
Petronius, A. (2020) Satyricon. Translated by Edited and translated by Gareth Schmeling. Cambridge and Massachusetts: Harvard University Press.
Pliny the Elder (1942) Natural History. Translated by H. Rackham. Cambridge and Massachusetts: Harvard University Press.
Pliny the Younger (1969) Letters. Vol. 1 Books.1-7. Translated by Betty Radice. Cambridge and Massachusetts: Harvard University Press.
Procopius (2002) De Aedificiis. Translated by H. B. Dewing. Cambridge and Massachusetts: Harvard University Press.
Strabo (1917) Geography. Translated by H. L.Jones. London: William Heinemann & New York: G.P. Putnam’s Sons.
Internet Sources
http://apamee.org/fr/le-site-approche-topographique (last access: June 2021)
http://monumentsofsyria.com/wpid899-2005-09-04-sl-20-apamea-house-of-consoles-jpg/ (last access: June 2021)
http://vrc.princeton.edu/archives (last access: June 2021)
http://vrc.princeton.edu/archives (last access: January 2020)
https://artmuseum.princeton.edu/collections (last access: December 2019)
https://artmuseum.princeton.edu/collections/objects/29551 (last access: January 2020)
362
https://commons.wikimedia.org/wiki/File:Great_Colonnade_at_Apamea,_Syria,_reconstruction_-_Cinquantenaire_Museum_-_Brussels,_Belgium_-_DSC09011.jpg (last access: June 2021)
https://commons.wikimedia.org/wiki/File:Musée_Cinquantenaire_Mosaïque_de_la-Chasse_01.jpg (last access: June 2021)
https://earth.google.com/web/ (last access: July 2021)
https://psl.eu/en/news/archives-french-syrian-mission-dura-europos-mfsed-online-soon-psl-explore (last access: March 2021)
https://spacedmm.com/pics/437/16-10%20226.JPG (last access: March 2021)
https://tr.m.wikipedia.org/wiki/Dosya:Qalat_el-Mudiq.jpg (last access: June 2021)
https://www.bbc.com/turkce/multimedya/2015/08/150824/gallery/palmyra (last access: June 2021)
https://www.khanacademy.org/humanities/ancient-art-civilizations/palmyra/a/temple-of-bel-palmyra (last access: Sptember 2020)
https://www.khanacademy.org/humanities/ancient-art-civilizations/palmyra/a/temple- of-baalshamin (last access: September 2020)
https://www.mediastorehouse.com/worldinprint/ruins-colonnade-1224471.html(last access: June 2021)
https://www.pinterest.com/pin/292804413265457383 (last access: December 2020)
https://www.romeartlover.it/Apamea2.html (last access: June 2020)
363
APPENDICES
A. FIGURES
Fig. 1 Map of the Roman East at its greatest extent. Five cities studied in the thesis are marked. (Edwell, 2008: 21)
364
Fig. 2 Geographical map showing the location of the five cities studied in the thesis (www. earth.google.com)
Fig. 3 Aerial photo of modern Antakya taken in 1930s. The city lays between the Orontes River and Mount Silpius (Kenfield and Moss, 2014: fig. 1)
365
Fig. 4 The Yakto Complex, The Megalopsychia mosaic Dunbabin (1999: fig 194)
Fig. 5 The Yakto Complex, Details from the topographical border of the Megalopsychia mosaic (Photos taken by the author in the Hatay Archaeology Museum)
366
Fig. 6 Antioch on the Orontes, Urban Plan (Becker and Kondoleon, 2005: xv)
367
Fig. 7 Antioch on the Orontes, Remains of the walls and towers in 1977, Engraving by L. F. Cassas, (Foss, 2000: 22)
Fig. 8 Antioch on the Orontes, Medina Gate in 1977, Engraving by L. F. Cassas, (Downey, 1961: 768, fig.21)
368
Fig. 9 Antioch on the Orontes, Urban scenes
Left: Details from the topographical border of Megalopsychia mosaic (Photos taken by the author in the Hatay Archaeology Museum)
Right: Snapshots by Lassus in modern Antakya in 1930’s (1935: Plates XII-XVIII)
369
Fig. 10 Daphne, Excavation photo of the House of the Buffet Supper in 1937 (http://vrc.princeton.edu/archives/items/show/15449)
Fig. 11 Daphne, Plan of the excavated residential sector (Levi II, 1947: Plan III)
370
Fig. 12 The Atrium House, Plan (after Fisher, 1934: Plate IV)
371
Fig. 13 The Atrium House, Computer generated reconstruction photograph of the mosaic pavement in room 82 (Becker and Kondoleon, 2005: 18, fig. 2)
Fig. 14 The House of the Calendar (highlighted section), Plan of the excavated part (The attached building to the north-east is the House of Drunken Dionysus) (http://vrc.princeton.edu/archives/items/show/14645)
372
Fig. 15 The House of the Calendar, Excavation photo of room 2 and the nymphaeum beyond (http://vrc.princeton.edu/archives/items/show/14109)
Fig. 16 The House of the Calendar, Mosaic pavement of room 2 (Photo taken by the author in the Hatay Archaeology Museum)
373
Fig. 17 The House of Cilicia, Plan of the excavated part (http://vrc.princeton.edu/archives/items/show/15781)
Fig. 18 The House of Cilicia, Excavation photo (http://vrc.princeton.edu/archives/items/show/15613)
374
Fig. 19 The House of the Drinking Contest, Plan (Gruber and Dobbins, 2010: fig. 1)
Fig. 20 The House of the Drinking Contest, Axonometric view (Gruber and Dobbins, 2010: fig. 3)
375
Fig. 21 The House of the Drinking Contest, Mosaic pavement of triclinium A (https://artmuseum.princeton.edu/collections/objects/29551)
Fig. 22 The House of the Drinking Contest, Mosaic pavement of portico B (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 23 The House of the Drinking Contest, Excavation photo with Mountain Casius in the background (http://vrc.princeton.edu/archives/items/show/17412)
376
Fig. 24 The House of Dionysus and Ariadne, Restored Plan (Levi I, 1947: fig. 56)
Fig. 25 The House of Dionysus and Ariadne, Excavation photo towards the harbor (Levi I, 1947: fig. 57)
377
Fig. 26 The House of Dionysus and Ariadne, Mosaic pavement of the reception room (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 27 The House of Dionysus and Ariadne, Mosaic pavement of the portico (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 28 The House of Dionysus and Ariadne, Excavation photo from triclinium towards the northern room (http://vrc.princeton.edu/archives/items/show/15764)
378
Fig. 29 The House of the Buffet Supper, Plan of the 3rd c. AD phase (After Stillwell, 1941: figs.32)
379
Fig. 30 The House of the Buffet Supper, Excavation photo of rooms C1, C2, C3 and the portico in front (http://vrc.princeton.edu/archives/items/show/16135)
Fig. 31 The House of the Buffet Supper, Mosaic pavement of room C2 (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 32 The House of the Buffet Supper, Mosaic pavement of room C3(Photo taken by the author in the Hatay Archaeology Museum)
380
Fig. 33 The House of the Buffet Supper, Mosaic pavement of the portico (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 34 C300 Complex, Excavation photo (http://vrc.princeton.edu/archives/items/show/16346)
381
Fig. 35 The House of the Buffet Supper, Plan of the 5th c. AD phase (After Stillwell, 1941: figs.33)
382
Fig. 36 Narrow streets between the house and the entrance doors in modern Antakya (Photos taken by the author)
Fig. 37 The House with Floral Pavements, Opus sectile floor pavement of room B13 (http://vrc.princeton.edu/archives/items/show/15499)
383
Fig. 38 The House with Floral Pavements, Plan of the late 5th c. AD phase (http://vrc.princeton.edu/archives/items/show/15783)
384
Fig. 39 The House of the Boat of Psyches, Restored plan by Levi (1947: fig. 63)
Fig. 40 The House of the Boat of Psyches, Restored plan by Stillwell (1961: fig. 16)
385
Fig. 41 The House of the Boat of Psyches, Plan by Kondoleon (2000: fig.5)
Fig. 42 The House of the Boat of Psyches, Mosaic pavement of room 1 (Levi II, 1947: Plate XXXVa)
386
Fig. 43 The House of the Boat of Psyches, Mosaic pavement of room 3 (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 44 The House of the Boat of Psyches, Mosaic pavement of corridor 4 (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 45 The House of the Boat of Psyches, Mosaic pavement of room 6 (http://vrc.princeton.edu/archives/items/show/13638)
387
Fig. 46 The House of Menander, Plan (Stillwell, 1941: plan VII)
388
Fig. 47 The House of Menander, Colored plan with five different sections (Dobbins, 2000: 50)
Fig. 48 The House of Menander, Excavation photo of the southeast suite with court 1 and room 3 in the foreground (http://vrc.princeton.edu/archives/items/show/16987)
389
Fig. 49 The House of Menander, Excavation photo of courtyard 1 with two layers of mosaic pavement (Levi I, 1947: fig.27)
Fig. 50 The House of Menander, Excavation photo of the late pool with rooms 11, 12 and 2 behind (http://vrc.princeton.edu/archives/items/show/16982)
390
Fig. 51 The House of Menander, Court 17 Top: Excavation photo (Levi II, 1947: Plate XLVIIc) Bottom: Mosaic pavement of the court as exhibited in the Hatay Archaeology Museum (Photo taken by the author)
391
Fig. 52 The Constantinian Villa, Plan of the excavated part (Levi, 1947: fig.85)
Fig. 53 The Constantinian Villa, Excavation photo (Levi I, 1947: fig. 86)
392
Fig. 54 The Constantinian Villa, Mosaic Pavement of Room 1 (Photo taken by the author in Louvre Museum)
Fig. 55 The House of Ge and the Seasons, Plan of the excavated part (Levi 1, 1947: fig. 139)
Fig. 56 The House of Ge and the Seasons, Plan of room 1 in the earlier phase (http://vrc.princeton.edu/archives/items/show/14648)
393
Fig. 57 The Yakto Complex, Plans of the 3rd c. AD and 5th c. AD phases (Lassus: 1938 figs. 4 and 5)
394
Fig. 58 The Yakto Complex, Plan of the 5th c. AD phase (Lassus, 1938: Plan IX)
395
Fig. 59 The Yakto Complex, Mosaic pavement of corridor 5 (Lassus, 1938: fig. 8)
Fig. 60 The Yakto Complex, Central panel of opus sectile pavement of room 52 (Lassus, 1938: fig. 32)
Fig. 61 The Yakto Complex, Excavation photo of room 10 with nymphaeum at the back (Lassus, 1938: fig. 39)
396
Fig. 62 The Yakto Complex, Mosaic pavement of Room A (Photo taken by the author in the Hatay Archaeology Museum)
Fig. 63 The Yakto Complex, Mosaic pavement of Room B (Photo taken by the author in the Hatay Archaeology Museum)
397
a. Parthian Palace excavated in Assur
(Andrae& Lenzen, 1933: Taf. 11)
b. Parthian house in insula H10 excavated in Assur (Andrae& Lenzen, 1933: Taf. 11)
c. House A excavated in (Venco Ricciardi, 1992: Plate 1)
Fig. 64 Parthian Palace Plans
398
Fig. 65 Apamea on the Orontes, The google earth view of the site (www.earth.google.com)
Fig. 66 Apamea on the Orontes, The Acropolis Hill (Modern Qal’at al Mudiq) (https://tr.m.wikipedia.org/wiki/Dosya:Qalat_el-Mudiq.jpg)
399
Fig. 67 Apamea on the Orontes, Urban Plan (The location of the excavated houses are highlighted) (After Balty, J., 1984: fig.3)
400
Fig. 68 Apamea on the Orontes, The Colonnaded Street (https://www.stevensklifas.com/apamea-syria-hellenistic-ancient-city)
Fig. 69 The Grand Colonnade of Apamea on the Orontes, Reconstruction in Cinquantenaire Museum in Brussels, Belgium (https://commons.wikimedia.org/wiki/File:Great_Colonnade_at_Apamea,_Syria,_reconstruction_-_Cinquantenaire_Museum_-_Brussels,_Belgium_-_DSC09011.jpg)
401
Fig. 70 The Building with Triclinos, Plan (After Balty, J. Ch., 1984: fig. 2)
402
Fig. 71 The Building with Triclinos, Capital of the pilaster in the north gallery of the peristyle (Balty, 1969: Plate XLI.1)
Fig. 72 The Building with Triclinos, Marble wall decoration of room Q (Balty, 1969: Plate XLIV.2)
Fig. 73 The Building with Triclinos, Schematic drawing of the mosaic pavement in room AB (Balty, 1969: 110, fig.3)
Fig. 74 Scene from the mosaic pavement of room AB (Balty, 1969: Plate XLII.1)
403
Fig. 75 The Building with Triclinos, Amazon mosaic in room T (Dulière, 1968: Plate I)
Fig. 76 The Building with Triclinos, Staircase in room M (Balty, 1969: Plate XXXIX)
404
Fig. 77 The Building with Triclinos, Hunting mosaic in room (https://commons.wikimedia.org/wiki/File:Musée_Cinquantenaire_Mosaïque_de_la_ Chasse_01.jpg)
Fig. 78 The Building with Triclinos, Inscription on the threshold of room A (https://spacedmm.com/pics/437/16-10%20226.JPG)
405
Fig. 79 Insula plan with three adjoining houses (After Balty, J. Ch., 1984: fig. 1)
Fig. 80 The House of the Consoles, Plan showing the construction phases (Balty, J., 1984: fig. 1)
406
Fig. 81 The House of the Consoles, Plan (After Balty, J. Ch., 1984: fig. 2)
407
Fig. 82 The House of the Consoles, Western façade with the entrance door (http://apamee.org/fr/le-site-approche-topographique)
Fig. 83 The House of the Consoles, Details from the entrance door (https://www.romeartlover.it/Apamea2.html)
408
Fig. 84 The House of the Consoles, Peristyle courtyard (http://monumentsofsyria.com/wpid899-2005-09-04-sl-20-apamea-house-of-consoles-jpg/)
Fig. 85 The House of the Consoles, Opus sectile floor pavement of Room A’ (Balty, J., 1984: Plate X.1-2)
