Architecture and Material Culture from the Earthquake House at Kourion, Cyprus: A Late Roman Non-Elite House Destroyed in the 4th Century AD 9781407312729, 9781407342399

In the late 4th century AD, the site of Kourion, Cyprus was destroyed by an earthquake that struck with little or no war

235 20 38MB

English Pages [157] Year 2014

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Architecture and Material Culture from the Earthquake House at Kourion, Cyprus: A Late Roman Non-Elite House Destroyed in the 4th Century AD
 9781407312729, 9781407342399

Table of contents :
Front Cover
Title Page
Copyright
Acknowledgements
TABLE OF CONTENTS
Abbreviations Used in the Text
List of Figures
Chapter 1: Introduction
Chapter 2: Kourion: History and Excavation
Chapter 3: Methodology
Chapter 4: Architecture
Chapter 5: Room Stratigraphy/Formation Processes
Chapter 6: Room Assemblages
Chapter 7: Activities and Aspects of Household Organization
Chapter 8: Conclusions
Appendix A: Catalogue of Individual Room Assemblages
Appendix B: Kourion Amphora Typology and Capacity Estimates
Appendix C: Tables
Bibliography

Citation preview

BAR S2635 2014 COSTELLO ARCHITECTURE AND MATERIAL CULTURE FROM THE EARTHQUAKE HOUSE

B A R

Architecture and Material Culture from the Earthquake House at Kourion, Cyprus Benjamin Costello IV

BAR International Series 2635 2014

Architecture and Material Culture from the Earthquake House at Kourion, Cyprus A Late Roman Non-Elite House Destroyed in the 4th Century AD

Benjamin Costello IV

BAR International Series 2635 2014

ISBN 9781407312729 paperback ISBN 9781407342399 e-format DOI https://doi.org/10.30861/9781407312729 A catalogue record for this book is available from the British Library

BAR

PUBLISHING

Acknowledgements project and introducing me fully to the archaeological community on the island. Special thanks are also due to Vathoulla Moustoukki and Diana Constantinides for all their help while at the institute. Following my time in Cyprus I received financial support in the form of a generous grant from the A.G. Leventis Foundation and a teaching assistantship from the University at Buffalo, both of which were instrumental for completing the initial draft of this work.

It is my pleasure to acknowledge the help and support of those who contributed to the successful completion of this work. A great deal is owed to the many professors I worked with during my time at the University at Buffalo, SUNY, whose guidance and support made this project possible. J. Theodore Peña first proposed the idea of working with the Earthquake House assemblage and his experience in material culture studies was instrumental in helping to develop the questions that formed the basis of my research. Throughout the process, his criticism and suggestions helped guide and push me to continually dig deeper, ask questions, and bring my work to the next level. Bradley Ault was a constant source of feedback and encouragement during the course of the project. His friendship and collegiality have been an inspiration and remain one of the most enduring and appreciated aspects of this experience. Steve Dyson contributed a vast knowledge of all things archaeological and provided invaluable advice in matters ranging from the initial proposal to the final write up. A special acknowledgement and thank you is due to David Soren, who generously permitted me to access the material and excavation records of the Earthquake House and whose classes as an undergraduate at the University of Arizona sparked my desire to pursue the study of Roman archaeology.

This project could not have been possible without the assistance of the Department of Antiquities, Cyprus, particularly its former and current directors Pavlos Flourentzos, Maria Hajicosti, Despina Pilides, and Marina Solomidou-Ieronymidou. I am extremely grateful to Limassol District Officer Yiannis Violaris and the staff of the Kourion Museum in Episkopi for their support and willingness to assist in my research. I also wish to express my gratitude to the faculty and students of the Archaeological Research Unit of the University of Cyprus, particularly Demetrios Michaelides and Stella Demesticha for their warm hospitality and assistance with questions that arose during my study of the material. Others to whom I owe many thanks for their advice, opinions, and friendship are Frank and Anthea Garrod, Mark Lawall and John Lund, Stuart and Helena Swiny, Danielle Parks, Sherry Fox, Sotiris and Bettina Hadjichara, Vance Watrous, Matt Buell, David Lightbody, Smadar Gabrieli, John Leonard, and Jackie McGuire.

Numerous institutions provided the financial and logistical resources necessary to complete my research. In 2007-2008 I received a Fulbright Fellowship to Cyprus and am indebted to the Cyprus Fulbright Commission and its former Executive Director, Daniel Hadjittofi for their sponsorship. Additional Fulbright support was provided by Anna Argyrou and Ioli Kythreotou who were an endless source of help and information during my stay. Institutional support and additional funding in Cyprus was supplied by the Cyprus American Archaeological Research Institute. I am especially grateful to CAARI Director Tom Davis for taking a personal interest in my

In the end, none of this would be possible without the unwavering support of my father Ben, my mother Roxanne, my sister Megan, and especially my dearly beloved Yota, who believes in me when I don’t believe in myself, picks me up when I fall, keeps me going when I don’t think I can, and makes my life far better than I ever dreamed it could be.

iii

TABLE OF CONTENTS Appendix A: Catalogue of Individual Room Assemblages.................................. 109

Acknowledgements............................................... iii Abbreviations Used in Text..................................... v List of Figures...................................................... vi

Appendix B: Kourion Amphora Typology and Capacity Estimates.......................... 135

1. Introduction....................................................... 1 1.1: Scope and Justification of the Present Study..... 1 1.2: Method and Theory......................................2 1.3: Material Culture Studies in the Greco-Roman World....................................................... 3

Appendix C: Tables........................................... 137 Bibliography...................................................... 141

2. Overview of Kourion: History and Excavation......... 6 2.1: Location and Topography............................. 6 2.2: Site History............................................... 6 2.3: History of Excavation................................. 17 2.4: The Earthquake Complex Excavations.......... 20 3. Methodology.................................................... 23 3.1: The University Museum Mission Excavations: 1934-35................................................... 23 3.2: The University of Arizona Excavations: 1984-87................................................... 24 3.3: The Present Study...................................... 27 4. Architecture..................................................... 31 4.1: Construction............................................. 31 4.2: Architectural Phases/Structural Development.. 32 4.3: Discussion................................................ 42 5. Room Stratigraphy/Formation Processes............... 44 5.1: Stratigraphic Anomalies.............................. 44 5.2: Room Stratigraphy..................................... 45 6. Room Assemblages........................................... 52 6.1: Terminology............................................. 52 6.2: Artifact Typology...................................... 52 6.3: Possible Factors Affecting Vessel Competeness............................................ 53 6.4: Room Assemblages................................... 54 7. Activities and Aspects of Household Organization...85 7.1: Methodologies and Calculations Used in Analysis.................................................. 85 7.2: Individual Room Activities......................... 86 7.3: Aspects of Household Organization.............. 97 8. Conclusion..................................................... 105 8.1: Architecture............................................ 105 8.2: Artifact Assemblage................................. 105 8.3: Directions for Future Research................... 107

iv

Abbreviations Used in the Text Alab Amph ASL AU AZLS BD BL BSK BVD ca. CA CC CDOA C’less cm Cook CRSW D DavisFN DavisFR DFN FE Fine Gab GFR Gran H Ind. KT I KT II KT III

: Alabaster : Amphora : Above Sea Level : Gold : Arizona Locus Sheet : Base Diameter : Bottom Level : Basket : Broken Vessel Density : Circa : Copper Alloy : Cypro-Classical : Cyprus Department of Antiquities : Colorless : centimeter : Cookware : Cypriot Red Slip Ware : Diameter : Davis Field Notebook : Davis Final Report : Daniel Field Notebook : Iron : Fineware : Gabbro : Gardener Final Report : Granite : Height : Indeterminate : Kourion Type I Amphora : Kourion Type II Amphora : Kourion Type III Amphora

KT IV/LR4 KTV/LR1 L LBA LC LR3 LRC LS lt. m Max. MaxD MBA MC Min. PB Plain RD RDAC SFN TL TAA TDA TUA TRIII Unk. UR VTC W

v

: Kourion Type IV/Late Roman 4 Amphora : Kourion Type V/Late Roman 1Amphora : Length : Late Bronze Age : Late Cypriot : Late Roman 3 Amphora : Late Roman C Ware : Limestone : liter(s) : meter(s) : Maximum : Maximum Diameter : Middle Bronze Age : Middle Cypriot : Minimum : Lead : Plainware : Rim Diameter : Report of the Department of Antiquities, Cyprus : Schiffer Field Notebook : Top Level : Total Available Area : Total Discard Area : Total Usable Area : Trench III : Unknown : Unregistered : Vasa Type C Amphora : Width