409
Fig. 86 The House of the Consoles, Marble and painted plaster wall coverings of Room A-A’ (Balty, J., 1984: Plate XIV.1-2)
Fig. 87 The House of the Consoles, North gallery with the apse at its end (Balty, J., 1984: Plate III.2)
410
Fig. 88 The House of the Pilasters, Plan (After Balty, J. Ch., 1984: fig. 2)
Fig. 89 The House of the Pilasters, Plan showing construction phases (Gisler and Huwiler, 1984: fig.1)
411
Fig. 90 The House of the Pilasters, Reconstituted plan of the initial construction (Gisler and Huwiler, 1984: fig.2)
Fig. 91 The House of the Pilasters, Room A
Left: Axonometric drawing (Gisler and Huwiler, 1984: fig.4)
Right: Steps with column bases (Gisler and Huwiler, 1984: Plate XXXbis.1)
412
Fig. 92 The House of the Pilasters, Column bases between rooms L and T (Gisler and Huwiler, 1984: Plate XXX.1)
Fig. 93 The House of the Pilasters, Niches on the west wall of room T (Gisler and Huwiler, 1984: Plate XXX.2)
413
Fig. 94 The House of the Console Capitals, Plan (After Balty, J. Ch., 1984: fig. 2)
414
Fig. 95 The House of the Console Capitals, Plan showing construction phases (Baratte, 1984: fig.1)
Fig. 96 The House of the Console Capitals, Polylobed basin in the peristyle courtyard (After Baratte, 1984: Plate XXXIV.1)
415
Fig. 97 The House of the Console Capitals, A console capital found in the courtyard (Baratte, 1984: Plate XLV.2)
Fig. 98 The House of the Console Capitals, Late cistern in the peristyle courtyard (Baratte, 1984: Plate XLIII.1)
Fig. 99 The House of the Console Capitals, Later wall built with column shafts (Baratte, 1984: Plate XLI.2)
416
Fig. 100 The House of the Deer, Plan (Donnay-Rocmans and Donnay, 1984: fig.2)
Fig. 101 The House of the Deer, Details of the column bases in the peristyle (Donnay-Rocmans and Donnay, 1984: figs.7, 8)
417
Fig. 102 The House of the Deer, Schematic drawing of the mosaic of room A (Donnay-Rocmans and Donnay, 1984: fig. 5)
Fig. 103 The House of the Deer, Green marble top found in room A (Donnay-Rocmans and Donnay, 1984: Plate LII.2)
418
Fig. 104 The House of the Deer, Mosaic pavement of room C (Donnay-Rocmans and Donnay, 1984: Plate LIV.2)
Fig. 105 The House of the Deer, Mosaic pavement of room F (Donnay-Rocmans and Donnay, 1984: Plate LV.2)
Fig. 106 The House of the Deer, White marble table top found in room F (Donnay-Rocmans and Donnay, 1984: fig. 6 and Plate LII.1)
419
Fig. 107 The House of the Deer, East wall of room A with windows opening to courtyard D (Donnay-Rocmans and Donnay, 1984: Plate LI.3)
Fig. 108 The House of the Deer, Plan showing later constructions (Donnay-Rocmans and Donnay, 1984: fig. 1)
420
Fig. 109 The House with the Bilobed Columns, Two excavation photos taken in 1934 and 1970 (Mayence, 1935: fig. 10& Balty, 1972: Plate II)
Fig. 110 The House with the Trilobed Columns, Plan (Raepsaet-Charlier: 1984: fig. 1)
421
Fig. 111 The House with the Trilobed Columns, Mosaic pavement of the western portico (Raepsaet-Charlier: 1984: Plate LXI)
Fig. 112 The House with the Trilobed Columns, Opus sectile pavements of rooms O&M (Raepsaet-Charlier: 1984: Plate LXIII)
422
a. Ephesus, The Villa above the Theatre-Ephesus (Ellis, 1991: fig. 6.1)
b. Xanthos, The Northeast House on the Lycian Acropolis (Manière-Lévêque, 2002: 239)
c. Perge, The Late Antique Residence (Abbasoğlu, 2001: 184)
Fig. 113 Asia Minor, Late antique house plans with apsidal audience halls
423
Fig. 114 Seleucia and Apamea on the Euphrates, Reconstruction model (Görkay, 2020a: fig. 3)
Fig. 115 Apamea on the Euphrates, Urban plan (Abadie-Reynal, Ergeç and Bucak, 1999: fig. 17)
424
Fig. 116 Zeugma, Aerial photos taken in 2000 showing the territory of the ancient city (Başgelen, 2000: 5-6)
Fig. 117 Seleucia and Apamea on the Euphrates, Urban plan showing the borders and fortifications of the city in the Hellenistic and Roman periods (Görkay, 2020a: fig. 1)
425
Fig. 118 Zeugma, Belkıs Tepe as seen from the modern road approaching the site (Photo taken by the author at the archaeological site)
Fig. 119 Roman period coin depicting Belkıs Tepe and the temple (Görkay, 2015: 32)
426
Fig. 120 Zeugma, Urban plan (Locations of the excavated houses studied or mentioned in the thesis are highlighted) (After Aylward, 2013a: Plate 2)
427
Fig. 121 The House of the Fountain, Plan of the excavated part (Tobin, 2013: Plate 77)
Fig. 122 The House of the Fountain, Mosaic panel of room 11D (Photo taken by the author in the Zeugma Mosaic Museum)
428
Fig. 123 The House of the Helmets and The House of the Bull, Plan of the excavated part (Tobin, 2013: Plate 10)
Fig. 124 The House of the Helmets, Peristyle courtyard (Tobin, 2013: Plate 26-C)
429
Fig. 125 The House of the Bull, Room 2M (Aylward, 2013: Plate 38-A)
Fig. 126 The House of the Bull, Room 2K (Tobin, 2013: Plate 36-A)
Fig. 127 The House of the Bull, Entrance of room 2L from room 2K (Tobin, 2013: Plate 34-C)
430
Fig. 128 The House of Dionysus and Ariadne and the The House of Danae, Sheltering Structure (Photo taken by the author at the archaeological site of Zeugma)
Fig. 129 The House of Dionysus and Ariadne (on the left) and the The House of Danae (on the right), View towards south (Photo taken by the author at the archaeological site of Zeugma)
Fig. 130 The House of Dionysus and Ariadne, View towards north overlooking the river (Photo taken by the author at the archaeological site of Zeugma)
431
Fig. 131 The House of Dionysus and Ariadne and The House of Danae, Plan of the excavated part (Önal, 2005: fig. 1)
432
Fig. 132 The House of Dionysus and Ariadne, Reconstruction plan (Photo taken by the author at the archaeological site of Zeugma)
Fig. 133 The House of Dionysus and Ariadne, Peristyle courtyard (D3/13) (Photo taken by the author at the archaeological site of Zeugma)
433
Fig. 134 The House of Dionysus and Ariadne, Three-dimensional reconstruction of the peristyle courtyard with the impluvium (Görkay, 2020b: fig. 1)
Fig. 135 The House of Dionysus and Ariadne, Room D2/16 and the entrance door of room D1/19 (Photo taken by the author at the archaeological site of Zeugma)
434
Fig. 136 The House of Dionysus and Ariadne, Mosaic pavement of room D1/19 (Photo taken by the author in the Zeugma Mosaic Museum)
Fig. 137 The House of Dionysus and Ariadne, Mosaic Pavement of room D2/16 (Ergeç, 2000: 21)
435
Fig. 138 The House of Dionysus and Ariadne, Three-dimensional reconstruction of room D1/19 (Görkay, 2020b: fig. 6)
Fig. 139 The House of Dionysus and Ariadne, Rooms D11 and D12 (Photo taken by the author at the archaeological site of Zeugma)
436
Fig. 140 The House of Dionysus and Ariadne, Peristyle courtyard and room 6 in the upper level (Photo taken by the author at the archaeological site of Zeugma)
Fig. 141 The House of Dionysus and Ariadne, The two columns behind the southern portico of the courtyard and the cistern on the south wall (Photo taken by the author at the archaeological site of Zeugma)
Fig. 142 The House of Dionysus and Ariadne, Reused column drums in the wall of room D20/5 (Photo taken by the author at the archaeological site of Zeugma)
437
Fig. 143 Plan of the contiguous houses in Trench 8 (The House of Poseidon, The House of Euphrates and The House without Mosaics) (After Önal, 2013: Plan 2)
438
Fig. 144 The House of Poseidon, Plan with mosaic pavements (Önal, 2013: plan 4)
Fig. 145 The House of Poseidon, Reconstruction of peristyle A6/P9 (Photo taken by the author in the Zeugma Mosaic Museum)
439
Fig. 146 The House of Poseidon, Fountain in peristyle A6/P9 (Photo taken by the author in the Zeugma Mosaic Museum)
Fig. 147 The House of Poseidon, Three-dimensional reconstruction of the fountain (Barbet, 2005: Plate XVIII)
440
Fig. 148 The House of Poseidon, Court A2/P4 with the impluvium (Önal, 2013: fig. 21)
Fig. 149 The House of Poseidon, Three-dimensional reconstruction of court A2/P4 (Görkay, 2020b: fig. 4)
441
Fig. 150 The House of Poseidon, Three-dimensional reconstruction of triclinium A1/P3 (Görkay, 2020a: fig. 9)
Fig. 151 The House of Poseidon, Mosaic pavement of triclinium A1/P3 (Photo taken by the author in the Zeugma Mosaic Museum)
442
Fig. 152 The House of Poseidon, Three-dimensional reconstruction of room A5/P5 (Görkay, 2020d: fig. 20)
Fig. 153 The House of Poseidon , Mosaic pavement of room A5/P5 (Önal, 2013: 35, fig. 33)
443
Fig. 154 The House of Poseidon, Wall paintings of room A11/P6 (Önal, 2013: fig. 38)
Fig. 155 The House of Poseidon, Mosaic pavement of room A14/P38 (Photo taken by the author in the Zeugma Mosaic Museum)
444
Fig. 156 The House of Poseidon, Mosaic pavement of room A13/P37 (Photo taken by the author in the Zeugma Mosaic Museum)
Fig. 157 The House of Poseidon, Three-dimensional reconstruction of room A13/P37 (Görkay, 2020e: fig. 58)
445
Fig. 158 The House of Poseidon, Peristyle courtyard B6/P13 (Barbet, 2005: Plate. VII)
Fig. 159 The House of Poseidon, Reconstruction of peristyle courtyard B6/P13 in the Zeugma Mosaic Museum (Photos taken by the author)
446
Fig. 160 The House of Poseidon, Rooms B1/P10, B2/P11, B3/P12 and B4/P23 (Önal, 2013: fig. 53)
Fig. 161 The House of Poseidon, Room B1/P10 towards south (Önal, 2013: fig. 92)
447
Fig. 162 The House of Poseidon, Three-dimensional reconstruction of room B2/P11 (Görkay, 2020e, fig. 63)
Fig. 163 The House of Poseidon, Mosaic pavement of room B2/P11 (Photo taken by the author in the Zeugma Mosaic Museum)
Fig. 164 The House of Poseidon, Wall paintings of room B2/P11 (Önal, 2013: fig. 88)
448
Fig. 165 The House of Poseidon, View from room B2/P11 towards rooms B3/P12 and B4/P23 (Önal, 2013: fig. 87)
Fig. 166 The House of Poseidon, Doors opening to room B9/P27 from the courtyard (Önal, 2013: fig. 95)
Fig. 167 The House of Poseidon, Statue of Mars as found in the store room (Önal, 2013: fig. 99)
449
Fig. 168 The House of Poseidon, Latrine (Önal, 2013: fig. 105)
Fig. 169 The House of Euphrates, Plan with mosaics (Önal, 2013: 90)
450
Fig. 170 The House of Euphrates, Corridor C3/P21 towards south (Önal, 2013: fig. 113)
Fig. 171 The House of Euphrates, Drawing of graffiti on the wall of corridor C3/P21 (Önal, 2013: fig. 114)
Fig. 172 The House of Euphrates, Mosaic pavement of triclinium C1/P19 (Photo taken by the author in the Zeugma Mosaic Museum)
Fig. 173 The House of Euphrates Impluvium court C2/P20 (Önal, 2013: 109, fig. 132)
451
Fig. 174 The House of Euphrates, Three-dimensional reconstruction of impluvium court C2/P20 (Görkay, 2020c: fig. 5)
Fig. 175 The House of Euphrates, Paintings on the eastern wall of impluvium court C2/P20 (Önal, 2013: 109, fig. 133)
Fig. 176 The House of Euphrates, Mosaic pavement of room C8/P17 (Önal, 2013: fig. 122)
452
Fig. 177 The House of Euphrates, Rooms C4/P16, P2/C15 and P15/C5 (the latter two provided passage between C2/P20 and C6/P22) (Önal, 2013: fig. 112)
Fig. 178 The House of Euphrates, Reconstruction of peristyle courtyard C6/P22 (Photo taken by the author in the Zeugma Mosaic Museum)
453
Fig. 179 The House of Euphrates, View from room A11/P6 of the The House of Poseidon towards peristyle C6/P22, with rooms C13/P26 and C12/P25 in the background (Önal, 2013: 120, fig. 151)
Fig. 180 The House of Euphrates, Wall paintings of room C13/P26 (Photo taken by the author in the Zeugma Mosaic Museum)
454
Fig. 181 The House without Mosaics, Section of the eastern part of the building (Önal, 2013: fig. 33)
Fig. 182 The House without Mosaics, Wall paintings on the southern wall of room D6/P30 (Önal, 2013: fig. 172)
Fig. 183 The House of Synaristosai, Schematic plan of the house(s) in the 2nd c. AD (Abadie-Reynal, 2012a: Plate II)
455
Fig. 184 Rural house dated between 2nd – 6th c. AD in Northern Syria (Tchalenko, 1953: Plate V)
456
Fig. 185 The House of Synaristosai, Schematic plan in the first half of the 3rd c. AD (Abadie-Reynal, 2012a: Plate III)
Fig. 186 The House of Synaristosai, Plan showing walls of subsequent constructions (Abadie-Reynal, 2012a: Plate I)
457
Fig. 187 The House of Synaristosai, Partly hypothetical axonometric drawing (Abadie-Reynal, 2012a: pl. VII)
Fig. 188 The House of Synaristosai, Three-dimensional reconstruction towards south-east (Abadie-Reynal, 2012a: Plate VIIIa)
458
Fig. 189 The House of Synaristosai, Excavation photo of room P4 (Abadie-Reynal, 2012a: fig. 15)
Fig. 190 The House of Synaristosai, Three-dimensional reconstruction of room P4 towards south (Abadie-Reynal, 2012a: Plate IX)
459
Fig. 191 The House of Synaristosai, Excavation photo of the eastern part with room P1 in the foreground (Abadie-Reynal, 2012a: fig. 4)