List of Figures⃰ Figure 1.1: Map of Cyprus. Megaw 2007, vi. Figure 2.1: Map of the Kourion District showing Bronze and Iron Age sites discussed in the text and the relationship between the acropolis, stadium, sanctuary, and basilicas. Figure 2.2: Plan of the Sanctuary of Apollo Hylates at Kourion. Adapted from The Sanctuary of Apollo Hylates at Kourion, Cyprus by David Soren. © 1986 The Arizona Board of Regents. Reprinted by permission of the University of Arizona Press. Figure 2.3: Plan of Kourion Acropolis. Modified from Megaw 2007, 51 fig 1A. Figure 2.4: Map showing the water conduits that supplied Kourion. Dotted sections indicate probable routes that have not been confirmed archaeologically. After Last 1975, 40 Map 1. Figure 2.5: Kourion beach basilica. Figure 2.6: Plan of Trench III showing architectural remains uncovered during excavation. Figure 2.7: Victims recovered in House 1, Room 1 by J. F. Daniel during the Trench III excavation. Courtesy of the Penn Museum, image #100369. Figure 2.8: Earthquake complex state plan. After Buell et al. 2009, fig. 1. Figure 3.1: Trench III excavation notebook page illustrating recording protocols discussed in text. DFN 6/14/1934, 38. Courtesy of the Penn Museum. Figure 4.1: Stone foundations of the Earthquake House showing construction techniques. Figure 4.2: Room 14 doorway illustrating the monolithic blocks used for the threshold and jambs with a stone block as a step. Figure 4.3: Pano Kivides, Cyprus. Traditional roof construction that would have been used extensively in the Earthquake House. Figure 4.4: Earthquake House – Initial Construction. Adapted from Buell et al. 2009, fig. 1. Figure 4.5: Area of previously hypothesized “second doorway” in Wall E between Room 1 and Room 8. Top: North face of Wall E looking south; Bottom: South face of Wall E looking north. Figure 4.6: Room 8 plan showing locations of semipermanent fixtures. Reconstructed by the author from AZLS 8. Figure 4.7: Ceramic fragments used as wall chinking in Wall T, Room 14. Figure 4.8: Earthquake House – Plan showing additions and modifications. Adapted from Buell et al. 2009, fig. 1.

Figure 4.9: Proposed “double” threshold in Wall D linking Room 2 and Room 3. Note the layer of fill between the upper and lower stones. Figure 4.10: Intersection of Wall J and Wall F showing the lack of evidence for a doorway to Room 6. Figure 4.11: Stone trough in Room 2. Figure 4.12: Blocked doorway in Wall E originally linking Room 1 and Room 8. Figure 4.13: Blocked doorway in Wall K joining Room 6/7 and Room 8. Top: Looking east toward Room 8. Note the “finished” appearance of the blocking and its flushness with the wall face. Bottom: Looking west toward Room 6/7 showing the unevenness of the blocking on this side. Figure 4.14: Room 15 and Room 18 plan indicating the locations of late features and loci. AZLS: Room 15. Figure 4.15: Comparative plans of the House of the Gladiators and the Earthquake House illustrating their difference in size. After Loulloupis 1971, fig. 1; Buell et al. 2009, fig 1. Figure 5.1: Room 1 – E-W section looking south. Reconstructed by the author from elevations recorded in AZLS from AZLS: TRIII/Room 1. Figure 5.2: Room 2 and Room 3 – E-W sections looking south. Reconstructed by the author from elevations recorded in AZLS: Room 2. Hatching in Room 3 indicates portion excavated in 1985. Dashed line dividing Locus 007 indicates horizon between earthquake and post-earthquake deposits. Figure 5.3: Room 6 and Room 7 – N-S sections looking west. Reconstructed by the author from elevations recorded in AZLS: Room 6 and AZLS: Room 7. Figure 5.4: Room 8 – N-S section looking east. Reconstructed by the author from elevations recorded in from AZLS: Room 8. Figure 5.5: Room 11 – E-W section looking north depicting the relationship between the doorway in Wall N and the strata recorded in the western part of Room 11. The elevation of the threshold block is the horizontal surface of Wall N. Reconstructed by the author from elevations recorded in AZLS: Room 8 and AZLS: Room 11. Figure 5.6: Room 12A and Room 14 – N-S sections looking west. Reconstructed by the author from elevations recorded in AZLS: Room 12 and AZLS: Room 14.

⃰ All images by author unless otherwise credited.

vi

Figure 6.28: V560 – Plainware basin. (Note: Missing rim fragments found in loose sherd material post-photograph.) Figure 6.29: V568 – Pithos. Figure 6.30: 579/601 – CRSW Form 1 bowl. Figure 6.31: (L – R) M178, M179, M185, M192 – CA objects, likely pins. Figure 6.32: M177 and M180 – Iron spearhead. Figure 6.33: M175 – Four tine iron fork. Likely an agricultural implement or possibly used for fishing. Figure 6.34: M176 – Lead “cup.” Top and side views. Figure 6.35: S75, S86, S87 – Stone mortarium. Figure 6.36: Room 14 plan. Reconstructed by the author from DFN: 6/4/1934, 21 and AZLS: Room 14. Figure 6.37: V617 – Pithos. Note ancient plaster repairs on side. Figure 6.38: Field photograph recording the group of objects recovered in the northeast corner of Room 14 during the Trench III excavation in 1934. A: P129-1934; B: P130-1934; C: P131-1934; D: P133-1934; E: ST13-1934. Photo courtesy of the Cyprus Department of Antiquities archives. Figure 6.39: V615 – KT II amphora. Figure 6.40: V619 – Vasa Type C amphora. Figure 6.41: V622 – CRSW Form B1 dish. Figure 6.42: V618 – Vasa Type 3G jug. Figure 6.43: G147 – Rim and fragments consolidated in situ during excavation. Figure 6.44: B58 – Worked bone object. Possibly an awl or punch. Figure 6.45: Room 20 Plan. Reconstructed by the author from AZLS: Room 20. Figure 6.46: UR/BSK 566 – Iron finger ring. Side and top views. Figure 6.47: UR/BSK 1412 – Iron and CA tintinnabulum. Object is uncleaned. Figure 6.48: Room 25 Plan. Reconstructed by the author from AZLS: Room 25. Figure 6.49: V677 – CRSW Hayes Form 12A jar. Note missing rim as a result of removal by maintenance processes. Figure 6.50: P684 – Two-handled pithos/amphora. Figure 6.51: S109 – Mortar. Figure 6.52: S110 – Pestle. Figure 7.1: Artifacts reflecting storage in Room 1, Room 2, and Room 3. Reconstructed by the author from DFN: 6/6/1934, 27 and AZLS: TRIII/Room 1, AZLS: Room 2, and AZLS: L9/Room 3. Figure 7.2: Provenienced coins from Room 1, Room 2, and Room 3. Reconstructed by the author from DFN: 6/6/1934, 27and AZLS: TRIII/ Room 1, AZLS: Room 2, and AZLS: L9/Room 3. Figure 7.3: Artifacts reflecting storage in Room 6 and Room 7. Reconstructed by the author from AZLS: Room 6 and AZLS: Room 7.