Fig. 192 The House of Synaristosai, Excavation photo of Room P9 (Abadie-Reynal, 2012a: fig. 21)
460
Fig. 193 The House of Synaristosai, Three-dimensional reconstruction of room P2 towards south-east, (Abadie-Reynal, 2012a: Plate X)
Fig. 194 The House of Synaristosai, Mosaic pavement of room P13 (Photo taken by the author in the Zeugma Mosaic Museum)
461
Fig. 195 The House of Synaristosai, Excavation photo of rooms P8 and P13 (Abadie-Reynal, 2012a: fig. 36)
Fig. 196 The House of Synaristosai, Three-dimensional reconstruction of rooms P8, P13 and the upper floor (Abadie-Reynal, 2012a: Plate XI)
462
Fig. 197 The House of Synaristosai, Three-dimensional reconstruction of rooms P13-P11 (Abadie-Reynal, 2012a: Plate XIIa)
Fig. 198 The House of Synaristosai, Excavation photos of room P8 and the apse (Abadie-Reynal, 2012a: 51, figs. 35&39)
463
Fig. 199 The House of Synaristosai, Excavation photo of the western part of the building (Abadie-Reynal, 2012a: fig. 56)
Fig. 200 The House of Synaristosai, Three-dimensional reconstruction of court P3 (Abadie-Reynal, 2012a: Plate XIIb)
464
Fig. 201 The House of Synaristosai, Three dimensional reconstruction of court P17 (Abadie-Reynal, 2012a: Plate XIII)
Fig. 202 The House of Synaristosai, Schematic plan of the later constructions (Abadie-Reynal, 2012a: Plate V)
465
Fig. 203 The Late Imperial Peristyle House, Plan of Trench 7B with the house in 7B (Tobin, 2013: Plate 96)
Fig. 204 The Late Imperial Peristyle House, Excavation photo of the courtyard towards southeast (Tobin, 2013: Plate 103A)
466
Fig. 205 Palmyra, Urban plan showing different urban quarters (Zuchowska, 2011: 142)
467
Fig. 206 Palmyra, Plan of the Hellenistic quarter, the structures marked with gray are a set of four houses with courtyards (Al As’ad, Schmidt-Colinet, 2000: fig 3)
Fig. 207 Palmyra, Urban plan of the south-western quarter (Grassi, 2008:6)
468
Fig. 208 Palmyra, Grand Colonnade with Tetrapylon at the back (https://www.pinterest.com/pin/292804413265457383)
Fig. 209 Palmyra, Consoles on the columns that carried statues (https://www.mediastorehouse.com/worldinprint/ruins-colonnade-1224471.html)
Fig. 210 Palmyra, Temple of Bel (https://www.khanacademy.org/humanities/ancient-art-civilizations/palmyra/a/temple-of-bel-palmyra)
Fig. 211 Palmyra, Temple of Baalshamin (https://www.khanacademy.org/humanities/ancient-art-civilizations/palmyra/a/temple-of-baalshamin)
469
Fig. 212 Palmyra, Urban plan showing the location of the houses that are examined in the thesis (After Zuchowska, 2011: fig. 1)
470
a. House 38 b. House 45
c. House 39
Fig. 213 House Plans by Gabriel (1926: figs. 3, 5& Plate XV)
Fig. 214 Plan of Houses 39, 40 and a third house (Krencker, 1932: Plate 19)
471
a. Plan of the first phase (2nd c. AD)
b. Longitudinal section
c. Three-dimensional reconstruction
Fig. 215 House F (Gawlikowski, 2007: figs. 6, 14, 13)
472
Fig. 216 House F, Excavation photo (Gawlikowski, 2007: fig.12)
Fig. 217 House F, Plan of the latest phase (7th-9th c. AD) (Gawlikowski, 1996: fig.2)
473
Fig. 218 The Peristyle House, Plan (Grassi, Rocca and Piacentini, 2015: fig. 2)
Fig. 219 The Peristyle House, Excavation photo (Grassi, Rocca and Piacentini, 2015:fig. 3)
474
Fig. 220 The Peristyle House, Reused inscribed block and small altar (Grassi, Rocca and Piacentini, 2015: figs. 11& 18)
Fig. 221 The House(s) of Achilles and Cassiopeia, Plan (Gawlikowski, 2007: fig. 4)
475
Fig. 222 The House(s) of Achilles and Cassiopeia, Plan (Delplace, 2013: 44)
Fig. 223 The House(s) of Achilles and Cassiopeia, Peristyle courtyard 17 (Delplace, 2013: fig. 2)
476
Fig. 224 The House to the Southeast of the Theatre, Plan (Gawlikowski, 2007: fig. 5)
Fig. 225 The House to the Southeast of the Theatre, Aerial view of the remains of the house (in the foreground), the Theatre and the Colonnaded Street (in the background) (https://www.bbc.com/turkce/multimedya/2015/08/150824_gallery_palmyra)
477
Fig. 226 The Houses to the East of the Temple of Nabu, Plan (Saliby, 1996: fig.7)
478
Fig. 227 Dura-Europos, Aerial photo of the ruins of the ancient city (Archives from the French-Syrian Mission to Dura-Europos (MFSED)
Fig. 228 Dura-Europos, Isometric reconstruction of the Hellenistic city by Rostovtzeff (1938, 11-fig. 5)
Fig. 229 Dura-Europos, Isometric reconstruction of the Roman city by Perkins (1973: fig. 8)
479
Fig. 230 Dura-Europos, Urban plan (Baird, 2014: 9)
480
Fig. 231 Dura-Europos, Hellenistic agora (Brown, 1944a: fig. 9)
Fig. 232 Dura-Europos, Agora in the 3rd c. AD (Brown, 1944a: fig. 78)
Fig. 233 Dura-Europos, The military camp (Hopkins and Rowell, 1934: Plate III)
481
Fig. 234 Dura-Europos, Urban plan The houses examined in the thesis are highlighted (After Baird, 2014: fig. 1.4)
482
Fig. 235 The Citadel Palace, Aerial view of the remains in 1929 (Pillet, 1931: Plate XXIX)
Fig. 236 The Citadel Palace, Plan (Pillet, 1931: Plate IV)
Fig. 237 The Citadel Palace, Hypothetical reconstruction plan by F. Brown (Rostovtzeff, 1938: fig. 9)
483
a. Palace, plan
(Bernard, 1976: fig. 3)
b. The house outside the walls, plan c. The house in the southern sector, plan
(Bernard, 1974: fig. 1) (Bernard, 1971: fig. 11)
Fig. 238 Ai Khanoum, Hellenistic residential buildings
484
Fig. 239 The Redoubt Palace, Plan (Pillet, 1933: Plate III)
485
Fig. 240 The Redoubt Palace/Strategeion, Three-dimensional reconstruction (Rostovtzeff, 1938: Plate V)
Fig. 241 The Redoubt Palace/Strategeion, Plan (Hoepfner and Schwander, 1986: fig. 216)
486
Fig. 242 Insula C7, Plan (Baird, 2014: fig. 6.2)
Fig. 243 Insula C7, Housing Plots (Saliou, 2005: fig. 5)
487
Fig. 244 Insula C7, Plan (After Baird, 2014: fig 6.2)
488
Fig. 245 Insula C7, Plan of the latest phase of occupation (After Baird, 2014: fig 6.2)
489
Fig. 246 House C7A, Courtyard with cistern/cesspool (A), cooler (B) and trough (C)
(Baird, 2006: fig. 70)
Fig. 247 House C7A, Cooler next to door from courtyard to room 7 (Baird, 2006:
fig. 71)
Fig. 248 House C7B2, The door between room 2 and room 4, the plaster cornice
above the door (A) and bench on the floor (B). (Baird, 2006: fig. 78)
490
Fig. 249 House C7C, View of room 14 from the courtyard (A is the entrance of
room 4, B and E are the column and central pier against the wall, C is the storage jar
and D is the 'flag-pole base')
Fig. 250 House C7C, Graffiti from room 4 (Hopkins, 1934: Plates
XXXIII&XXXIV)
Fig. 251 House C7C2, Niches in the place of the blocked door of room 3 (Baird,
2006: fig. 87)
491
Fig. 252 House C7F, Fresco on the wall of room 8 (Rostovtzeff, 1935: fig. 82)
Fig. 253 House C7G, View from court towards the entrance of room G4 (Baird,
2006: fig. 95)
Fig. 254 Houses C7G2 and C7G3, Excavation photo from the west (Baird, 2006: fig.
100)
492
Fig. 255 Insula G1, Plan (After Brown, 1944c: fig. 80)
Fig. 256 Houses G1A and G1B, Excavation photo taken from the east (Brown, 1944c: Plate XII)
493
Fig. 257 House G1A, Plan (Brown, 1944c: fig. 61)
Fig. 258 House G1B, Plan (Brown, 1944c: fig. 65)
494
Fig. 259 Insula G3, Plan (After Brown, 1944c: fig. 79)
495
Fig. 260 Houses G3B-G3C-G3G, Early phase of occupation (Brown, 1944c: fig.30)
Fig. 261 Houses G3B-G3C-G3G, Later phase of occupation (Brown, 1944c: fig.31)
496
Fig. 262 Insula B2, Excavation photo from south-west (Allara, 2002: fig. 31)
Fig. 263 Insula B2, Plan (Allara, 2002: fig. 35)
497
Fig. 264 House B2C, Plan (Allara, 2002: fig. 43)
Fig. 265 House B2C, Excavation photo from north-west (Allara, 2002: fig. 44)
498
Fig. 266 House B2A, Plan (Allara, 2002: fig. 96)
Fig. 267 House B2A, Excavation photo from north-west (Allara, 2002: fig. 97)
499
Fig. 268 The House of Lysias (Insula D1), Plan (Baird, 2014: fig. 6.6)
500
Fig. 269 The House of Lysias, North entrance. (Baird, 2006: fig. 410)
Fig. 270 The House of Lysias, Plan with different sections of the house (Pontbriand,
2012: 84-fig. 10)
501
Fig. 271 The House of Lysias, Plan with different parts of section A (Pontbriand, 2012: 84-fig. 11)
Fig. 272 The House of Lysias, View of Courtyard 1 from north-east (Baird, 2014: fig.6.7)
502
Fig. 273 The House of Lysias, View of Courtyard 1 towards south (Baird, 2014: fig.6.9)
Fig. 274 The House of Lysias, Fanciful reconstruction of south side of the courtyard 1 by Herbert Gute (Baird, 2014: fig. 6.12)
Fig. 275 The House of Lysias, Stables (Pontbriand, 2016: 123-fig. 6)
503
Fig. 276 The House of Lysias, Sections (Pontbriand, 2016: figs. 7-8)
Fig. 277 The House of Lysias, Axonometric reconstruction (Pontbriand, 2016: fig. 10)
504
Fig. 278 The House of the Large Atrium/Insula D5, Plan showing housing units in early phases of occupation (After Baird, 2014: fig: 6.3)
505
Fig. 279 The House of the Large Atrium/Insula D5, Plan showing housing units in latest phase of occupation (After Baird, 2014: fig: 6.3)
506
Fig. 280 The House of the Large Atrium/Insula D5, Staircase F11 (Baird, 2006: fig. 184)
Fig. 281 The House of the Large Atrium/Insula D5, Central door from court F1 to room F4 (Baird, 2006: fig.186)
Fig. 282 The House of the Large Atrium/Insula D5, Niche on the south wall of room F4 (Baird, 2006: fig. 182)
507
Fig. 283 The House of the Large Atrium/Insula D5, Terracotta tiles (Pillet: 1933, Plate VI-3, 4)
Fig. 284 The House of the Large Atrium/Insula D5, Friezes from room F4 (Baird, 2006: figs. 135, 136, 139, 140)
Fig. 285 The House of the Large Atrium/Insula D5, Courtyard D1 (Baird, 2006: fig. 162)
508
Fig. 286 House E4, Plan (Baird, 2014: fig. 3.12)
Fig. 287 House E4, North-east corner of courtyard 14 (Baird, 2006: fig. 199)
509
Fig. 288 Insula M8, Plan (Kraeling, 1967: plan II)
Fig. 289 The Christian House, Plan (Baird, 2006: 500-fig. 392)
510
Fig. 290 The Christian House, Graffiti on the south wall of room 4B (Goldman,
1999, figs. 13b&14b)
Fig. 291 The Christian House, View of rooms 4B and 4A with windows on the back
wall (Güney, 2012: fig. 37)
511
Fig. 292 The Christian House, Plan of the building after its conversion into a church (Kraeling, 1967: fig. 1)
Fig. 293 The Christian House, Isometric Drawing of the building after its conversion into a church (Kraeling, 1967: plan III)
512
Fig. 294 The Roman Palace/The Palace of Dux Ripae, Plan (Detweiler, 1952: fig. 7)
Fig. 295 The Roman Palace, Longitudinal Sections (Detweiler., 1952: fig. 8)
513
Fig. 296 The Roman Palace, Apse of room 2 during excavation (Detweiler, 1952: Plate III. 1)
Fig. 297 Roman Palace/Palace of Dux Ripae, Painted Ceiling Patterns of Rooms 3, 6, 19 and 61 on the left and of rooms 12 and 13 on the right (Detweiler, 1952: Plates VIIII& IX)
Fig. 298 The Roman Palace, Apse 21 during excavation(Detweiler, 1952: Plate V. 2)
514
Fig. 299 The Roman Palace, Isometric reconstitution (Detweiler, 1952: fig. 2)
Fig. 300 The Roman Palace, Reconstructed Plan of the northern section (Downey,
1993: fig. 5)
515
Fig. 301 Dura-Europos, Isometric reconstruction of an insula (Hoepfner and Scwander, 1986: fig. 222)
Fig. 302 Dura Europos, Geophysical map of insula M3 (Benech 2010, fig. 6a)
516
Fig. 303 Dura Europos, Remains and reconstructions of roof construction (Baird, 2006: figs. 48, 49, 50, 51)
Fig. 304 House M7W, Panels of wall paintings with dining scenes (Baird, 2006: figs. 381& 382)
517
B. CURRICULUM VITAE
PERSONAL INFORMATION
Surname, Name: Ayatar, Esra
Nationality: Turkish (TC)
Date and Place of Birth:
Marital Status:
Phone:
email:
EDUCATION
Degree Institution Year of Graduation
MA METU-Architectural History 2005
BS METU-Architecture 2002
High School Atatürk Kız Lisesi, Konya 1997
PROFESSIONAL EXPERIENCE:
Year Firm Enrollment
2003-2007 Aysel Construction Co. Architect
Participated Porjects:
Canadian Chancery Building - Ankara/TURKEY
Jordan Armed Forces General Headquarters - Amman/JORDAN
Navoi Airport Air Control Buildings - Navoi/UZBEKISTAN
Hyatt Regency Hotel - Kiev/UKRAINE
Almaty Finance Center Almaty/KHAZAKISTAN
518
2007-2014 Yüksel Constructıon Co. Architect
Participated Porjects:
Al Fateh University Science, Buildings of Science, Veterinary and Agriculture Faculties Tripoli - LIBYA
U-CELL Office Building- Tashkent - UZBEKISTAN
Naval Academy Complex- Turkmenbashi - TURKMENISTAN
2015-2016 METU-GSSS Research Asistant
CONGRESS AND COLLOQUIUM
2021 “The Glamor of ‘Public’ in the Domestic Context: Continuity and Change in Reception Rituals and Spaces in the Roman East” Architecture and Endurance. European Architectural History Network. 30 September-02 October 2021.
2019 “İş Bankası Yeni Şehir Şubesi” Ankara’da İz Bırakan Mimarlar (Architects who left their marks in Ankara project”/VEKAM. 17.May 2019.