Figure 5.7: Room 12 and Room 19 – E-W sections looking south. Reconstructed by the author from elevations recorded in AZLS: Room 12 and AZLS: Room 19. Figure 5.8: Room 20 and Room 25 – E-W sections looking north. Reconstructed by the author from elevations recorded in AZLS: Room 20 and AZLS: Room 25. Figure 6.1: Room 1 Plan. Reconstructed by the author from DFN: 6/6/1934, 27 and AZLS: TRIII/Room 1. Figure 6.2: M26 – Unidentified CA object, possibly part of a buckle or clasp. Figure 6.3: P1 – Two handled pithos. Figure 6.4: P221 – KT IV/LR4 amphora. Note small hole drilled in base to facilitate emptying. Figure 6.5: S5 – Possible column drum or suspensura. Note traces of burning on surface of object denoting likely reuse. Figure 6.6: Room 2 Plan. Reconstructed by the author from AZLS: Room 2. Figure 6.7: BI2 – Bone pin, top and side views. Figure 6.8: BI5 – Bone pin. Figure 6.9: BI6 and BI7 – Carved bone object, probably inlay or veneer. Top and side views. Figure 6.10: M117 – Iron curb bit. Part of the mule tack assemblage. Figure 6.11: M116 - Unidentified iron and lead(?) object. Originally identified as a lock, but fits no known typology. From top to bottom: top view, back view, side view. Figure 6.12: S20, S21, S33, S34, S41, S42, S44, S47, S51, S52, S59 – Round marble tabletop with carved border. Figure 6.13: Room 3 plan. Reconstructed by the author from AZLS: L9/Room 3. Figure 6.14: 187 - Jug. Similar to Vasa Form 1H. Figure 6.15: 161/170/209 – KT III amphora. Figure 6.16: 174/196/ 197/ 208/ 213/ 214/ 217/ 390 Partially reconstructible pithos rim. The two lighter colored fragments on the left were recovered from Room 11; the remainder from the eastern side of Room 3. Figure 6.17: M48 – Unidentified lead object. Figure 6.18: BI1 – Bone pin with a metal covered head. Metal identified by the excavator as silver. Figure 6.19: Room 6 Plan. Reconstructed by the author from AZLS: Room 6. Figure 6.20: V443 – Pithos. Figure 6.21: V476 – Vasa Type C amphora. Figure 6.22: 520 – Plainware funnel. Figure 6.23: 474-1 – KT I amphora side and bottom view showing deliberately cut hole. Figure 6.24: G138 – Glass bottle. Figure 6.25: Room 7 plan. Reconstructed by the author from AZLS: Room 7. Figure 6.26: V552 – KT V/LR1 amphora. Figure 6.27: Room 8 Plan. Reconstructed by the author from AZLS: Room 8.

vii

Figure 7.4: Artifacts reflecting storage in Room 8. Reconstructed by the author from AZLS: Room 8. Figure 7.5: Artifacts reflecting storage in Room 14. Reconstructed by the author from AZLS: Room 14. Figure 7.6: Artifacts reflecting storage activities in Room 25 and Room 25. Reconstructed by the author from AZLS: Room 20 and AZLS: L9/Room 25. Figure 7.7: Plan of the Earthquake House showing victim locations. Plan adapted from Buell et al. 2010, fig. 1. Reconstructed by the author from AZLS: TRIII/Room 1, AZLS: Room 2, AZLS: Room 11, and AZLS: Room 20. Figure 7.8: Plan of the Earthquake House showing artifacts reflecting storage. Plan adapted from Buell et al. 2009, fig. 1. Figure 7.9: Detail plan of Room 12, Room 14, and Room 19. Plan adapted from Buell et al. 2009, fig. 1. Reconstructed by the author from AZLS: Room 12 and AZLS: Room 14. Figure 7.10: Cookpot V657 rim fragment (A) recovered from provisional discard area of Room 14. Additional joining fragments recovered from Room 19 (B and C) and Room 12A (D). Figure 7.11: Joining fragments of a Vessberg 18 lamp of the “galloping horse” variety. Unbroken discus (P635) and lower portion of nozzle recovered from the provisional discard area of Room 14, remainder from Room 19. Figure 7.12: Plan of the Earthquake House showing artifacts reflecting food preparation and food/drink consumption. Plan adapted from Buell et al. 2009, fig. 1.

viii

Chapter 1: Introduction Material culture, the human-made objects of everyday life1 surrounds us almost without notice. However, its existence is essential for all aspects of society to function, from the basic component of the individual to large military, religious, and governmental institutions. This basic fact was also true for the ancient Greco-Roman world. While generally associated with iconic images of monumental architecture and works of art, these civilizations were dependent upon the mundane articles of everyday life that today make up the bulk of the archaeological record. Because material culture is meaningfully constituted,2 its study by archaeologists, anthropologists, and social historians can be used to illuminate aspects of the people, society, and culture responsible for its creation. It is upon this premise that the present study is based.

century AD. This preserved not only the entire assemblage in situ, but also the remains of seven individuals who perished during the event. Such sudden and complete destruction makes Kourion a “Pompeii” of the late Roman Empire. It is one of only a handful of ancient Mediterranean sites that provide an opportunity to study artifact assemblages in the context of their daily use rather than as abandonment or discard deposits. Conditions of complete preservation without significant post-deposition disturbance combined with excavation using the most methodologically and theoretically sophisticated techniques available is relatively rare in archaeology, making the Earthquake House an exceptional site for studying domestic life in the Roman world. Because of this, analysis of the architecture and artifact assemblage has the potential to reveal a great deal about the processes, behaviors, and conditions present in the structure during its period of terminal use. These insights can then be used to elucidate general aspects of daily life at Kourion during the late 4th century AD.

1.1: Scope and Justification of the Study: The purpose of this work is to present a detailed examination of the architecture and artifact assemblage recovered from the “Earthquake House,” a multi-room structure located at the site of Kourion on the southwest coast of Cyprus (Fig. 1.1). Like the rest of the city, the Earthquake House was destroyed with little or no warning by a massive earthquake sometime during the late 4th

At present, only a few preliminary reports and one book for a popular audience have been published on the Earthquake House and its assemblage; a lacuna necessitated a detailed inquiry into the significant body

Figure 1.1: Map of Cyprus. Megaw 2007, vi. 1

For detailed definitions and discussions of what constitutes “material culture” see especially Deetz 1977; Prown 1982; Berger 1992; Schlereth 1982; Hurcombe 2007. 2 Hodder 1986, 4.

1

that necessitated a detailed inquiry into the significant body of data preserved in the structure. Contextual analysis of material culture assemblages has recently gained momentum within the larger framework of household archaeology, providing evidence that challenges long held assumptions concerning ancient domestic practices recorded by ancient authors. Studies of this type that rely predominantly on the archaeological evidence and only secondarily on literary models are particularly suitable to Roman Cyprus since the island is rarely mentioned in the written sources. This facilitates examining and interpreting the archaeological evidence on its own merit, eschewing any a priori assumptions influenced by accounts of architecture or domestic organization preserved in the textual record.

present study. The three subsequent chapters are devoted to investigating specific aspects of the archaeological data: Chapter 4 presents the architecture and provides a diachronic interpretation of how the structure evolved; Chapter 5 discusses the archaeological stratigraphy and formation processes that affected the structure both preand post-earthquake; and Chapter 6 presents a detailed room by room examination of the in use and associated artifact assemblages recovered during the excavations. Chapter 7 provides a synthesis of the architectural, stratigraphic, and artifactual material previously discussed. It offers interpretations concerning the overall organization of the structure as well as investigating evidence for the domestic economy. Finally, Chapter 8 provides a brief conclusion that discusses some of the results obtained from the study and offers possible directions for future work on the Earthquake House and the site of Kourion generally.

The need for studies of in situ assemblages, particularly from the period of Late Antiquity, has been stressed by Luke Lavan and others working in this area of inquiry. This approach has been advocated as a way to transform traditional paradigms of scholarship (based predominantly on architectural and literary evidence) to one that examines how structures and objects were used through contextual analysis and material spatiality.3 This change is believed to be particularly necessary in the field of late antique studies, since by this time structures could have been in use for significant amounts of time. As a result, the function of a particular space, even one that can be “identified” based on its architectural form, may have been completely different than originally intended; altered by the inhabitants’ particular needs and practices.

1.2: Method and Theory: The methodology used in material culture studies is partly derived from semiotic principles and linguistic theory developed by Ferdinand de Saussure. In his work, Saussure distinguished between parole, the words spoken by individual speakers, and langue, the system of language that makes any particular speech act possible. Based on these distinctions, comprehension is derived from understanding the entire underlying structure of the language, rather than any particular isolated component.5 These ideas were later adapted by the French anthropologist and ethnologist Claude Lévi-Strauss who incorporated them into the theoretical framework of Structuralism, whereby the meaning of cultural phenomena is derived through understanding the underlying systems (i.e. structures) that created them. A further progression of this theory was provided by the discipline of hermeneutics (interpretation), particularly the work of Paul Ricoeur. Ricoeur believed that material culture should not be viewed in terms of language, but rather as a text whose meaning is ultimately derived from the context in which it is read.6

While the value of such studies toward understanding late antique society has not been disputed, Lavan concedes that presently the “basic work on assembling corpuses of material has not been done and even analyses of rich occupation deposits are few and far between.”4 In light of these statements, this study of the Earthquake House architecture and artifact assemblage is timely, and seeks to address this significant deficit identified by late antique scholars. The purpose of this work is not to overturn the conclusions of the Earthquake House excavators regarding its primary function. Rather, it seeks to augment them in a number of ways and look at the archaeological data from a new and different perspective.