FOREIGN LANGUAGES
English: Advanced
French: Intermediate
FIELDS OF INTEREST:
Domestic Architecture, Roman Architecture, Late Antiquity, Roman East, Provincial Syria
519
C. TURKISH SUMMARY/TÜRKÇE ÖZET
DOĞU ROMA İMPARATORLUĞUNDA YAŞAM VE BARINMA:
SURİYE EYALETİNDE KONUT MİMARİSİ
1. Giriş
Bu tez Roma İmparatorluğu’nun doğu bölgelerindeki günlük yaşam pratikleri ve konut mimarisini incelemeyi amaç edinir. Çalışmanın temel odağı kent konutları olup, antik Suriye eyaletine bağlı beş farklı kente yoğunlaşır. Bu kentler konut mimarisi için nitelik ve nicelik bakımından daha doyurucu arkeolojik ve/veya epigrafik veriler sunan Asi’deki Antakya, Asi’deki Apamea, Zeugma, Palmyra ve Dura-Europos’dur. Bu kentlerden bugüne kadar elde edilmiş olan arkeolojik ve epigrafik veriler ışığında Roma döenmi evleri sosyal, kültürel ve mimari bağlamları ile birikte belgelenir ve tartışılır. Tezin çalışma alanı Helenistik dönemden antik çağın bitimine kadar olan zaman dilimini kapsar. Bu şekilde, Roma hâkimiyeti boyunca, bölgenin politik, ekonomik ve sosyal dinamiklerine de bağlı olan konut mimarisindeki, devamlılık, değişim ve dönüşümler izlenir. Bu coğrafi ve kronolojik çerçeve doğrultusunda ve karşılaştırmalı bir bakış açısıyla doğu Roma kentlerindeki konutlarda tespit edilen ortak özellikler ve farklılıklar, bölgeye özgü olgular ile kültürel ve mimari alanlardaki özümlemeler tartışılır. Bu tartışmalar doğrultusunda, bu tez, bölgesel verileri daha geniş bir bağlama oturtmayı ve Roma dönemi konut mimarisi ile ilgili güncel tartışmalara yeni perspektifler ve daha geniş çerçeveler sunmayı hedefler.
Gündelik yaşam tartışmalarının önde gelen isimlerinden olan, Fransız sosyolog ve filozof Henri Lefebvre toplumu analiz etmeye bireylerden başlar ve onları kendi yaşamlarını, bilinçlerini ve mekânlarını üreten sosyal varlıklar olarak tanımlar (1991: 68-169). Lefebvre’ye göre mekân sadece bir nesne veya hacim değil, bir sosyal gerçeklik ve bir çeşit ilişkiler ve formlar bütünüdür. Mekân aynı anda hem sosyal/toplusal bir ürün hem de bir üretim aracıdır. Bu nedenle Roma evini anlamak
520
için onu sadece günlük yaşam pratiklerinin ve geleneklerin mimari bir yansıması olarak değil, aynı zamanda sosyal/toplumsal ilişkilerin sürekli yeniden üretildiği sosyal bir mekân olarak da düşünmek gerekir. Knights’a göre Roma evinin barındırdığı sembol ve detaylar Romalıların evrenle ilişkisinin bir ifadesidir (1994: 102-131). Bu ortak evrensel algı ve onun konut mekânlarındaki ifadeleri, Roma toplumuna ait olmanın da bir göstergesidir (Grahame, 1998: 163; Hales, 2003:1). Bu bağlamda Roma konutu ile ilgili çalışmalar, sadece günlük yaşam pratikleri, konut mimarlığı ve sanatı ile ilgili önemli bilgiler sunmakla kalmaz aynı zamanda Roma toplumunun bir ‘mikrokozmos’u olarak Roma dünyasının sosyal ve kültürel yapısına da ışık tutar.
Son yıllarda Roma konutunu sosyal bağlamları ile birlikte tartışan güncel çalışmalar konuyla ilgili farklı çerçeveler ve bakış açıları sunmuş olsa da, bu çalışmaların büyük bir bölümü hala İtalya’nın Campania bölgesinden ve bazı batı eyaletlerinden gelen verilere odaklanır. Doğu eyaletlerinden gelen ve miktarı gün geçtikçe artmakta olan arkeolojik veriler ise Roma konut mimarisi yazımında henüz yeterli yeri bulamamıştır. Bu tez, bu açığı kapatmayı hedefler ve imparatorluğun doğu kesimlerine, Suriye eyaletinde kazılan konutlara, odaklanır. Yerel ve bölgesel verileri imparatorluğa ait daha geniş bir panoramaya oturtarak, Roma konut mimarisi çalışmalarına yeni bağlamsal perspektifler ve daha kapsayıcı bir çerçeve sunmayı hedefler.
Roma İmparatorluğunun doğu eyaletlerinin bulunduğu ve bugün “Yakın Doğu” olarak tanımlanan bölgenin sahip olduğu zengin arkeolojik miras 17. yüzyıldan itibaren batılı gezginlerin ve araştırmacıların ilgi odağı olmuştur. Bölgede 20. yüzyılda başlayan ilk sitemli arkeolojik kazılar anıtsal kamu yapılarına odaklanmıştır. Antik konut yapıları ise ancak son yıllarda, arkeolojide ve sosyal bilimlerde günlük yaşama olan ilginin artması ile birlikte ilgi odağı haline gelmiştir. Bugün Türkiye ve Suriye sınırları içinde kalan ve Roma İmparatorluğunun Suriye eyaletini kapsayan topraklarda yapılan arkeolojik çalışmalarda önemli sayıda konut yapısı kazılmıştır. Bu sayı, batı eyaletlerinden gelen arkeolojik verilerle kıyaslandığında hala oldukça kısıtlı olsa da, imparatorluğun doğu bölgelerindeki günlük yaşam ve konut mimarisi çalışmaları için
521
önemli bir veri seti oluşturur. Bu bağlamda bu tez, Roma konut mimarisini Suriye eyaletinde bulunan ve nitelik ve nicelik bakımından daha doyurucu arkeolojik veriler sunan Asi’deki Antakya, Asi’deki Apamea, Zeugma, Palmyra ve Dura-Europos kentleri üzerinden tartışır.
Tezin ana kaynağını yayınlanmış kazı raporları, araştırmalar, planlar, epigrafik kaynaklar ve bazı kentler için arkeolojik sit alanlarında veya müzelerde sergilenen kalıntılar ve buluntular oluşturur. Bu kaynaklar mevcut akademik çalışmalar için önemli veriler sunsa da ilgili kazıların kapsam, yöntem ve raporlanmaları ile gün yüzüne çıkartılmış kalıntıların korunma durumuna bağlı olarak bazı eksikler ve olumsuzluklar içerir. Bu durum söz konusu verilerin güncel Roma konutu tartışmalarında yeterli yer bulamamasının da başlıca nedenlerinden biridir. Bölgede yapılan ilk kazılarda konut yapıları ya tamamen göz ardı edilmiş ya da müze koleksiyonları için değerli eserler bulmak amacıyla kazılmıştır. Kazı raporlarında bu yapılardan çoğunlukla yüzeysel olarak bahsedilir. Raporlara eklenen yapı planları ise genellikle eskiz niteliğinde olup eksikler kimi zaman varsayımlarla tamamlanmıştır. Zeugma ve Asi’deki Apamea’da gerçekleştirilen daha geç dönem kazıları dışında yapılan arkeolojik çalışmaların genelinde buluntular bağlamları dışında incelenmiş ve raporlanmıştır. Konu ile ilgili araştırma ve çalışma yapmayı zorlaştıran ve bilim insanları arasında iletişim ve işbirliğini zorlaştıran bir diğer etken de günümüzdeki politik ve kültürel sınırlar ile yapılan yayınların dilidir. Suriye’deki arkeolojik yayınların dili yakın zamana kadar çoğunlukla Fransızca olup, Türkiye’dekiler genellikle İngilizce ve Fransızcadır. Bununla beraber bölgede arkeolojik çalışma yapan yerel ekiplerin çoğu raporlarını sadece Türkçe, Arapça, İbranice gibi dillerde yayınlamış, bu da çalışmaların daha geniş akademik çevrelere ulaşmasını engellemiştir. Son olarak bu tezde çalışılan beş antik kent ile bu kentlerde kazılan konutlara ait kalıntıların mevcut durumundan bahsetmek gerekir. 2011 yılında Suriye’de başlayan iç savaş nedeniyle, tezin yazım sürecinde, bu ülke sınırları içinde kalan Asi’deki Apamea, Palmyra ve Dura-Europos antik kentlerini ziyaret etmek ve kalıntıları yerinde incelemek mümkün olmamıştır. Ayrıca iç savaş sürecinde her üç şehirdeki kalıntılar hasar görmüştür. Zeugma’daki kazılarda açığa çıkartılan birçok
522
yapı ise Birecik barajının su tutmaya başlaması ile birlikte sular altında kalmıştır. Asi’deki Antakya’da kazılan sayısız konut, mozaikleri kaldırıldıktan sonra tekrar kapatılmış olup bugün yerleri dahi tam olarak bilinmemektedir.
Mevcut veriler ışığında bu tez bahsedilen beş kentte yapılan kazılarda gün yüzüne çıkmış ve bazıları kazı raporları dışındaki akademik çalışmalarda çok kısıtlı yer bulabilmiş olan konut yapılarını bir araya getirir, arkeolojik ve epigrafik veriler yardımıyla inceler ve onları daha geniş kronolojik ve coğrafi çerçevelere oturtmaya çalışır.
Tezin giriş kısmından sonraki ilk bölümde, her kentin coğrafi, sosyal, politik ve ekonomik bağlamları özetlenir ve kent nüfusunun etnik, sosyal, dini vb. kimlikleri sorgulanır. Bütün bu olguların kentin sosyal ve kültürel katmanları ile yapılı çevresinin oluşmasında, özellikle de konut mekânlarının sosyal ve mimari üretiminde etkili olduğu vurgulanır. Tezin takip eden beş bölümü beş ayrı kente odaklanır. Her bölüm kent planlarının oluşumu ve gelişimi ile konutların kent içindeki konumlarının irdelenmesi ile başlar. Daha sonra, o kentlerde yapılan arkeolojik çalışmaların tarihçesi ile mevcut verilerin kapsamı özetlenir. Bu verilere dayanarak o kentte kazılan ve belgelenen konutlar detaylı bir şekilde incelenir. Her bölümün son kısmında ise konutlardan elde edilen veriler doğrultusunda, o kentin günlük yaşam pratikleri ve mimarisi daha kritik bir bakış açısıyla değerlendirilir. Tezin son bölümünde, eldeki tüm veriler, yapılan çıkarım ve yorumlar karşılaştırmalı bir bakış açısı ile daha geniş bir çerçeveye oturtulur. Bu şekilde Helenistik dönemden antik dönemin sonuna kadar Roma İmparatorluğun doğudaki kentlerinde, günlük yaşam ve konut bağlamlarında görülen devamlılık, değişim ve dönüşümler ortaya konulur.
523
2. Doğu Roma: Coğrafi, Politik ve Sosyal Bağlamlar
2.1 Asi’deki Antakya: Suriye Eyaleti’nin Başkenti
Greko-Romen döneminin en önemli kentlerinden biri olan Asi’deki Antakya bugün Türkiye’nin güney kesiminde bulunan modern Antakya şehrinin bulunduğu noktada kurulmuştur. Antik kent batıda Asi nehri ve doğuda nehre paralel uzanan Silpius Dağları (günümüzdeki adıyla Habib Neccar Dağı) arasında konumlanmıştır. Kentin 9 km kadar güneyinde verimli arazileri ve doğal su kaynaklarına sahip Defne platosu (bugünkü adıyla Harbiye) yer alır. Antakya’nın yakın çevresi, özellikle de aşağı Asi havzası ve Amuk yaylası, bugün olduğu gibi, antik dönemde de kente verimli tarım toprakları sağlamıştır.
Asi’deki Antakya kenti M.Ö. 300 civarında 1. Seleukos Nikator tarafından, kolonileşme programının bir parçası olarak askeri amaçlarla kurulmuştur. M.Ö. 64 yılında Roma ordusu generallerinden Pompey tarafından Seleukos İmparatorluğuna son verilmesiyle Suriye toprakları Roma İmparatorluğunun yeni eyaleti olmuş, Antakya da bu eyaletin başkenti yapılmıştır. Antakya, politik, ekonomik ve sosyal alanlarda hem Seleukos hem de Roma İmparatorluğunun en önemli ve görkemli şehirlerinden biri olmuştur. Kentte Roma dünyasının en önemli festivallerinden olan Olimpiyat Oyunları düzenlenmiş, Konstantin döneminden itibaren Hıristiyanlık için önemli bir merkez haline gelmiştir. 4. yüzyıl ve sonrasında kent çeşitli depremler, veba salgınları, büyük yangınlar, ayaklanmalar ve istilalara maruz kalmış ve giderek zayıflamıştır. M.S. 638 yılında kent Araplar tarafından ele geçirilmiş ve yaklaşık bin yıl süren Greko-Romen hâkimiyeti sona ermiştir.
Bir Yunan kolonisi olarak kurulan Helenistik Antakya kentinin nüfusu 17.000-25.000 arasında tahmin edilmekte ve bu nüfusun çoğunluğunu Yunan ve Makedon kökenliler oluşturmaktadır (Downey, 1961: 80). Roma dönemindeki nüfusa ait tahminler ise 250.000 ve 800.000 arasında değişmektedir (Downey, 1958: Kloeg, 2013).
524
Antakya’daki kentli nüfusun kullandığı başlıca dil Yunanca olup, Latince, İbranice, Arami, Kıpti vb diller de halk arasındaki konuşma dilinde kullanılmıştır.
2.2 Asi’deki Apamea: Syria Secunda Eyaletini Başkenti
Asi’deki Apamea kentinin kalıntıları modern Suriye’de Hama kentinin 55 km kadar kuzey batısında Asi nehri kıyısında yer alır. Antik kent nehrin kavis yaparak sardığı yarımada biçimli verimli bir platoda yayılır.
Kazılardan elde edilen bulgulara göre Apamea kentinin bulunduğu bölge tarih öncesi çağlardan itibaren yerleşim görmüş ve birçok medeniyete ev sahipliği yapmıştır. Büyük İskender’in bölgeyi fethinden sonra küçük bir Pers kenti olan Pharnake’ye bir Makedon askeri birliği yerleştirilmiş ve kentin adı Pella olarak değiştirilmiştir. M.Ö. 300 civarlarında 1. Seleukos Nikator burayı bir koloni yerleşimi yapmış ve kente Apamea adını vermiştir. Helenistik dönem boyunca Apamea önemli bir askeri merkez olmuştur. M.Ö. 64’de kent Pompey tarafından Roma topraklarına katılmış ve zamanla Roma İmparatorluğunun en önemli kentlerinden biri haline gelmiştir. Erken 5. yüzyılda Apamea Syria Secunda eyaletinin başkenti olmuş ve geç antik dönemde önemini ve ihtişamını büyük ölçüde korumuştur. Kent M.S. 636 yılında Yarmuk Savaşı sonunda Arap hâkimiyetine geçmiştir.
6. yüzyılda Q. Aemilius Secundus tarafından yapılan sayımına göre Apamea’nın nüfus 117.000’dir. Bu sayının tam olarak nüfusun hangi kesimini kapsadığı net olarak bilinmemekle birlikte günümüzde antik kentin nüfusu için yapılan tahminler 125.000-130.000 (Butcher, 2003:106; Kennedy, 2006) ve 400.000-5000.000 (Cumont, 1934: 189; Balty, 1988: 96; Millar, 1993:250) arasında değişmektedir. Kentte bulunan mezar yazıtlarına bakarak nüfusun büyük bir çoğunluğunu Yunan ve Sami kökenlilerin oluşturduğu söylenebilir. Latin isimlere nadiren rastlanır. 2. yüzyılda kentteki nüfus artışına paralel olarak toprak sahibi yeni bir sosyal sınıfın ortaya çıktığı, kent çevresinde kazılan geniş mülklerden ve görkemli anıt mezarlardan anlaşılmaktadır.
525
2.3 Zeugma: Fırat’ta bir Köprü
Bugün Türkiye’nin güneydoğusundaki Gaziantep ili sınırlarında yer alan Zeugma antik kenti o dönemde Fırat nehrinin en kolay geçilebileceği noktalardan birine kurulmuş ve iki yaka arasında bir köprü görevi üstlenmiştir. Kent nehrin kuzey kıyısında Hobap platosunun verimli topraklarına, güney kıyısında ise dik ve engebeli bir araziye yayılmıştır.
Zeugma M.Ö. 300 yılı civarında 1. Seleukos Nikator tarafından askeri bir koloni olarak kurulmuştur. M.Ö. 64’de Romalı General Pompey müttefikliğinin ödülü olarak kenti Komagene Kralı Antiochus’a hediye etmiştir. Kısa süre içinde kent gelişmiş ve Komagene Krallığının dört büyük kentinden biri olmuştur. M.S. 72 yılında Komagene toprakları Roma İmparatorluğu’nun Suriye eyaletine bağlanmıştır. Zeugma ticaret yolları üzerindeki kilit konumu ve bulundurduğu askeri birlikler sayesinde Roma İmparatorluğunun en önemli sınır kentlerinden biri olmuştur. M.S. 250 civarında Sasani Kralı Şapur kenti ele geçirmiş ve yakıp yıkmıştır. Sonrasında kent kısmen toparlansa da tekrar eski görkemine kavuşamamıştır. M.S. 7. yüzyılda Arap hâkimiyetine geçtiğinde kent küçük bir yerleşim yerinden ibarettir.
Helenistik dönem için kentin nüfusu 25.000 civarı, Roma dönemi için ise 50.000-75.000 arası olarak tahmin edilmektedir (Grainger, 1990: 91-92, Kennedy, 1998a: 33). Yazıtlardan ve mezar taşlarından edinilen bilgiye göre kentte kullanılan başlıca dil Yunancadır. Sivil topluma ait yazıtlar Yunanca olup, yazıtlarda geçen isimler çoğunlukla Sami isimleridir. Latince yazıtların neredeyse tümü orduya aittir. Mezar taşlarındaki figürler çoğunlukla Part veya Yunan kıyafetleri ile resmedilmişlerdir
2.4 Palmyra: Çölde bir Ticari Vaha
Palmyra antik kenti modern Şam şehrinin 230 km kuzey doğusunda, Suriye çölünün ortasında, etrafı palmiye ağaçları ile çevrili bir vahada kurulmuştur. M.Ö. 2. binyılın başlarından itibaren bölgede bulunan yazılı kaynaklarda kent Tadmor adı ile anılır.
526
Tadmor, Seleukos hâkimiyetine geçmeden önce Asurlular ve Persler tarafından yönetilmiştir.
Palmyra M.Ö. 1. yüzyılın ortasına kadar küçük bir yerleşim yeridir. Helenistik dönemde dört büyük aşiretin temsilcileri tarafından yönetilen kent, Vespasian döneminde bir meclis ve konsey kurulması ile diğer Helenistik dönem şehirlerinin yönetim şemasını benimsemiştir. M.S. 213-216 yılları civarında Palmyra kenti bir Roma kolonisi olmuş, Romalı yöneticiler tarafından yönetilmeye başlamıştır.
İki büyük imparatorluğun sınırında ve Fırat nehri ile Akdenizin tam ortasında konumlanan Palmyra doğu ile batı arasındaki ticari ağın kilit noktalarından biri olmuş ve kervanların uğrak noktası haline gelmiştir. Ayrıca Palmyralı okçu birlikler çölde seyahat eden kervanların güvenliğini sağlamış ve bunlardan aldıkları vergilerden önemli bir gelir elde etmişlerdir. M.S. 2. yüzyılda Roma topraklarının Fırat nehrinden doğuya doğru genişlemesiyle Palmyra ve çevresi askeri açıdan da önem kazanmıştır. M.S. 3. yüzyılda Roma imparatorluğunda baş gösteren iç savaşlar ve Sasani İmparatorluğunun güçlenmesi bölgedeki Roma gücünü zayıflatmış ve ticaret ile ekonomiyi olumsuz yönde etkilemiştir. Palmyra 7. yüzyılda Arap hâkimiyetine geçene kadar küçük bir askeri yerleşim olarak kalmıştır.
Palmyra ve çevresinde yaşadığı bilinen ilk halk Amoritelerdir. M.Ö. 2. ve 1. binyıllarda bölgeye Araplar ve Aramlar gelmiş ve kentin nüfusunun büyük bir çoğunluğunu oluşturmuşlardır. Helenistik dönemde kente çok sayıda Yunan kökenli ve Roma hâkimiyetiyle birlikte sayıları az da olsa Latin kökenli nüfus yerleşmiştir. Palmyra kenti yerleşik halk ile göçebe aşiretlerin sürekli etkileşimde olduğu bir yerleşimdir. Özellikle M.S. 1. ve 2. yüzyıllara tarihlenen yazılı kaynaklarda biyolojik ve sosyal ilişkileri belirtmek için akrabalık terimleri sıkça kullanılmıştır (Dirven, 1999: 22-23; Yon, 2002: 57-97).
Palmyra halkı Arami dilinin yerel bir lehçesini konuşmuştur. M.S. 1. yüzyıldan itibaren yazılı kaynaklarda bu dille birlikte Yunanca da görülmeye başlar. Kamu veya
527
sivil bağlamlardaki yazıtlarda genellikle iki dil birden kullanılırken dini yazıtlar ve mezar taşlarında yalnızca Aramice kullanılmıştır. Az sayıdaki yazıtta ise Aramca, Yunanca ve Latince bir arada kullanılmıştır. İki dil kullanımı isimlerde de yaygındır. Özellikle M.S. 2. yüzyıldan itibaren kişiler Semitik isimlerine ek olarak Greko-Romen isimleri almaya başlamışlardır. Palmyra toplumunun sahip olduğu kültürel çeşitlilik ve sentez dini yapılar ve mezar anıtlarında da görülür.
2.5 Dura-Europos: Fırat’ta bir Kale
Dura-Europos kentine ait kalıntılar modern Suriye sınırları içinde al-Salihiyeh köyü yakınlarında yer alır. Antik kent nehirden 40 m. kadar yüksekte kayalık bir platonun uç kısmına kurulmuştur. Nehre ve ovaya tepeden bakan ve iki taraftan vadilerle sınırlanan coğrafi oluşum kente doğal bir savunma sağlar.
Dura-Europos 1. Seleukos Nikator’un yeğeni Nikanor tarafından kurulmuş ve M.Ö. 2. yüzyıl ortalarına kadar küçük bir askeri yerleşim yeri olarak kalmış ve şehir statüsüne yükseltildikten sonra hızlı bir kentsel gelişim ve nüfus artışına sahne olmuştur. Dura-Europos M.Ö. 2. yüzyılın sonlarından M.S. 165 yılına kadar Part hâkimiyetinde kalmıştır. Yazıtlardan ve antik kaynaklardan anlaşıldığı üzere Partlar kenti mevcut Yunan yönetim şekli ve kurumlarında fazla bir değişiklik yapmadan yönetmişlerdir. Yunanca hem kamu hem de sivil bağlamlarda kullanılan başlıca dil olmaya devam etmiştir. M.S. 165 yılında Dura-Europos Roma topraklarına katılmış ve Suriye eyaletine bağlanmıştır. Yazılı kaynaklar ve arkeolojik veriler Roma hâkimiyetinde de kentin düzen ve yönetiminde fazla bir değişiklik yapılmadığını gösterir. Ancak M.S. geç 2. ve 3. yüzyıllarda Septimus Severus ve haleflerinin yaptığı yeni düzenlemeler ile kentin askeri önemi artmış ve daha büyük ve sürekli bir ordu kente yerleşmiştir. Dura-Europos tahminen M.S. 256/257 yıllarında Sasaniler tarafından işgal edilmiş ve sonrasında tamamen terkedilmiştir.
Dura-Europos’un en yoğun olduğu dönemde nüfusu 5.000-6.000 olarak tahmin edilmektedir (Hoepfner and Schwandner, 1986: 258; Will, 1988: 315-321). Nüfusun
528
çoğunluğunu Greko-Makedonlar ve bölgenin yerlisi olan Semitik Mezopotamyalılar oluşturmaktadır. Kentte hatırı sayılır sayıda Palmyralı nüfusun olduğu da bilinmektedir. Arkeolojik ve epigrafik veriler kent halkının etnik, kültürel ve dini kimliklerinin çeşitliliğini gözler önüne serer. Kentte Latince, Yunanca ve birçok farklı Semitik dil kullanılmıştır. Arkeolojik kazılarda hem Yunan hem de bölgenin yerel tanrılarına adanmış birçok farklı tapınak ortaya çıkartılmıştır.
3. Asi’deki Antakya
3.1 Bir Eyalet Başkentinin Kent Planı ve Kentsel Gelişimi
Antik dönemin en önemli ve gösterişli metropollerinden biri olan Asi’deki Antakya kenti ile ilgili mevcut arkeolojik veriler oldukça kısıtlıdır. Kentte bugüne kadar yapılan arkeolojik kazılar süre ve kapsam bakımından oldukça sınırlıdır. Buna karşın özellikle geç antik döneme tarihlenen birçok yazılı kaynak kent ile ilgili önemli bilgiler içerir.
Seleukos döneminde Antakya, Asi Nehri ile bölgede var olan bir ticaret yolu arasındaki düz bir araziye kurulmuş ve surlarla çevrelenmiştir. Izgara planına sahip kentin alanı yaklaşık 100-110 hektardır (fig. 6) (Cabouret, 1999: 134). 2. Seleukos zamanında kentin başkent olması ile birlikte nehir üzerindeki adada bir inşa programı başlatılmış ve burası etrafı surlarla çevrili, ızgara planlı ve kente köprülerle bağlanan ayrı bir kentsel merkeze dönüşmüştür. Yaklaşık yüzyıl sonra 4. Antiochus Epiphanes kenti doğuda Silpius dağının eteklerine doğru genişletmiş ve bu bölge Epiphania olarak adlandırılmıştır. Kentin doğuya doğru genişlemesiyle daha önce doğu sınırını oluşturan cadde yerleşimin ortasında kalmış ve genişleyerek anıtsal bir karakter kazanmıştır. Antakya’nın Roma hâkimiyetine geçmesinden sonra, kent imparatorların ve kentin önde gelenlerinin katkılarıyla gelişmiş ve ihtişamlı bir başkente dönüşmüştür.
Antakya kent planının en dikkat çekici unsuru sütunlu caddesidir. MS 2. yüzyılda depremlerle hasar gören cadde onarılmış, genişletilmiş ve mermer sütunlarla bezenmiştir. Yaklaşık 3 km uzunluğunda olan sütunlu caddenin genişliği her iki
529
tarafındaki geniş portikolarla birlikte 41 m.ye ulaşmıştır. Libanius söylevlerinde bu caddenin gece ve gündüz devam eden kalabalığından, canlılığından, ihtişamından ve sosyal hayata katkısından söz eder (Or. XI. 215-218, 266-267). Roma dönemi Antakyasında tiyatro, hamam, su kemeri gibi birçok kamu yapısı inşa edilmiştir. Bununla birlikte, nehrin üzerinde yer alan adadaki binalar yenilenmiş, yeni bir saray, hipodrom ve hamamlar yapılmış ve bu bölge “Yeni Şehir” adını almıştır. Justinian döneminde kentin kapladığı alan yaklaşık 500 hektardır (Cohen, 2006: 84).
Roma hâkimiyetinde Antakya kent merkezindeki büyümeye paralel olarak çevre yerleşimler de gelişmiş ve genişlemiştir. Kentin 8 km güneyinde zengin su kaynakları ve yeşil doğasıyla kentliler için serin ve çekici bir sayfiye yeri olan Defne’de de büyük kamu yapıları ve gösterişli villalar inşa edilmiştir.
Antakya kentinin refahı ve ihtişamı M.S. 6. yüzyıla kadar devam etmiştir. 526 ve 528 yıllarında yaşanan iki büyük deprem kentte büyük hasara yol açmış imparator Justinian’ın katkılarıyla kısmen ayağa kaldırılmışsa da eski ihtişamına bir daha ulaşamamıştır.
3.2 Asi’deki Antakya’da Konut Mimarisi
Antakya antik kenti ve çevresi 1932-1939 yılları arasında Princeton Üniversitesi tarafından kazılmış ve bu kazılarda çoğunluğu özel konutlardan oluşan seksen civarında yapı tespit edilmiştir. Konut yapılarının büyük çoğunluğu sadece kısmen kazılmıştır ve/veya ortaya çıkarılan kısımları iyi korunamamış durumdadır. Bölgede antik dönemde meydana gelen depremler, toprak kaymaları ve bunlara bağlı inşaat faaliyetleri nedeniyle birçok yapı hasar görmüş, yıkılmış ve yeniden yapılmıştır. Buna ek olarak hasarlı binaların yapı malzemeleri yenilerin yapımında kullanılmak üzere taşınmıştır. Diğer taraftan, zamanın kısıtlı olması ve özellikle Defne’de kalıntıların üzerindeki fıstık ağaçlarının bulunması gibi nedenlerle konutlar ancak kısmen (sadece mozaikli odalar) kazılabilmiş ve mozaikler kaldırıldıktan sonra kalıntıların üzerleri tekrar örtülmüştür. Ayrıca yapıların duvarları korunmadığından dolaşım şemaları da
530
çoğunlukla belirsizdir. Bu nedenlerle bugüne kadar kentte kazılan konutlara ait yayınlanan planlar natamam olup, çoğunlukla varsayımlarla şekillenmiştir.
Antakya ve çevresinde kazılan konut yapılarının çoğu kentin yüksek kesimlerinde, Silpius ve Staurin dağlarının yamaçlarında, liman kenti olan Seleucia Pieria’da ve Defne’dedir. M.S. 2. yüzyıl ile 6. yüzyıl arasına tarihlenen bu yapılardan nispeten daha iyi korunmuş ve/veya belgelenmiş olanlar tezin bu bölümünde yapım yılına göre kronolojik bir sıralamayla detaylı olarak incelenmiştir. Bu evler: Atrium Evi, Takvim Evi, Kilikya Evi, İçki Yarışması Evi, Dionysus ve Ariadne Evi, Açık Büfe Akşam Yemeği Evi/Çiçekli Döşeme Evi, Psycheların Botu Evi, Menander Evi, Konstantin Villası, Ge ve Mevsimler Evi ile Yakto Kompleksi’dir (Ardabur’un Evi)103. Bu detaylı çalışma sonucu kentteki konut dokusu ve mimarisi hakkında elde edilen bilgi ve çıkarımlar bölümün son kısmında özetlenmiştir.