Integrating the basic principles from these bodies of theory permits individual objects to be interpreted as basic language components. When examined collectively, these components in effect become archaeological “texts.” Both the individual components as well as the “text” as a whole encode information about the culture that created it. These ultimately are “read” by the archaeologist within their find context to “decode” information about the culture “stored” by the artifacts.7

The following work is divided into eight chapters. Chapter 1 provides a brief background of the method and theory upon which material culture studies are based and previous scholarship employing this approach in the Greco-Roman world. Chapter 2 presents a historical overview of Kourion as it developed from the Bronze Age to the Byzantine period and as an archaeological zone beginning in the late 19th century. It is here that the Earthquake House excavations will be contextualized within the history of work carried out at the site. Chapter 3 focuses on the methodologies and procedures utilized in excavating the structure and the collection of data for the

5

Tilley 1990, 6. In discussing this theory, Tilley provides a useful comparison to the game of chess, where the basic understanding of the game is derived from the underlying system of rules that govern how the pieces are moved, not the individual moves that make up any particular match. 6 Moore 1990, 111-3; Hodder 1986, 153-4; Kingery 1996. The work of Ricoeur maintains the validity of the structuralist approach to material culture, since it emphasizes that the dimensions of the “text” can only be discovered by a thorough analysis of its “language” components. 7 Tilly, 1990b, 65; Hodder 1986, 4; 153; Van der Leeuw 1990, 94.

3 Lavan et al. 2008a. The arguments put forth by the authors echo those articulated by Allison in her work on Pompeii. 4 Lavan et al. 2008b, 3.

2

Based on this methodology, archaeological context, defined by Hodder as “the totality of the relevant environment,” is the primary means through which the material culture “text” can be accurately read and understood.8 Since each object exists within numerous contexts simultaneously (e.g. spatial, temporal, typological, etc.), careful examination and evaluation of the relationships between the two is critical. This process can lead to the recognition of patterns, whose presence or absence may reflect underlying attitudes and practices of individuals and societies concerning object creation, use, and disposal. Ultimately, it is the analysis of context that allows the “meaning” of an individual object and consequently the material record to be established and understood.

that may have occurred within a particular architectural space. Since this approach is dependent upon the recovery of well-preserved assemblages, studies of this type have largely been limited to sites that suffered catastrophic and sudden destruction by human and/or natural agents. Material from these sites has provided compelling physical evidence that impugns previous interpretations of Greco-Roman household organization, especially in areas such as gender specificity, domestic production, and religion. Detailed autopsy of artifact assemblages has formed the basis of several noteworthy studies in the Greek, Roman, and most recently, Byzantine worlds. One of the earliest was undertaken at Olynthos in Northern Greece; a site sacked by Philip of Macedon in 348 BC and subsequently abandoned. Because of how it was destroyed, Olynthos has provided one of the most comprehensive architectural and artifactual data sets for investigating aspects of the ancient Greek household.13 In conducting this research, Nicholas Cahill examined the assemblages from 13 of the 107 excavated houses, identifying groups of objects that indicated where specific domestic activities had been carried out. The results from individual houses were then compared to formulate general observations concerning organizational aspects of the ancient Greek household.14 Perhaps the most significant conclusion reached by Cahill’s study was that ancient domestic space was not functionally specific, particularly in terms of gender, a situation that had long been assumed and accepted by scholars. Although the assemblages from some rooms in the houses at Olynthos were more indicative of female activities (e.g. food preparation, weaving, etc.), in general, the evidence did not support the type of gender exclusivity found in sources such as Xenophon. On the contrary, the data indicate that use of domestic space appears to have been relatively fluid, with a single space capable of being the locus for many different activities.15

1.3: Greco-Roman Material Culture Studies: Studies of material culture have long been accepted as a method of inquiry in New World and American historical archaeology.9 For the majority of its history as a discipline, however, Classical archaeology has been reluctant to adopt such theoretical frameworks as investigative tools, clinging tenaciously (some would say detrimentally) to its antiquarian, art historical, and philological roots.10 Because of this, interpretations of remains from Greco-Roman sites have traditionally relied on accounts preserved in the textual record. The material evidence was largely considered of secondary importance and used predominantly to support accounts of domestic activities recorded by authors such as Xenophon, Vitruvius, and Pliny the Younger. As a result, early excavations focused primarily on recovering architectural ground plans, with the identification and function of individual spaces regularly assigned based on their similarity to textual descriptions. In such cases, the artifacts counted for little in determining room function unless they happened to support the textually derived interpretation.11 Except in certain circumstances (i.e. where the object was deemed to possess either intrinsic or artistic value), artifact assemblages were traditionally only superficially examined. If a detailed study was undertaken, objects were often de-contextualized, examined by specialists (mainly according to typology), and interpreted based on questions concerning trade, production patterns, or chronology.12

A similar study to Cahill’s was undertaken by Bradley Ault, who examined the assemblages from five houses at Halieis in the northeast Peloponnesus. The site was suddenly abandoned for unknown reasons sometime at the end of the 4th century BC and not subsequently disturbed. This resulted in the recovery of a significant quantity of in-situ artifacts during excavation.16 Although the number of houses examined by Ault is smaller than Cahill, the Halieis dataset is more comprehensive since the assemblages were recovered stratigraphically and virtually all of the objects were recorded as part of the excavation notebooks. This permitted the entire corpus of artifacts from each excavation unit to be assembled and examined together

Recently, a new research paradigm has arisen in classical archaeology, particularly in the area of domestic studies that advocates a holistic and contextual approach for studying artifact assemblages. In this model, all objects are recovered, with their findspots recorded and examined to identify the types and spatial distribution of activities 8 Hodder 1986, 143. In this definition, “relevant environment” refers to any significant relationship that is necessary for determining a particular object’s meaning. 9 Deetz 1977; Ames 1981; Fleming 1981; Meltzer 1981; Schlereth 1982; 1985. 10 For discussion of this topic see especially Dyson 1999, 1998, 1993, 1989; Renfrew 2004, 1980. Ault (2005, 2-3) provides a good overview of the issue. 11 Allison 1995, 148. 12 Allison 2004, 4, 175; 1997a, 78; Ault 2005, 3; Dyson 1995, 33.

13

Cahill 2002, 45. For ancient accounts of gender separation in the Greek house see especially Xen. Ec. 9.2-11; Aristoph. Thes. 424-17; Lysias 1.9, 3.6; Vitruvius 6.7.2. Discussion on this subject is provided by Jameson 1990, 104; Cahill 2002, 148-53; Nevett 1999, 19-20; 1994, 103. See also Conkey and Spector 1984. 15 Cahill 2002, 191-3. 16 Ault 1994, 30-1. 14

3

within its find context.17 In addition to providing confirmation for the flexible nature of domestic space and lack of rigid segregation between males and females within the Greek household, Ault was able to elucidate numerous aspects of ancient domestic life including evidence for discard processes, economy (particularly household production), and the practice of domestic cult.18

space at Pompeii was predominantly mono-functional,22 a condition contradicted not only by domestic assemblage studies from the Greek world but also ethnographic studies that indicate domestic space is only rarely limited to a single function.23 While the uncertainty of continuing seismic activity at Pompeii may have had an impact on the distribution of artifacts, the “unsettled” nature of the Pompeii assemblages more likely reflect multiple activities occurring within a single architectural space combined with the behavioral vicissitudes of individual inhabitants.24