Antakya ve çevresinde kazılan konutlar farklı büyüklüklere ve plan tiplerine sahiptirler. Evlerin birçoğu mütevazı boyutta olsa da tümü belirgin bir lüks ve ihtişam sergiler. Plan şemalarının bütününde bir simetri veya eksenellik olmamakla beraber bazı mekân organizasyonlarındaki simetri dikkat çeker.
Antakya evlerinin mekânsal odağı triclinium yani yemek odalarıdır. Her evde en az bir yemek odası olup bu odalar boyut, düzen ve dekorasyonlarıyla öne çıkar. Bazı evlerde yemek odaları ile bağlantılı ve mozaikli yan odalar vardır. Yemek odaları ve yan odalar sütunlu portikolar aracılığı ile iç avlulara açılır.
Antakya evlerinin iç avluları genellikle küçük boyutlu olup hem görsel hem de fonksiyonel olarak yemek odalarının bir uzantısı niteliğindedir. Avlular sert zeminli olup, İtalya’daki evler gibi bahçe niteliği taşımaz. Atrium Evi ve Yakto Kompleksi dışında ortaya çıkartılan hiçbir evde peristile rastlanmaz. Avluların bir veya iki
103 Antakya ve bu tezde çalışılan diğer şehirler ile ilgili kazı raporları ve yayınlar çoğunlukla İngilizce ve Fransızca olup evlerin isimlendirmeleri de bu dillerde yapılmıştır. İsimler yazar tarafından Türkçeye çevrilmiştir.
531
tarafında Yunan evlerinde bulunan pastas veya prostas benzeri portikolar vardır. Antakya ve çevresinde, özellikle Defne’de zengin su kaynakları bulunduğundan çeşmeler, havuzlar vb su öğeleri Antakya evlerinin önemli unsurlarıdır. Bu öğeler konutlarda sadece hijyenik amaçlarla kullanılmamış aynı zamanda mekanlara lüks ve komfor sağlayan dekorasyon nesneleri olarak da kullanılmışlardır.
Antakya evlerinin ihtişamını sağlayan en önemli unsurlardan biri mozaik zemin kaplamalarıdır. Mozaiklerde Yunan mitolojisi ve edebiyatından sahneler betimlenir. Mozaik zemin kaplamaları mekânların görsel kalitesini artırmakla kalmaz aynı zamanda mekânlar arasında görsel ve fiziksel yönlendirme sağlar.
Arkeolojik ve epigrafik kaynaklar Antakya evlerindeki zenginlik ve lüksün geç antik dönemde de devam ettiğine ve bu dönemde evlerin mimari ve dekorasyon şemalarında bazı değişim ve dönüşümler yaşandığına işaret eder. Bu değişimlerden en dikkat çekici olanı yemek odalarındaki T şeklindeki mozaik panel düzeninin ortadan kaybolması ve bu odaların boyutlarının büyümesidir. Bu dönemde özellikle misafir/kabul odaları için yeni formlar tanımlanmıştır. Bu formların en dikkat çekici ve yaygın olanı geniş bir alanın eyvan benzeri daha dar bir alana açıldığı iki bölümlü mekânlardır.
Geç antik dönemde mozaik panellerde farklı temalar kullanılmaya başlamış, mitolojik repertuar yerini geometrik formlara, bitki ve çiçek motiflerine ve av sahnelerine bırakmıştır. Önceki dönemlerde daha çok kamu yapılarında görülen mermer kullanımı geç antik dönemde, çoğunlukla opus sectile şeklinde, konut dekorasyonunda da yaygın hale gelmiştir.
4. Asi’deki Apamea: Geç Antik Dönemin İhtişamlı Konutları
4.1 Bir Geç Antik Dönem Başkentinin Kentsel Planı ve Gelişimi
Asi’deki Apamea kentinde ilk yerleşime ait izler, nehir kıyısında, nehre ve vadiye hâkim konumda olan bir tepe üzerinde yoğunlaşır. Seleukoslar döneminde bu tepe
532
kentin savunma merkezi (akropol) olarak düzenlenmiş ve kent tepenin yamaçlarına doğru yayılmıştır (fig. 65, 66, 67). M.Ö. 2. yüzyılda surlarla çevrili kentin kapladığı alan yaklaşık 250 hektar olup, Helenistik dönemde bu alanın kısmen boş kalmış olduğu düşünülmektedir (Balty, J. Ch., 1991: 17). Kentin ızgara planı kuzey-güney yönünde uzanan bir sütunlu cadde ve onu doğu-batı yönünde kesen iki ana cadde etrafında şekillenir.
Asi’deki Apamea kenti Roma hâkimiyeti altında hızla gelişmiştir. M.S. 115 yılındaki depremde büyük hasar gören kentte geniş çaplı bir yenileme ve inşaat süreci başlamış ve bu sürece katkıda bulunan yöneticiler ve kentin ileri gelenleri ödüllendirilerek heykelleri kentin belirli yerlerine yerleştirilmiştir.
Deprem sonrası yapılan en dikkat çekici yenileme faaliyetlerinden biri sütunlu caddede gerçekleştirilmiştir. Portikolarla birlikte genişliği 30 m. olan Cardo Maximus 37 m.ye, genişliği 16 m. olan Decumanus ise 22 m. ye genişletilmiştir. Helenistik dönemin Dorik sütunları Korint sütunlarla değiştirilmiş ve cadde doğu Roma kentlerindeki en görkemli sütunlu caddelerden biri haline gelmiştir. Bu dönemde kentte görülen diğer dikkat çekici inşaat faaliyetleri tiyatronun büyütülmesi ve sahneye hidrolik altyapı döşenmesi, su kemerleri ile kente su getirilmesi ve birçok anıtsal çeşme yapılmasıdır.
Apamea kenti M.S. erken 5. yüzyılda Syria Secunda eyaletinin başkenti olmuş ve kentsel refah ve ihtişamı geç antik dönemde de sürmüştür. MS 526 ve 528 yıllarında kenti tahrip eden depremler sonrası yenileme faaliyetleri yapılmış ve başta sütunlu cadde, kiliseler ve özel konutlar olmak üzere birçok yapı dönemin mimari ve dekorasyon tercihlerine uygun şekilde restore edilmiştir.
M.S. 6. yüzyıldan itibaren Apamea’da orta çağ kentine doğru bir dönüşüm başlamış, sütunlu cadde araç trafiğine kapatılmış, portikolar zamanla kapatılarak dükkânlar oluşturulmuş ve cadde ticari aktivitelerin yoğun olduğu bir arter haline gelmiştir.
533
4.2 Asi’deki Apamea’da Konut Mimarisi
Asi’deki Apamea kentindeki arkeolojik çalışmalar 1928 yılında başlamış ve antik kent 1930-1938 yılları arasında Belçikalı arkeolog Franz Cumont tarafından kazılmıştır. 1947 ve 1953 yıllarında iki kısa kazı çalışması yürütülmüştür. 1965 yılından itibaren Belçikalı arkeolog Jean Charles Balty önderliğinde kentte düzenli kazılar yeniden başlamış ve bu kazılarda birçok konut yapısı kazılmış ve belgelenmiştir. 2001-2011 yılları arasında kazılar Didier Viviers başkanlığında devam etmiştir.
Bugüne kadar yapılan kazılarda Apamea’da en az dokuz konut yapısı tespit edilmiş, bunlardan dört tanesi neredeyse tamamen açığa çıkartılmış diğerleri ise kısmen kazılmış veya hiç kazılmamıştır (fig. 67). Bu bu tezde incelenen konutlar: Triclinoslu Yapı, Konsollar Evi, Pilasterler Evi, Konsollu Başlıklar Evi, Geyik Evi, İki Loplu Sütunlar Evi ve Üç Loplu Sütunlar Evi’dir. Bu konut kalıntılarına ait plan ve belgelerin detaylı incelenmesi sonucu kentteki konut dokusu ve mimarisi hakkında elde edilen bilgi ve çıkarımlar bölümün son kısmında özetlenmiştir.
Apamea’da kazılan antik konut yapıları 2.000 m2-4.500 m2 arasında değişen büyüklükleri, gösterişli mimarileri ve dekorasyonları ile dikkat çeker. Konutlara giriş, bazı kamusal yapıların kapıları ile benzerlik gösteren büyük ve gösterişli kapılardan sağlanır. Konutların plan şemasının merkezinde büyük sütunlu avlular yer alır. Avlunun bir tarafındaki sütunlar genellikle diğerlerinden daha büyük çapta ve daha yüksektir. Bu kolon sırasının arkasında yapının en büyük ve gösterişli odaları konumlanır. Muhtemelen kabul/misafir odaları olarak kullanılan bu odalar büyüklükleri, üç açıklıklı geniş giriş kapıları, kimi zaman da duvarlarında kullanılan yüksek kaliteli malzeme ve işçilik ile öne çıkarlar. Bu odaların her iki tarafında ve kimi zaman simetrik şekilde konumlanmış daha küçük yan odalar vardır. Bazı kabul/misafir odaları sütunlu avluya ek olarak geniş arka avlulara da açılır. Üst kotlardan düşmüş halde bulunan birçok mozaik parçası, mimari eleman ve bazı evlerde bulunan merdiven kalıntıları, konutların en az iki katlı olduğuna işaret eder. Aile bireylerine ait yaşam mekânları ve yatak odaları bu üst katlarda olmalıdır.
534
Apamea’da kazılan evler uzun kullanıma bağlı olarak belirli zamanlarda tadilata ihtiyaç duymuş olsa da bu tadilatlar binanın ana plan şemasının korunduğu küçük ölçekli işlerdir. M.S. 6.yüzyılda ise muhtemelen depremlerin yarattığı hasarların da etkisiyle konutlarda geniş çaplı tadilatlar yapılmış ve yapılar dönemin değişen ihtiyaç, tercih ve zevklerine göre dönüşüme uğramıştır. Kabul/misafir odaları yeniden düzenlenmiş ve bu büyük odaların geride kalan bölümleri yerden birkaç basamak ile yükseltilerek geri kalan kısmından farklılaştırılmış, bazılarına apsis eklenmiştir. Bu yükseltilen bölüme geçiş, kimi zaman iki taraflı sütunlar ve/veya kemerlerle vurgulanmıştır. Bazı konutların cephelerine yeni giriş kapıları açılmış ve bu kapılar giriş koridoru ve/veya cadde ile sütunlu avlu ve kabul/misafir odaları arasında eksenel bir görsel ve fiziksel bağlantı sağlamıştır. Kazılan birçok evde bulunan sigma mermer masa tablaları, yemek odalarının apsisli formları ve T biçiminde düzenlenen mozaiklerin ortadan kalkması geç antik dönemde yemek yeme ritüellerinde ve yemek odalarının hem tefriş hem de mekânsal düzenindeki değişikliğe işaret eder. Geç antik dönemde Apamea’daki konut yapılarında ön plana çıkan diğer bir unsur da suyun yaygın kullanımıdır. Bu dönemde evlere tuvalet ve banyolar eklenmiş ve bazı avlular ile misafir/kabul salonlarına dekoratif çeşmeler yerleştirilmiştir. Geç antik dönemde konutların dekorasyon şemalarında da birtakım değişiklikler olmuştur. Opus sectile zemin ve duvar kaplamaları yaygın olarak kullanılmaya başlamış ve mozaik döşemelerde av sahneleri sıkça resmedilmiştir. Apamea’da kazılan konutlarda M.S. 6. yüzyıl sonrasında yeni bir mekân tipi ortaya çıkar. Bu mekânlar avludaki sütun sıralarından birinin kapatılmasıyla oluşturulan lineer ve apsisli odalardır ve fonksiyonları tam olarak bilinememektedir.
M.S. 7. yüzyılın başlarından itibaren Apamea’nın kent konutları görkemini kaybeder ve bölünerek aynı anda birden çok aileyi barındırabilecek daha küçük ve kırsal yaşamın gerekliliklerine uygun konutlara dönüşür.
535
5. Zeugma: Nehrin İki Yakası Arasında Bir Dönüşüm
5.1 Fırat Kıyısındaki İkiz Kentlerin Planı ve Gelişimi
Fırat nehrinin karşılıklı iki yakasında kurulan Seleucia ve Apamea kentleri birlikte planlanmış olsalar da farklı kent planlarına sahiptirler ve zaman içinde farklı gelişimler göstermişlerdir.
Apamea kentinin konumlandığı nehrin doğu kıyısındaki düz arazi daha düzenli bir yerleşime izin vermiş ve surlarla çevrelenmiş kent ızgara planına göre şekillenmiştir. Surların çevrelediği alan yaklaşık 45 hektardır. Apamea kenti kısa bir süre yerleşim görmüş ve M.Ö. 2. yüzyılın sonu veya 1. yüzyılın başlarında terkedilmiştir.
Nehrin batı kıyısı tepelerden ve dik yamaçlardan oluştuğundan burada konumlanan Seleucia/Zeugma kenti daha düzensiz bir yerleşim sergiler. Zeugma kenti de Apamea gibi Helenistik dönemde surlarla çevrelenmiş, Roma döneminde kentin genişlemesine paralel olarak yeni surlar inşa edilmiştir. Kentin en yüksek noktası olan Belkıs Tepe bir kült merkezi niteliğindedir. Helenistik kent merkezi nehir kıyısında Bahçedere’nin güneydoğusundadır. Nehrin iki yakasını birleştiren köprünün de bu çevrede olabileceği düşünülmektedir (Abadie-Reynal et al., 1996: 319; Kennedy, 1998a: 41). Zeugma M.Ö. 2. yüzyılda Apamea’nın terkedilmesiyle eş zamanlı olarak hızlı bir kentsel gelişme göstermiştir. M.S. 1. yüzyılda 4. Scythica Lejyonunun kente gelmesi ile birlikte Zeugma’nın nüfusu artmış ve yeni kent merkezleri oluşmuştur. Roma döneminde kent güneyde Belkıs Tepenin yükselen yamaçlarına, doğuda Helenistik nekropole ve güney batıda At Meydanı platosuna doğru genişlemiştir. Yapılan kazılarda Belkıs Tepe’nin kuzey yamaçlarında bir zafer takı veya hamama ait olduğu düşünülen tonozlu bir yapı, tiyatro ve stadyum binalarının kalıntılarına rastlanmıştır. Kentin doğu ve batı sınırlarında ise askeri kamplar yer alır. Konut yapıları kentin doğusunda Helenistik nekropolü de içine alan bölgede yoğunlaşır.
536
Arkeolojik veriler Zeugma’da M.S. 3. yüzyılda büyük bir yıkım ve yangına işaret eder. Tahminen Sasani akınları sırasında oluşan bu yıkım sonrasında kent gerilemiş, birçok yapı tamamen terkedilmiş ve zamanla toprak altında kalmıştır. M.S. 5. yüzyılda kentte bir miktar canlılık görülse de, Zeugma kenti geç antik dönem boyunca nispeten küçük ve düzensiz bir yerleşim yeri olarak kalmıştır.