In the Roman world, the principle contextual material culture study was conducted by Penelope Allison on the assemblages from 30 atrium houses at Pompeii. While it should theoretically provide the optimum scenario for this type of research, in reality, the situation at Pompeii was complicated by a variety of factors including ancient post-eruption salvage and a lack of systematic excavation methodology/recording procedures throughout much of its history.19 Because of the latter, it was necessary for Allison to reconstruct the in situ artifact assemblages recovered from each house. This involved using accounts preserved in excavation archives such as the Giornali degli Scavi, which are notoriously lacking in detailed recording of object descriptions and find contexts. While the work by Allison in reconstructing these assemblages is commendable, her dataset is nevertheless irrevocably compromised since the documented objects almost certainly account for only a portion of what was discovered.20 Despite this problematic aspect, Allison’s work made significant contributions to Roman material culture studies and household archaeology. Her results revealed no apparent correlation between the assumed function of a space based on its architectural interpretation and the actual activities that occurred within it based on the artifactual evidence.21 This result is consistent with those obtained from the aforementioned studies from the Greek world. In explaining the variance between her observations of artifact and activity distribution in Pompeian households and the literary models, Allison looks to the period between the AD 62 earthquake and the AD 79 eruption for explanation. Rather than a static, stable interval between these two major events, probable continuous seismic activity would have caused a certain degree of uncertainty and instability within the population. These conditions are reflected in the distribution of artifacts and activities, which were interpreted by the early excavators as evidence that Pompeii was a city in decline when it was destroyed. Allison believes that these conditions caused the “unsettled” nature observed in some of the assemblages and accounts for the presence of objects in spaces where they did not “belong.” However, this interpretation assumes that prior to the AD 62 earthquake domestic

Since Allison’s study, contextual analyses of Roman material culture have continued to gain momentum. Most notable is the work of Joanne Berry, who examined the assemblage from the Casa di M. Epidius Primus (I.8.14) at Pompeii. A wide range of artifacts with a variety of functions were recovered throughout this structure, complicating the identification of activity areas within individual rooms. Most spaces produced mixed assemblages, an indicator of multiple activities taking place within a single space. The conclusions of Berry are informative and indicate that the distribution of artifacts within a structure is more likely to reflect patterns of storage rather than use. Because of this, true activity areas may be difficult to discern since their location was probably not fixed within a particular architectural space and likely organized based upon temporal/seasonal factors rather than spatial ones.25 While Berry acknowledges that domestic artifact assemblages are by no means a “magic bullet” for unlocking the social and spatial dynamics of the Roman house, she stresses that combining their analysis with the architectural data constitutes a valuable line of primary evidence that leads to a more complete understanding of household activities and organization. In recent years, investigations of Roman domestic artifact assemblages have begun to expand geographically outside of Greece and Italy proper (predominantly in the Roman/Byzantine east) and temporally to include the late Roman and early-middle Byzantine periods.26 For example, Crawford systematically examined assemblages from a series of what he interpreted as shops destroyed by fire at Sardis in the 7th century AD.27 At the site of Pella in the Jordan Valley, work has begun on the contextual analysis of domestic assemblages from a set of six courtyard houses destroyed by a massive earthquake in the mid-8th century AD. Preliminary work on these has provided significant insights into the use of domestic space during this period but, at present, no comprehensive study and publication from these well preserved deposits

17 Ault 1994, 12. In comparison, at Olynthos it was necessary for Cahill to incorporate “data for ca. 3000 objects which were omitted from the final publications” and correct “the published provenience of another 1500.” This number does not include artifactual material discarded during the excavation of which no record was made. 18 See Ault 1999; 2005. 19 Allison 2004, 3-4. For a summary of the history of excavations at Pompeii see Nappo 2004, 16-9. 20 Allison acknowledges this problem in the data set (2004, 33). 21 Allison 2004, 157, 201. See also Adam 1989.

22

Berry 1997, 185. Kent 1984; 1987. Nevett 2010, 98-113. 25 Berry 1997, 194. 26 Crawford 1990; Walmsley 2007; Putzeys et al. 2008. 27 Cf. Crawford 1990. To the knowledge of the author, this is the only contextual material culture study undertaken to date on a non-domestic structure. The identification of these spaces as retail installations has been challenged by Harris (2004) who believes that they were used for domestic occupation during their final phase. 23 24

4

has been produced.28 To the knowledge of the author, the most recent study to employ this methodology is one proposed at Ephesos by the Institute for Culture History of Antiquity of the Austrian Academy of Sciences. This multi-year project intends to examine the assemblages recovered from Unit 7 of Terrace House 2 in order to better understand the “use of space and household activities in the various building periods.”29 The present work on the Earthquake House exists in the same vein of scholarship as those previously mentioned, and seeks to contribute to the ever growing body of artifact assemblage data available to archaeologists. It is hoped that this work will serve as a useful tool for future studies and offer insights into the processes of daily life in the ancient world. It is the first study of this type to be carried out in Cyprus, and aims to provide a significant addition to the growing corpus of scholarship of the island for the late Roman period.

28

Cf. Walmsley 2007. Thür 2006, 4. Special thanks to Dr. Bradley Ault, one of the referees for this proposal, for informing me of this project. 29

5

Chapter 2: Kourion: History and Excavation The following chapter provides an overview to Kourion as a city and as an archaeological site. Such an overview is a standard feature of previous scholarship, most notably the Swiny guidebook on Kourion and the Akrotiri peninsula and the Christou guidebook that focuses primarily on the excavated monuments of the ancient city. Since the publication of these works, a great deal of new fieldwork and research has been undertaken that has enhanced our knowledge and understanding of Kourion. The present discussion does not attempt to fundamentally alter previous accounts of the site or its history. Rather, it utilizes them as a foundation upon which the results of subsequent studies are added, thus updating the picture of Kourion as it exists today.30

part of the Pakhna Formation. This geological unit is composed of bioclastic limestone and formed by uplift from the subduction of the African plate below Cyprus.31 The promontory measures ca. 900 meters long, 700 meters wide, and rises to an elevation of ca. 80 meters above sea level. It forms a natural boundary that delineates the western edge of the coastal plains and valleys that stretch along the south coast of the island to the east. The location of Kourion is recorded on the Tabula Peutingeriana as being 16 miles west of Amathous, and 22 miles east of Paphos.32 2.2: Site History Despite the dominance of the Kourion acropolis over the surrounding landscape, evidence for its occupation can only be traced archaeologically to the Cypro-Classical period (475-325 BC). Of the remains observable today, the majority belong to the Roman and Late Roman /Byzantine phases of the site from the mid-late 1st – 5th/6th centuries AD. The lack of remains pre-dating the CyproClassical is due to the fact that Kourion was not a single

2.1: Location and Topography The site of Kourion (Latin Curium) is located on the southwest coast of Cyprus where the western edge of the Akrotiri peninsula curves to form Episkopi Bay (Fig. 2.1). Dominating the topography of this area is the Kourion acropolis (Palaikastro), a large outcrop that is

inhabited sitediscussed throughout its history like Figure 2.1: Map of the Kourion District showingcontinuously Bronze and Iron Age sites in the text and the relationship between the acropolis, stadium, sanctuary, and basilicas. 30 Comprehensive publication of Kourion excavations, while better than many other sites in Cyprus is still generally lacking. The results of numerous major excavations and their conclusions concerning the overall chronology and development of the site are only cursorily described in preliminary reports. The material presented in this chapter is derived from these published sources. However, the situation at Kourion appears to be changing with a number of major works reportedly nearing completion.

31

For a more detailed description of this formation see Bullard 1987 in Soren 1987a; Eaton and Robertson 1993. 32 Bekker-Neilson 2004. For Tabula Peutingeriana dating see Bosio 1983, 154; Weber 1989, 116-7.

6

the “tells” of the Near East or cities such as Athens or Rome. Instead, it developed as a series of settlements in the same general area that were occupied and dominant during different chronological periods.

period (ca. 475 BC). This necropolis, east of Kourion at Kaloriziki, has provided valuable information about Cypriot culture during this transitional period, whose wealth and sophistication is attested by the elaborate grave goods found in the tombs.37 The true value of the Kaloriziki necropolis, however, is in providing confirmation that habitation in the general area continued and flourished following the abandonment of Bamboula. It thus forms a bridge between the prehistoric and historic eras when the site known today as Kourion begins to emerge in the historical and archaeological record.38

2.2.1: Prehistoric Herodotus records that Kourion was founded by Argive settlers from Greece at the end of the Mycenaean period.33 Archaeological investigations, however, have discovered remains that indicate the area was first inhabited during the Neolithic period, with a significant settlement at Episkopi Phaneromeni established in the Middle-Late Bronze Age (MBA/MC – LBA/LCIA). This site was gradually abandoned in favor of a ridge top location 500 meters to the northwest at Episkopi Bamboula (LBA/LC). During the transition from the Late Bronze to early Iron Age (normally placed around 1050 BC in Cyprus) Bamboula was abandoned for a new and as yet undiscovered site that served as the seat of the Iron Age kingdom. Although evidence for settlement atop the Kourion Acropolis during these early periods is lacking, recent archaeological investigations have revealed possible Bronze Age activity in the form of three rock-cut tombs discovered during excavations of the Forum/Agora area by the Cyprus Department of Antiquities (CDOA). They have not been published and are only cursorily described in a preliminary report, which states they were looted and/or filled in sometime during the Hellenistic period.34 The description of a large chamber accessed by means of a stepped dromos is consistent with late Bronze Age tombs discovered in the area, most notably at Bamboula.35 However, more irregularly shaped tombs with sloping stepped dromoi also occur on Cyprus from the latter part of the Cypro-Geometric period (roughly 850-750 BC) to the 2nd century BC.36 The presence of these tombs indicates funerary activities occurred on the Kourion acropolis prior to the Hellenistic period, but its chronology, duration, and intensity cannot presently be determined.

The first written record mentioning Kourion specifically is a cuneiform inscription on a clay prism from Nineveh dating to the reign of the Assyrian king Esarhaddon (680669 BC).39 Inscribed in 673/2 BC, the prism commemorates rebuilding the royal palace at Nineveh and records the names of ten Cypriot rulers who sent raw materials to support this project. One of these is “Damasu, King of Kuri,” which is generally accepted as “Damasos, King of Kourion.”40 The same individual is recorded in an inscription of Assurbanipal (668-626 BC), the successor of Esarhaddon that chronicles a military campaign undertaken against Egypt. Among the list of vassals supporting this expedition are the ten kings of Cyprus previously recorded on the prism of Esarhaddon, who “brought their rich (lit. heavy) gifts [before me] and kissed my feet. Those kings], together with their forces…”41 Unfortunately, the remainder of the text in this section is lost, as it seems to indicate that the vassal kings of Cyprus not only provided financial support but also contributed troops for the cause. While the rich tombs discovered in the Kaloriziki necropolis demonstrate that Kourion possessed the wealth spoken of by the inscription, it also appears to attest that the city and its chora had a sufficient population base to provide

37 Buitron-Oliver 1997, 27-8. Among these is the famous “Scepter of Kourion.” 38 Swiny 1982, 51. The location of the settlement to which the necropolis belonged is a matter of debate and several possibilities have been put forth. These include a new settlement on the Kourion acropolis itself (Daniel 1937; McFadden 1954), on the plain adjacent to the necropolis (cf. Benson 1973), or possibly underneath the village of Episkopi. 39 Although it has been proposed that the kingdoms of Cyprus first emerged during the Late Bronze Age (cf. Baurain 1984 for the textual and archaeological evidence), this is not without controversy (Merrillees 1987; Masson 1990). The earliest un disputed epigraphic testimony for the presence of kingdoms in Cyprus is the Stele of Sargon II (721-705 BC), which records the tribute of seven kings from Cyprus. However, it does not list the kingdoms individually. 40 Luckenbill 1927, 265. This object forms the basis for the division of Cyprus into at least ten kingdoms by the second quarter of the 7th c. BC. The kingdoms listed on the prism are: Edi’al (Idalion), Kitrusi (Chytroi), Sillûa (Salamis or Soloi), Pappa (Paphos), Silli (Salamis or Soloi), Kuri (Kourion), Tamesu (Tamassos), Kartihadasti (usually equated with Kition), Lidir (Ledra), and Nûria (Uncertain). The “Display Inscriptions” of Sargon II from Khorsabad as well as the “Sargon Stele,” supposedly erected in the area of Citium about 707 BC both mention the kings of Cyprus that submitted voluntarily to the Assyrian king but provides neither the names of the kings nor their associated kingdoms (Luckenbill 1927, 35-6). 41 Luckenbill 1927, 341. Based on the dating of the Prism of Esarhaddon, the continuing rule of all the kings into the reign of Assurbanipal is possible.

2.2.2: Cypro-Geometric – Cypro-Classical The primary data concerning settlement in the Kourion area between the Bronze and Iron Ages comes from a large necropolis dating from the very end of the LBA/LCIIIB to the beginning of the Cypro-Classical 33

Hdt. 5.113 Christou 1985, 269 ff. 35 Swiny 1982, 43. Cf. Benson 1972. 36 Vessberg and Westholm 1956. See also Rupp 1978 for another chamber tomb fragment on the eastern slopes of the Kourion acropolis beneath a Hellenistic pebble mosaic. Only a fragment of the tomb dromos is preserved. Rupp states “So far on the acropolis of Kourion, the earliest small chamber tombs to be excavated date only to the fourth century.” No reference for this conclusion is provided, but it seems to be based on the report of Karageorghis (1977, 775) where he states “Trois tombes de la période hellénistique ont été foullées dans les secteurs Sud-Est et Sud-Ouest du site. Elles comportaient un chamber funéraire unique et étaient taillées dans le rocher.” The presence of White Painted Ware may indicate a date during the Cypro-Geometric or Cypro-Classical period, but the absence of any context for these finds makes their use problematic. The statement by Karageorghis seems to contradict the statement of Christou in his report. It seems unlikely that the Hellenistic constructions excavated in the forum/agora of the city would have been built over an active or recently abandoned necropolis. 34

7

briefly in the ancient sources (most notably Strabo)42 and its identity has been confirmed through epigraphic evidence discovered at the site.43 During this early phase of the sanctuary, religious activity was focused around a circular open air altar made of rough stones that may have been enclosed by a temenos.44

military resources for overseas campaigns as well as internal security. Archaeologically, the earliest remains from Kourion date to the Cypro-Archaic period – roughly contemporaneous with the Esarhaddon inscription. However, they do not come from the city site but from the extra-urban sanctuary of Apollo Hylates located ca. 2 kilometers to the northwest (Fig. 2.2). The sanctuary is mentioned

A layer of fine gray ash containing the burned bones of young sheep and/or goats was found among and on top

Figure 2.2: Plan of the Sanctuary of Apollo Hylates at Kourion. Adapted from The Sanctuary of Apollo Hylates at Kourion, Cyprus by David Soren. © 1986 The Arizona Board of Regents. Reprinted by permission of the University of Arizona Press. 42

Strab. 14.6.3. See also Ael. N.A.. 11.7. The most notable being an inscribed stele found by the University of Pennsylvania on the first day of excavations, which reads “ΤΕΜΕΝΟΣ ΑΠΟΛΛΩΝ [ΟΣ].” See McFadden 1938a. 44 Scranton 1967, 6-8; Soren 1987b, 29; Sinos 1990. Evidence for a temenos is elusive. 43

8

of these altar stones with a significant number of Archaic terracotta votive figurines (7th – 6th centuries BC).45 During the late Cypro-Archaic to late Cypro-Classical periods (beginning 6th – mid 4th centuries BC), the sanctuary expanded through the addition of numerous structures, including the main street, a simple first “temple” (or some type of sacred structure below the later Temple of Apollo), the west enclosure, the round structure, and most probably the first stoa of the East Complex.46 Despite these and more radical subsequent architectural modifications, the Archaic precinct remained intact in its original form until the sanctuary ceased functioning in the 4th century AD.47

Following this event, Kourion disappears from the historical record for the remainder of the 5th and first half of the 4th centuries BC. While Cypriot contingents participated in the expeditions against the Greeks mounted by Xerxes, it is not known how or if Kourion contributed to these campaigns. The archaeological evidence indicates it was during this period that the acropolis was initially occupied and fortified. Unfortunately, contemporary dedicatory inscriptions from the sanctuary are generally short and do not provide significant evidence to reconstruct the early social and political history of the site.53 In 351/0 BC, Kourion was one of the nine extant kingdoms that declared independence during an uprising against Artixerxes III and the Persians.54 This revolt proved abortive, with all the kingdoms except Salamis reconquered by the following year.55 The period during which these events occurred is archaeologically unremarkable, consisting predominantly of short dedicatory inscriptions from the Sanctuary of Apollo, which had undergone little or no architectural development since the 6th century BC.56 An inscribed block discovered northwest of the city indicates the presence of a second sanctuary or shrine at Kourion, although its precise location has not been determined. The inscription, in Greek and Cypro-Syllabic script, belongs to a marble statue base that records a dedication on behalf of Hellooikos, son of Poteisis to Demeter and Kore.57 Based on the letter forms it has been dated to the late 4th c. BC – the transition between the Cypro-Classical and Hellenistic periods. Its bilingual nature is characteristic of this period when Greek fully replaced the earlier syllabic writing system. The presence of this inscription indicates that Demeter and Kore were worshipped at Kourion in some manner. However, its recovery in a reuse context makes it difficult to determine what role these deities played in the religious life of the city, as well as whether their cult was centered on the Kourion acropolis or in an extra urban sanctuary like that of Apollo.