5.2 Zeugma’da Konut Mimarisi
Yerel halk tarafından “Belkıs Harabeleri” olarak adlandırılan kalıntılar 1900’lü yıllarda Franz Cumont ve Jörg Wagner tarafından antik Zeugma kenti ile ilişkilendirilmiştir. Kentte 20. yüzyıl sonuna kadar birçok izinsiz kazı yapılmış ve bu kazılarda çıkarılan bazı mozaik ve buluntular yurtdışına kaçırılmıştır. Zeugma antik kentindeki yasal kazılar 1987 yılından itibaren Gaziantep ve Malatya Müzeleri tarafından yürütülmüştür. 1994 yılında bölgede bir baraj yapılacağı açıklanmış ve kazılara hız verilmiştir. 2000 yılında Birecik Barajı’nın su tutmaya başlaması ile birlikte Zeugma antik kenti ulusal ve uluslararası medyada yer bulmuş ve bir Amerikan kuruluşu olan Packard Humanities Institute geniş kapsamlı kurtarma kazıları organize etmiş ve bu kazılara maddi destek sağlamıştır. Kurtarma kazıları kapsamında Oxford Arkeoloji, Centro di Conservazione Archeologica ve sivil bir oluşum olan Zeugma Girişim Grubu arkeolojik çalışmalar yürütmüştür. Aynı yılın Ekim ayında baraj sularının maksimum seviyeye ulaşmasıyla kurtarma kazıları sona ermiştir ve Apamea kentinin tamamı ile Zeugma kentinin yaklaşık üçte birlik alanı sular altında kalmıştır. 2005 yılından bu yana antik kentteki kazılar Prof. Dr. Kutalmış Görkay başkanlığında Ankara Üniversitesi tarafından yürütülmektedir. Zeugma antik kentinde bugüne kadar yapılan kazılarda yaklaşık otuz adet antik konut yapısı kısmen veya tamamen açığa çıkarılmıştır. Bu konutlardan nispeten daha iyi korunmuş olan; Çeşme Evi, Miğferler Evi, Boğa Evi, Dionysus ve Ariadne Evi, Poseidon Evi, Fırat Evi, Mozaiksiz Ev, Synaristosai Evi (Zosimos Evi) ve Geç İmparatorluk Dönemi Peristilli Evi detaylı olarak incelenmiş ve kentin konut dokusu ve mimarisi hakkında elde edilen bilgi ve çıkarımlar bölümün son kısmında özetlenmiştir.
537
Zeugma antik kentinin kurulduğu bölgedeki eğimli arazi ve yumuşak ana kaya oluşumu, kent planında olduğu gibi mimari tasarımlarda da simetri ve keskin bir düzene mani olmuş yerine daha pratik çözümler ve adaptasyonları gerekli kılmıştır. Antik kentte konutlar yamaçlardaki ana kayanın kısmen oyulması kısmen de doldurulması ile oluşturulan yapay teraslara inşa edilmiştir. Terasların nehre bakan ön bölümleri dolaşımı sağlamak için boş bırakılmıştır. Yapıların büyüklüğü ve plan şeması kısmen bu terasların ölçüsü ve biçimi ile bağlantılıdır. Zeugma’da kazılan evler Asi’deki Apameadakilerle kıyaslandığında boyut olarak daha mütevazıdır. Kentte bugüne kadar kazılan en büyük ev olan Poseidon Evi yaklaşık 1.000 m2 lik bir alanı kaplar. Fakat bu evin her biri yaklaşık 500 m2 olan iki ayrı evin birleşmesiyle oluştuğu düşünülmektedir. Benzer şekilde nispeten daha kapsamlı kazılan Fırat Evi ve Synaristosai Evi’nin büyüklükleri de 400 m2 ve 500 m2 arasında değişir. Kazılarda tespit edilen merdiven kalıntıları evlerin en azından bazı bölümlerinde üst katların olduğunu gösterir.
Zeugma’da kazılan en erken dönem evleri geç Helenistik ve erken Roma dönemine tarihlenir. Bu evler sütunlu avlular etrafında şekillenmiştir. Bazı evlerde avlunun sadece bir tarafında bulunan sütun sıraları arkalarındaki loca niteliğindeki yarı açık mekânları vurgular. Ana kayaya oyulmuş odalar kentte erken dönemden itibaren yaygın olarak kullanılmıştır.
M.S. 2. yüzyılın sonları ile 3. yüzyılın ilk yarısında ekonomik canlılığa, kentsel gelişime ve kent nüfusunun artmasına paralel olarak konut yapılarında kapsamlı yenilemeler yapılmıştır. Bu dönemde avlular konutların mekânsal odağı olmaya devam etmiş fakat hem fonksiyon hem de görsel nitelikleri açısından dönüşüme uğramışlardır. Avlulara mozaik kaplı sığ havuzlar ve dekoratif çeşmeler eklenmiştir. Avlu çevresindeki bazı odalar, avluya açılan pencereleri ve kapıları aracılığı ile avludan gelen gün ışığı ve temiz hava ile birlikte suyun ferahlatıcı etkisi ve görselliğinden de faydalanmışlardır. Zeugma evlerinde plan şemasındaki yerleşimleri, boyutları ve dekorasyonları ile iki tip oda ön plana çıkar. Bunlardan biri T düzenli mozaik kaplamaları ve freskleri olan yemek odaları diğeri ise kayalara oyulmuş ve
538
zengin mozaik ve fresklerle dekore edilmiş odalardır. Bu odaların küçük toplantılar ve/veya özel görüşmeler için kullanılmış olması muhtemeldir.
Zeugma konutlarındaki ihtişamlı dönem uzun sürmemiş ve 3. yüzyılın ortalarına doğru, tahminen yaklaşan Sasani akınlarının beklentisiyle, konutlar yeniden düzenlenmiştir. Evlerin bazıları bölünerek birden çok konut ünitesi elde edilmiş, mevcut kapı ve pencerelerden bir kısmı kapatılmış, mozaik ve fresklerle dekore edilmiş odalardan bazıları depo olarak kullanılmaya başlamıştır. Evlerin büyük bir kısmı M.S. 252/253 yıllarındaki Sasani istilaları sırasında tahrip edilmiştir. Zeugma’da M.S. 5. yüzyıla tarihlenen ve nispeten daha iyi mimari standartlara sahip az sayıda konut yapısı tespit edilmiş olsa da geç antik dönem konutlarının büyük bir kısmı kırsal yaşamın gerekliliklerine uygun hale getirilmiş mütevazı yapılardır.
6. Palmyra: Çölde Yaşam
6.1 Bir Çöl Kentinin Planı ve Kentsel Gelişimi
Doğudaki birçok Greko-Romen şehrinin aksine Palmyra ızgara planlı bir kent değildir. Kentin farklı bölgelerinde farklı plan modelleri uygulanmıştır. Helenistik dönemde yerleşim Bel Tapınağı çevresinden al-Qubur vadisinin güney kesiminde batıya doğru yayılmıştır. Geç Helenistik dönemde nüfusun artması ile birlikte kent vadinin kuzeyine doğru genişlemiş ve yapılar kuzey-güney yönünde birbirine neredeyse paralel uzanan caddeler arsına yerleşmiştir. Roma hâkimiyeti boyunca sütunlu cadde, tiyatro ve tapınaklar gibi birçok büyük çaplı kamusal yapı inşa edilmiş ve kent genişlemiştir. M.S. 273 yılında kentte devlete karşı yapılan ayaklanmayı bastıran Aurelian ve birlikleri birçok binayı tahrip etmiş ve sonrasında bazı büyük projelerin yapımı yarım kalmıştır. M.S. 4. yüzyıldan sonra kent küçük bir askeri yerleşim yeri olarak varlığını sürdürmüştür.
539
6.2 Palmyra’da Konut Mimarisi
Palmyra antik kentindeki arkeolojik kazılar 1920’lerde başlamış ve 2. Dünya Savaşı’na kadar devam etmiştir. 1946 yılında Suriye Fransız Mandasından kurtulup bağımsızlığını kazanınca kazılar yeniden başlamış ve 2011 yılında ülkede iç savaşın çıkmasına kadar sürmüştür. Bu kazılarda elde edilen konutlara ait arkeolojik veriler diğer kentlerdekilere göre daha kısıtlı olsa da kentteki konut mimarisi ile ilgili önemli ipuçları verir.
Palmyra kentinin kamusal merkezi ve Bel Temenosu’nun dışında kalan alanların büyük bir kısmının konut bölgeleri olduğu düşünülmektedir. Palmyra’da bugüne kadar kazılmış konut yapılarından nispeten iyi korunmuş ve/veya belgelenmiş olanları tezin bu bölümünde detaylı olarak incelenmiş ve kentin konut dokusu ve mimarisi hakkında elde edilen bilgiler ve çıkarımlar özetlenmiştir. Tezde incelen konutlar 38,39 ve 45 Nolu Evler, F Evi, Peristilli Ev, Achilles ve Cassiopeia Evi (Evleri), Tiyatronun Güneydoğusundaki Ev ve Nabu Tapınağının Doğusundaki Evler’dir.
Palmyra kenti düzenli bir ızgara planına sahip olmadığından konut planları diğer kentlerdeki gibi sınırları belirli dikdörtgen biçimli arsalara uymak yerine daha organik bir plan şeması sergiler. Kentin batı kesimlerinde konutlar aynı yönde uzanan caddeler arsında lineer bir yayılım gösterirken, kent merkezindeki konutlar mevcut binaların arasındaki boşluklara göre şekillenir. Kentin sonradan gelişen ve daha seyrek bir yapı dokusuna sahip doğu kesimlerinde ise yapılar daha büyüktür.
Palmyra evlerinin temel plan şeması merkezi bir avlu etrafında konumlanan odalardan oluşur. Daha büyük evler bu temel birimin tekrarlanması ile şekillenir. Muhtemelen M.S. 2. yüzyılda evlerin avlularına sütunlar eklenmiştir. Sütunlar avlunun dört tarafını çevreleyebildiği gibi tek tarafında da olabilir. Avlu çevresinde konumlanan odaların bazıları fonksiyonlarına göre daha göz önünde olup kolay ulaşabilirken bazıları daha gerilere yerleşmiş ve koridor veya ön odalardan geçilerek ulaşılabilir durumdadır.
540
Palmyra konut mimarisi ve dekorasyonunda Greko-Romen unsurlara geç 2. yüzyıl veya 3. yüzyıla tarihlenen yapılarda rastlanır. Achille ve Cassiopeia Evi ile Tiyatronun Güneydoğusundaki Ev daha geniş ve gösterişli yapılar olup plan şemalarında kısmi de olsa eksenellik ve simetri görülür. M.S. 2. yüzyılın ikinci yarısına tarihlenen ve varlıklı bir aileye ait olması muhtemel olan F Evi ise ayrı girişleri olan iki kısımdan oluşması, nispeten küçük odaları ve avluları ile daha yerel bir mimari plan şeması sergiler.
Palmyra kenti M.S. 4. yüzyıldan itibaren bölgedeki diğer kentlere benzer şekilde, fakat onlardan daha erken bir dönemde, kırsallaşma yönünde bir dönüşüm geçirmiştir. Büyük kamu yapıları işgal edilmiş ve barınma amaçlı kullanılmaya başlamıştır. Aynı şekilde büyük ve gösterişli konut yapıları bölünerek daha küçük ve kırsal yaşantıya uygun konut birimlerine dönüşmüştür.
7. Dura-Europos: Dükkânlardan Saraylara
7.1 Bir Kale Kentin Planı ve Gelişimi
Seleukoslar tarafından kurulduğu dönemde küçük bir askeri yerleşim olan Dura-Europos Fırat Nehrine hâkim kayalık bir platonun uç noktasındaki bir tepenin üzerinde kurulmuştur. Tepenin üst bölümünde surlarla çevrili bir saray yapısı konumlanmıştır. Askerlerin konutları ise surların çevresinde yer almıştır. M.Ö. 2. yüzyıldan itibaren Dura-Europos dikkat çekici bir kentsel büyüme ve gelişim göstermiş ve bu dönemde geniş kapsamlı bir kentsel planlama ve inşaat programı başlatılmıştır.
Kentinin ızgara planı Palmyra Kapısından başlayan ve nehre doğru devam eden ana caddeyi referans alır. Kentin merkezindeki agora ana caddenin hemen kuzeyindeki toplam yedi parsele yayılır. Dura-Europos kentinde doğudaki diğer Greko-Romen kentlerinin önemli unsurlarından olan tiyatro, stadyum vb. anıtsal kamu yapılarının olmaması dikkat çekicidir.
541
M.Ö. 113’de kentin Partlar tarafından ele geçirilmesi sonucu büyük çaplı projeler yarım kalmış, kent surları terkedilmiştir. Partlar zamanında kentte tapınaklar dışında çok fazla yeni kamusal bina yapılmamış mevcut binalar dönüştürülerek kullanılmaya devam etmiştir.
Dura-Europos M.S. 165 civarında Roma hâkimiyetine geçmiş ve askeri önemini yeniden kazanmıştır. Bu dönemde surlar ve kent kapısı onarılmış, kentin kuzey kesimi duvarlarla ayrılarak bir ordu karargâhına dönüştürülmüştür. Roma hâkimiyeti döneminde agoranın ve ana caddenin bir kısmı sütunlarla bezenmiştir.
7.2 Dura-Europos’da Konut Mimarisi
Dura-Europos kentine ait kalıntılar 1920 yılında bölgede konaklayan bir İngiliz askeri birliğin dini bir yapının duvar resimlerini keşfetmesi ile arkeoloji dünyasının dikkatini çekmiştir. 1922-1923 yıllarında Franz Cumont antik kentte iki kazı gerçekleştirir. 1928 yılında Yale Üniversitesi ve Fransız Academie des Inscriptions et Belle-Lettres işbirliğinde yeniden başlayan kazılar on sezon boyunca devam eder ve bu kazılarda kentin yaklaşık dörtte biri açığa çıkartılır. Bu kazılarda konut yapıları da kazılmış fakat konutların mimarisinden çok buluntulara odaklanılmıştır. 1982 yılında Misison Franco-Syrienne d’Europos-Doura (MFSED) kurulmuş ve kazılar Pierre Leriche başkanlığında yeniden başlamıştır. 2011 yılına kadar devam eden bu kazılarda daha önce açığa çıkartılan kalıntılar ve bu kazılara ait Yale Üniversitesi arşivinde saklanan ve yayınlanmamış belgeler yeniden incelenmiştir.
Dura-Europos kentinde bugüne kadar açığa çıkartılan ve nispeten iyi korunmuş yüzden fazla konut kalıntısı kentin ve bölgenin konut mimarisini çalışmak için çok zengin bir veri seti sunar. Bu konutlar bir veya iki odalı küçük birimlerden saraylara kadar geniş bir örnekleme sağlar. Tezin bu bölümünde incelemek üzere konut örnekleri seçilirken, kentteki konumları, büyüklükleri ve fonksiyonları gibi konularda farklı bağlamlar sunmaları amaçlanmıştır. İncelenen konutlar, Kale Sarayı, Redoubt Sarayı/Strategeion, C7 parselindeki konutlar, agoranın G1 ve G3 parsellerinde bulunan
542
G1A, G1B, G3G, G3B ve G3C konutları, B2 parselindeki konutlar, Lysias Evi, Büyük Atrium Evi, E4 Evi (Part Evi), Hıristiyan Evi ve Roma Sarayı’dır. Bu yapıların detaylı incelenmesi sonucu elde edilen veriler bölümün son kısmında özetlenmiştir.