Following the breakup of the Assyrian Empire in 612 BC, Cyprus was initially conquered by Ahmose II/Amasis (569-525 BC) of Egypt. Eventually it became part of the Persian Empire, perhaps as early as 546 BC and almost certainly before the expedition of Cyrus against Babylon in 538 BC when the Cypriots voluntarily placed their forces at the disposal of the Persian leader.48 Because of this, the island and city kingdoms were granted a certain amount of autonomy, retaining their own rulers who maintained their royal status.49 The role played by Kourion in these events is not recorded, and no further mention is made of the island until the beginning of the 5th century BC. At this time, the city kingdoms of Cyprus led by Prince Onesilos of Salamis, attempted to throw off Persian domination following the Ionian cities in Asia Minor. To this end, Onesilos formed a coalition of all the city kingdoms except Amathous, which remained loyal to the Persians.50 A battle between the two sides occurred on the plain of Salamis that initially appeared to favor the Cypriots, especially following the death of the Persian general Artybius at the hands of Onesilos and his Carian shield-bearer. However, the sudden defection of the Salaminian war-chariots and Stanisor, King of Kourion, who commanded a significant number of troops, eventually led to a Persian victory and the death of Onesilos.51 The timely defection/betrayal of Stanisor seems to have earned Kourion a degree of clemency and nominal autonomy from the Persians not afforded to other city-kingdoms such as Paphos and Soloi that continued to resist Persian control.52

While archaeological evidence for the development of Kourion up to the late 4th century BC exists, it is rather ephemeral. The majority of what is known about the city until this time is derived from written sources or the scant epigraphic record. However, the remains preserved at Kourion begin to speak louder in the period following the death of Alexander the Great, creating a much clearer picture of the site and its history from the end of the 4th c. BC to its ultimate destruction during the Arab raids of the 7th c. AD.

45 Buitron-Oliver 1996; Christou 2001, 71. The votives depict warriors, horses and riders, chariot groups, male figures in long robes playing the lyre, centaurs, bulls, and other animals. 46 Soren 1987b, 39. 47 Buitron and Soren 1982, 62. 48 Herodotus (3.19) is not specific regarding the date of Cyprus allying itself with the Persians. Xenophon (Cyr. 7.4.1; 8.6.8) indicates the Cypriots assisted Cyrus with his campaign against the Carians and willingly joined the expedition against Babylon. 49 Xen. Cyr. 7.4.2; 8.6.8 50 The events surrounding the revolt, battle, and outcome are recorded by Herodotus (5.108-113). The reason for Amathous’ refusal to join the revolt is not recorded, but may have been due to a large Phonecian component among the population. This reason, however, is problematic since Kition, a city founded by the Phonecians, joined in the revolt (Hill 1972, 118.). 51 Hdt. 5.113. 52 The kingdom of Paphos was besieged before eventually being reduced and forced to accept a garrison of Persian troops. Soloi resisted

the longest, but was eventually taken by mining the fortification walls of the city (Hdt. 5.115). 53 Christou 1997, 916-7; see also Mitford 1971, 45-6 (#17), 46-9 (#18). 54 Diod. Sic. 16.42.4. Beloch (1923, 285-287) places this expedition in the year 344 BC and lists the extant kingdoms as Salamis, Citium, Amathus, Paphos, Marium, Soloi, Lapethos, Kerynia, and Curium. 55 Diod. Sic. 16.46.2. 343 BC in the accounting of Beloch. 56 Soren 1987b, 39. 57 C.f. Mitford 1971, 62-4 (# 26).

9

2.2.3: Hellenistic Cyprus remained a Persian possession until 332 BC. Following the Battle of Issus, the island transferred its allegiance to Alexander and dispatched a fleet of 120 ships to Sidon.58 The Cypriots accompanied the fleet to Tyre, a city whose capture Alexander deemed necessary before pursuing Darius. During the subsequent naval blockade, the Tyrians attacked the Cypriot ships resting at anchor. In the first exchange, the flagships of Pnytagoras of Salamis, Androcles of Amathus, and Pasicrates of Kourion were sunk with the remainder of the force driven ashore and broken up.59 The fate of Pasicrates following the defeat is not recorded, but it is possible that he survived since Pnytagoras is recorded as being rewarded by Alexander for his service and Androcles is mentioned by Arrian as one of the individuals that allied with Ptolemy in 321 BC.60 Following the account of the debacle at Tyre, Kourion virtually disappears from the historical record preserved in the ancient sources.

administration imposed on Cyprus was largely based on a military command structure, consisting of a governor general with the title strategos. This office, whose holder was directly responsible to the king of Egypt, encompassed a range of civil and military responsibilities.64 Under the strategos was an adjutantgeneral or quartermaster-general with the title grammateus. Garrisons stationed in smaller towns were led by commanders originally known as phrourarchoi, but later called ὁ ἐπὶ τῆς πόλεως (officer commanding the city).65 Kourion appears to have been a prosperous, moderatelysized settlement throughout much of the Hellenistic period. The epigraphic record provides evidence for numerous governmental officials, most notably Demetrios, son of Machatas who is recorded as the ὁ ἐπὶ τῆς πόλεως on a monument erected by the city in the Sanctuary of Apollo during the early 2nd century BC.66 Among the civic institutions at Kourion, the most notable is the mention of a boule from a fragmentary civic decree likely dating to the reign of Ptolemy II Philadelphos (283246 BC). This provides the earliest evidence for the existence of this council in Hellenistic Cyprus.67 The offices of archon, grammateus, civic strategos, and agoranomos at Kourion as well as the civic priesthood of ἱερεὺς τοῦ Ἀπόλλωνος are all similarly attested through inscriptions recovered at the site.68 Collectively, the epigraphic evidence indicates that Kourion enjoyed a certain degree of autonomy in its day to day affairs while remaining subject to representatives of the central government who were ultimately responsible for its administration.

During the struggle of the diadochoi following Alexander’s death, Cyprus was coveted for its strategic location and economic importance as a source of timber and copper. In 321 BC, Ptolemy I secured an alliance with four kings of Cyprus that strengthened his claim to the island; first against Perdiccas and later against Antigonos Monopthalmos and his son Demetrios Poliorcetes.61 Cyprus came under Antigonid control in 306 BC following the defeat of Ptolemy by Demetrios and his fleet at the Battle of Salamis. The continuous warfare between the various Hellenistic “kings,” however, ultimately caused Demetrios to abandon the island, which surrendered itself peacefully to Ptolemy in 294 BC.62

While the archaeological remains from the Hellenistic Period are far more numerous than those of previous eras, they are largely obscured by building from subsequent occupation phases. The remains that do exist, however, show the development of the site into a proper Hellenistic city with the standard accompanying structures. The most conspicuous of these is the theater, located on the southeast end of the acropolis north of the Amathus gate (Fig. 2.3, I). Although substantially modified during later periods, the earliest remains indicate a standard Greek

During this time, Cyprus underwent an administrative reorganization that was largely complete by the final Ptolemaic possession. Among these changes was suppressing or significantly reducing the power of the old city kingdoms and dividing the island into four administrative districts centered on the cities of Paphos, Amathus, Salamis, and Lapethos.63 The reorganization also transferred the capital from Salamis to Nea Paphos on the west coast of the island. This change was believed necessary since Salamis had assumed the role of leading city for much of the 4th century BC and was a stronghold of support for the Antigonid regime. The political

64 Hill 1972, 175. The primary work on the strategoi and other officials in Cyprus is Cohen (1912). In the time of Epiphanes (203-181 BC) the function of high priest (archiereus) was added to the duties of the strategos. 65 Hill 1972, 146-47; Cohen 1912, 42 ff. Hill equates this title with that of “Prefect” and suggests the change indicates a widening function of this office. It is argued that this change was due to the harsh military connotation associated with phrourarchos, but this is refuted by Bagnall (1976, 50). 66 Mitford 1971, 91-4 (#42). Watkin 1988, 281. At present this is the earliest attestation of this office on Cyprus. 67 Mitford 1971, 74-77 (# 32). Watkin 1988, 279. Bagnall and DrewBear 1974, 179-83. See Watkin 1988, 280 for arguments concerning the establishment of the demos during the Hellenistic period. 68 Archon cf. Mitford 1971, 79-81 (#34); Watkin 1988, 278-9; Grammateus cf. Mitford 1971, 100-105 (#s 46, 47, 48); Watkin 1988, 281-2; Civic Strategos cf. Mitford 1971, 81-3 ( #35), 100-4 (#s 46, 47); Watkin 1988, 284-5; Agoranomos cf. Mitford 1971, 79-81 (#34); Watkin 1988, 277-8; Civic Priesthood cf. Mitford 1971, 108-9 (#52) [79-81 (#34) and 89-91(#41) Probable]; Watkin 1988, 283-4.