Dura-Europosta yapılan kazılarda elde edilen arkeolojik verilere ek olarak, bulunan bir papirüs, kentteki konut mimarisi ve kullanımı ile ilgili önemli bilgiler verir. Bu belge Polemocrates adlı bir kişinin ölümünden sonra sahip olduğu mülkün çocukları arasında ne şekilde pay edileceğini açıklar.
Dura-Europos’da kazılan evlerin büyüklük, yön ve plan şemaları ızgara planındaki parsellere göre şekillenir. Yaklaşık 2.450 m2 lik parselleri tek bir büyük ev kaplayabildiği gibi aynı parsel çok sayıda küçük ev tarafından da paylaşılabilir. Evlerin çoğunluğu ortalama 300-350 m2’lik bir alan kaplar. Evlerin duvarları kerpiç olup teras çatılarda ahşap kirişler kamışlarla örtülmüş ve sıvanmıştır.
Dura-Europos’ta konut mimarisinin başlıca öğesi avludur. Saray niteliğinde olan yapıların avluları peristilli olup diğer konutların avlularının bir veya iki tarafı sütunlarla bezenmiş veya hiç sütun kullanılmamıştır. Avlularda bulunan tandırlar, değirmen ve havan gibi mutfak gereçleri burada yemek hazırlama ve pişirme işlerinin de yapıldığına işaret eder. Avluların ortasında muhtemelen atık suların toplandığı giderler kazılmıştır. Tüm evlerin avlularında bulunan merdivenler teras çatılara ulaşımı sağlar. Evlerin odaları avlu etrafında düzenlenmiştir. Bu odalardan bazıları, mimari ve dekorasyonları ile öne çıkar. Bunlar kabul/misafir odası olduğu düşünülen odalar olup, avludan genellikle birkaç basamakla çıkılır ve zemininde dört duvar boyunca devam eden yaklaşık 10 cm. yüksekliğinde sedir şeklinde yüükseltiler vardır. Avlu ve odaların zeminleri çoğunlukla sıkıştırılmış toprak bırakılmış veya toprak zemin sıvanmıştır. Dura-Europos’ta yapımı Helenistik veya Part dönemine tarihlenen konutlardan daha geç bir döneme tarihlenen ve Roma Sarayı olarak adlandırılan yapı mimari ve dekorasyon özellikleri ile diğer konutlardan farklılaşır. Bu yapıdaki apsisli odalar, simetrik ve eksenel yerleşimler, boyalı tonoz tavanlar vb. öğeler mimarideki Roma etkisine örnektir.
543
8. Sonuç: Suriye Eyaletinde bir ‘Romalı’ Olarak Yaşamak ve Barınmak
8.1 Suriye Eyaletinde bir ‘Romalı’ Olarak Yaşamak
Bu tezde odaklanılan beş kentin tümü Semitik bir çevrede gelişen Seleukos koloni yerleşimleri olarak kurulmuştur. Kentlerin ekonomisi temelde ticaret ve/veya tarıma dayalı alıp tüm kentler Seleukos ve Roma hâkimiyeti boyunca imparatorluk sınırlarının korunması için askeri önem taşımışlardır. Bu kentler farklı ölçü ve biçimlerde Greko-Romen etkisinde kalmışlar bu da kentlerin yapılı çevrelerine yansımıştır.
Beş kentin tarihi ve politik gelişimleri incelendiğinde ortaya çıkan en önemli şey Helenistik, Roma ve hatta Pars hâkimiyetleri arasında kentlerin düzen ve yönetim biçimlerinde çok keskin geçişler olmadığıdır.
Kentlerde nüfusun çoğunluğunu Greko-Makedon kolonistler ve bölgenin yerlisi olan Semitik Mezopotamyalılar oluştururken, nispeten daha az sayıda Latin, Yahudi, Arap vb kökenlere sahip kişilerde vardır. M.S. 212 yılında tüm özgür nüfus etnik orijinine bakılmaksızın Roma vatandaşlık hakkı kazanmıştır. Bölgedeki ‘Roma’halkı sadece etnik kökenleri ile değil, din, dil, kültür vb kimlikleri ile de büyük bir çeşitlilik gösterir. Kentlerdeki nüfusun en dikkat çekici özelliği çok dilli olmalarıdır. Halk arasında Yunanca ve Latinceye ek olarak Aramca, İbranice, Kıpti gibi birçok semitik dil konuşulmuştur. Yazıtlarda kullanılan başlıca dil ise Yunancadır. Palmyra ve Dura-Europos gibi yerel etkilerin daha yoğun olduğu şehirlerde Yunanca ile birlikte Aramca ve Süryanice de yazıtlarda kullanılmıştır. Doğu Roma kentlerinde yaşayan nüfusun sosyo-kültürel çeşitliliği, kişilerin isimlerinde, dini ritüeller ve yapılarda, ölü gömme gelenekleri ve mezar yapılarında da fark edilir.
544
8.2 Suriye Eyaletinde bir ‘Romalı’ Olarak Barınmak
Bu tezde çalışılan konutların planlamasında etkili olan başlıca faktörler, kent planı ve topografyası ile günlük yaşam pratikleri ve ritüellerine uygun mekânsal ihtiyaçlardır. İklim koşulları, ulaşılabilir inşaat malzemeleri ile kültürel ve mimari bağlamlarda süregelen yerel gelenekler ile Greko-Romen etkisi de konut mimarisinde önemli rol oynamıştır. Tezin bu son bölümünde önceki başlıklarda elde edilen veriler doğrultusunda konutların öncelikle kentsel çevre ile ilişkileri, sonra da mimari tasarım prensipleri, mekânsal özellikleri ve bu bağlamlardaki kronolojik gelişmeler özetlenir.
8.2.1 Kent Dokusunun Parçası Olarak Konut Yapıları
Suriye eyaletindeki birçok Roma kenti ızgara planına sahiptir. Bu planda yolların birbirine dik kesişmesiyle oluşan dikdörtgen inşaat parsellerinin en boy oranı genellikle 2:1’dir. Bu parseller üzerinde tek bir büyük konut yapısı veya çok sayıda küçük konut yapısı vardır. Asi’deki Apamea’da kazılan parsellerin ölçüleri 110 m. x 55 m. olup bu parseller üzerinde bir, iki veya üç konut yapılmıştır. Bu kentte kazılan konutların taban alanları 2.000 m2 ile 4.500 m2 arasında değişir. Dura-Europos’da ise parsel boyutları 70 m. x 35 m. olup, konutların bazıları tüm parsele yayılmış çoğunluğu ise parselde ortalama 300-400 m2 alan kaplamıştır. Palmyra ve Zeugma kentleri ızgara planına göre şekillenmemiş ve daha organik gelişimler göstermiştir.
Bu kentlerde kazılan antik dönem konutlarının bir kısmı büyüklükleri bakımından neredeyse bazı kamusal binalar ile yarışsa da kent dokusu içinde aynı ihtişamı göstermez. Apamea’da ortaya çıkarılan yapılar büyük taş bloklarla örülmüş duvarları ve anıtsal kapıları ile kent dokusu içinde dikkat çeker. Benzer şekilde Palmyra kentinde kazılan Tiyatronun Güneydoğusundaki Ev’in kuzey cephesi, önündeki sütunlu caddenin eğrisel aksına göre şekillenmiş ve caddede evin kapısına denk gelen iki sütunun arası açılarak evin girişi vurgulanmıştır. Diğer taraftan Dura-Europos’daki büyük konutlar ve Palmyra’da kazılan F Evi, daha küçük konutların birleştirilmesi veya plan şemalarının tekrarlanması ile oluştuğundan kerpiç cepheleri diğer
545
parsellerdeki cephelerden çok farklılaşmaz. Bu konutların mekânsal zenginliği ve büyüklüğü cephelere yansımaz.
Konutların hemen hemen hepsinde giriş kapıları ve koridorlar iç mekânlar ile sokak arasında doğrudan bir görsel veya fiziksel bağlantıyı engelleyecek şekilde konumlanmış ve böylelikle belirli bir mahremiyet sağlamıştır. Geç antik dönemde bu durum değişmiş, Apamea’daki Triclinoslu Yapı gibi bazı büyük yapılarda yeni giriş kapıları açılmış ve bu şekilde sokak, avlu ve kabul/misafir odaları arasında eksenel bir bağlantı sağlamıştır. Yapıların cephe duvarları iyi korunamadığından pencerelere ait veriler kısıtlıdır. Dura-Europos’da kazılan ve duvarları nispeten iyi korunmuş evlerde tespit edilen çok az sayıda pencere göz seviyesinden yüksekte olup dar açıklıklara sahiptir. Bunlar çevreyle bir görsel ilişki kurmaktan çok mekânlara ışık sağlamak için yapılmış olmalıdır. Kentte daha geç döneme tarihlenen Roma Sarayında ise durum değişmiş ve binanın muhtemelen özel yaşam alanlarının bulunduğu kuzey-doğu kesimindeki odalar Fırat nehrine hâkim bir terasa açılır. Apamea kentindeki evlerde ortaya çıkartılan büyük kabul/misafir odaları ise uzun pencerelerle caddeye veya çoğunlukla arka avlulara açılır. Antakya kentinde kazılan konutların cepheleri korunamamış olsa da Libanius özellikle yüksek kesimlerdeki ve Defne’deki ev sahiplerinin doğanın ve manzaranın tadını çıkartabildiklerinden bahseder (Or. XI. 200).
8.2.2 Mimari Tasarım ve Dil: Konut Mekânlarında Form, Fonksiyon ve Anlam
Bu tezde çalışılan konutların tümü mekânların avlular etrafında konumlandığı içe dönük bir tasarım anlayışı sergiler. Avlular Helenistik dönemden geç antik dönemin sonuna kadar konut mimarisinin en temel öğesi olarak kalmış olsa da, avlu etrafındaki diğer mekânlar gibi, farklı bölge ve dönemlerde farklı bağlamsal ve mimari yaklaşımlar sergilemiştir.
Bugüne kadar kazılan ve Helenistik döneme tarihlenen az sayıda konut yapısı peristilli ya da tek veya iki tarafında sütun sıraları olan avlular etrafında şekillenir. Sütun sıraları
546
genellikle yapının büyüklük ve eksenel yerleşimleri ile öne çıkan odalarına açılır. Bu plan şeması erken Roma döneminde de uygulanmaya devam etmiştir. M.S. 2. yüzyılın ikinci yarısından itibaren konut mimarisinde Helenistik etkinin azaldığı ve doğu geleneklerini de kapsayan Roma etkisinin artmaya başladığı gözlemlenir. Bu dönemde konut mimarisinde triclinium gibi yeni mekân tipleri tanımlanmıştır. Önceki dönemlerde inşa edilen mevcut konutlar yenilenmiş ve güncel ihtiyaçlar doğrultusunda yeniden şekillenmiştir. Bu düzenlemelerde özellikle kabul/misafir odaları ve yemek odaları genişlemiş, yeniden dekore edilmiş ve görsel ve fiziksel olarak daha çok vurgulanmıştır. Avlular da benzer şekilde mozaik döşemeler, dekoratif çeşmeler vb. öğeler ile zenginleştirilmiştir.
Suriye eyaletinde Roma dönemine tarihlenen konutların tümü aynı mimari yaklaşımları sergilemez. Dura-Europos’da kazılan evlerin neredeyse tümü ve Palmyra’da kazılan F Evi’nde Greko-Romen unsurlara nadiren rastlanır. Bu evler, ayrı giriş kapılarına sahip farklı bölümleri, toprak zeminli avluları ve odaları, daha küçük ve avlu etrafında daha düzensiz bir yerleşim sergileyen mekânları ile mimari ve dekorasyon bağlamında daha yerel bir yaklaşım sergilerler.
M.S. 3. yüzyıldan itibaren konut mimarisindeki en dikkat çekici değişim triclinia yani yemek odalarının plan şeması ve dekorasyonundaki değişimdir. Bu dönemde yemek odalarının birçoğu yeniden düzenlenmiş ve T şeklindeki mozaik panel düzeni zamanla kaybolmuştur. Bu odalara eklenen apsisler ve odalarda bulunan sigma şeklinde masalar üç klineli klasik yemek düzenin ve ilgili mekânların değişimine işaret eder. Konut mimarisinde tricliniumun kaybolmasına paralel olarak iki parçalı yeni bir oda tipi ortaya çıkar. Bu odalarda mekânın bir kısmı yükseltilerek ve geçişi kimi zaman kolon ve kemerlerle vurgulanarak diğer kısımdan farklılaştırılmıştır. Bazı örneklerde bu kısım apsisli bir forma sahiptir. Genellikle evin en büyük ve görkemli mekânı olan bu odaların yeni düzeni konuk ağırlama ve bazı toplantı ritüellerinde de bir dönüşüme işaret eder. Geç antik dönemde konutlar yeniden dekore edilmiş ve yeni dekoratif öğelerle bezenmiştir. Mozaik zemin kaplamalarındaki mitolojik ikonografi yerini bitkisel formlar, hayvan figürleri ve av sahneleri gibi temalara bırakmıştır. Kamusal
547
yapılarda sıkça kullanılan mermer konut mekânlarının dekorasyonunda da kullanılmaya başlamıştır. Geç antik dönemin sonlarına doğru Apamea’daki evlerde yeni bir mekân tipi ortaya çıkar. Peristil avluların herhangi bir tarafındaki portikoların kapatılması ile oluşan ve diğer eyaletlerde kazılan evlerde de benzerlerine rastlanan bu lineer ve apsisli mekânlar kabul/misafir odası olarak kullanılmış olmalıdır.
Antik dönemin sonuna doğru Roma İmparatorluğunun doğu topraklarında gücünü kaybetmeye başlaması ve tekrarlanan Arap akınlarının etkisiyle bölgedeki sosyal ve ekonomik yapıda ve buna bağlı olarak kentsel çevre ve konut mimarisinde radikal dönüşümler yaşanmıştır. Büyük kamusal yapılar küçük meskenler ve dükkânlar tarafından işgal edilmiş, büyük ve görkemli konutlar bölünerek çok sayıda ailenin aynı anda barınabileceği küçük yaşama birimlerine dönüşmüş, büyük odalar bölünerek daha küçük mekânlar elde edilmiş, mozaik kaplı odalar ahır, depo, atölye vb. olarak kullanılmaya başlamıştır. Zeugma ve Palmyra’da 4. ve 5. yüzyıllarda, Apamea ve Antakya’da da ise erken 7. yüzyılda başlayan bu dönüşümler neticesinde Roma kentleri kırsal yaşamın hüküm sürdüğü yerleşim yerlerine dönüşmüş ve M.S. 7. yüzyılın ortalarına doğru Arap hâkimiyetine geçmişlerdir.
548

LIVING AND DWELLING UNDER THE ROMAN ROOF IN THE EASTERN EMPIRE:
DOMESTIC ARCHITECTURE IN PROVINCIAL SYRIA