58

Arr. Anab. 2.19.20. Arr. Anab. 2.22.2. Hill notes that Arrian utilizes the title βασιλεύς only in reference to Pnytagoras, which he believes may be an indication that Androcles and Pasicrates were not the actual kings of their respective cities. However, it is possible that the title was omitted but understood for the other two (cf. Peristianes 1910, 276 no.1.) 60 For Pnytagoras see Arr. Anab. 2.21.8, 9; 2.22.1-5. For Androkles see Arr. Diad. fr. 24.6. He also appears in a Delian inventory of ca. 313 BC (IG XI.2.135) where he is recorded as: Ἀνδροκλῆς Ἀμαθουσίαν Βασιλεύς. 61 Arr. Diad. (ed. Roos, II, 280 ff.). The four kings allied to Ptolemy I were Nicocreon of Salamis, Nicocles of Paphos, Pasicrates of Soli, and Androcles of Amathus (cf. Hill 1972 156). 62 Plut. Vit.Demetr. 35, 38. Salamis, the stronghold of Demetrios, held out after the remainder of the island surrendered but eventually capitulated. 63 Swiny 1982, 90. 59

10

11 Figure 2.3: Plan of the Kourion Acropolis. Modified from Megaw 2007, 51 fig. 1A.

style plan with a round orchestra and high proskenion. In Greek fashion, the structure took advantage of the sloping topography in this area to form the cavea foundation. Based on its architectural form the theater was believed to have been constructed during the Hellenistic period with a more precise date of the latter part of the 2nd century BC established based on numismatic evidence recovered during excavation.69 Northeast of the theater, vestiges of what appears to be a large domestic structure have also been discovered, whose earliest phase appears to date to the late Hellenistic period (150-50 BC). However, these early remains have been largely obscured by substantial modification and reuse during the Roman and late Roman periods.70

been determined. Based on its size, construction, and location, the excavator believes it is either some type of public building, or a large private house.73 Given its location, it is possible that the remains are those of the Hellenistic gymnasium, whose presence is epigraphically attested at Kourion, but whose precise location remains unknown.74 Excavations in the Sanctuary of Apollo recovered a large concentration of coins and pottery from the late 4th c. BC, which coincides with an apparent period of major building activity. This included the construction of an ashlar building in the southeast part of the site,75 the stoa of the East Complex (dated ca. 300 BC), and a house that occupies much of the adjoining area south of the East Complex.76 Development of the sanctuary continued throughout the Hellenistic period, although the architectural evidence is obscured by successive remodeling during Roman times. The presence of a prytaneion is attested by an inscription recovered from the South Building, but like the gymnasium at the city site its exact location has yet to be determined.77 When the architectural and artifactual evidence are examined together,78 the picture of the sanctuary that emerges is one of a religious center whose architectural growth parallels that observed in the city and whose popularity may have been one of the reasons for the prosperity enjoyed by both.

On the extreme eastern edge of the acropolis, the remains of a pebble mosaic were discovered that have been dated stylistically to 215-185 BC. Lack of an associated archaeological context for this mosaic precludes a definitive identification, but its location over an underlying earlier chamber tomb led to the interpretation that these features formed part of a heroön used to worship a deceased individual.71 In the central area of the acropolis, the discovered remains indicate that the city’s infrastructure was enhanced by the construction of significant water storage facilities during the Hellenistic Period. Among these were a large rock cut reservoir measuring 9.5 X 8.5 X 3.5 m located in the Hellenistic agora and later Roman forum (Fig. 2.3, F).72 A large ashlar masonry structure was discovered in this same area, whose date is roughly contemporaneous with the reservoir. This structure was heavily robbed for building material in antiquity with the result that its identification and function have not yet

2.2.4: Roman Cyprus remained a Ptolemaic possession until 58 BC, when it was annexed by Rome and incorporated into the province of Cilicia. Control of the island was restored to the Ptolemies by Julius Caesar in 47 BC, but reverted back to Roman control upon the death of Cleopatra VII and Mark Antony. 79 With the establishment of the Principate in 27 BC, Cyprus became an independent minor province governed by legati of the emperor. This changed in 22 BC when control was ceded by Augustus to the Senate who governed it through Proconsuls of

69

Stillwell 1961, 77. McFadden posits a date for the theater in the 3rd c. BC based on the similarity of masonry techniques with a tomb in the Agios Ermogenes necropolis dating to the same period (cf. Mcfadden 1946). 70 Swiny 1982, 134; Daniel 1938, 4-6. This was initially identified as “The Palace” by the University of Pennsylvania Mission. It was later renamed the “House of Eustolios.” 71 Rupp 1978, 262 ff. On the basis of the elaborate provisions for the cult of the dead provided by these remains, Rupp postulates that they belonged to an important recently deceased individual from Kourion, perhaps the last king of the city, Pasicrates. If indeed this is a heroön, it would probably belong to the mythical founder of the city, Perseutes, who is known to have been worshipped at the site. The chamber tomb is of indeterminate date. 72 Christou 1983, 269-71. Nine cisterns have been discovered so far. Until this time, the citizens of Kourion relied primarily on rainwater captured in cisterns since the acropolis and sanctuary are devoid of springs and no indications for the sinking of wells has been found. The definitive work on the water supply of Kourion remains Last (1975), although much work has been done at both the sanctuary and city site since his death in 1969. Although no wells have been recovered at the sanctuary or city, Last notes that a small spring north of the acropolis and wells on the coastal plain would have been adequate to supply the settlement during its initial development. However, as the city grew the needs of the population would have quickly outstripped the supply provided by these sources (cf. Last 1975, 53). The date for the initial construction of the aqueduct is uncertain, with a terminus ante quem of the 2nd century AD. The pipes and conduit system discovered by Christou, if dated correctly, may provide evidence that Kourion was furnished with some type of formal water supply during the Hellenistic period.

73

Christou 1983, 267-9; 2001, 41. The remains of this building have not been published and the artifacts its dating is based on have not been subjected to any rigorous study. As a result, the conclusions put forth by the excavator concerning the chronology must be treated as preliminary. The excavator has identified three phases of use: 325-225 BC; 200-125 BC; 125-50 BC. 74 Nicolaou 1978, 533. 75 Soren 1987b, 39. 76 Scranton 1967, 14. 77 Mitford 1971, 144-6 (#77). Watkin 1988, 284. Mitford dates this inscription to the latter half of the 1st c. BC following the Roman conquest. In light of the other civic institutions attested at Kourion during the Hellenistic period, it is probable that the prytaneion was also established at this time. 78 Most notable being the inscribed statue base belonging to the ὁ ἐπὶ τῆς πόλεως Demetrios, son of Machatas discussed previously. 79 Livy Epit. 105; Vell. Pat. 2.45. The Tribune P. Clodius Pulcher carried a law to reduce Cyprus to the condition of a province, ostensibly due to the moral corruption and greed of Ptolemy, the King of Cyprus. In reality it appears to have been done to confiscate the king’s property, which was auctioned to fill the treasury of Rome. The actions of Clodius are generally condemned by the ancient sources, particularly Cicero (Sest. 26.57; 27.59; Dom. 8.20). See also Badain 1965 and Oost 1955.

12

praetorian status.80 The reorganization of the empire by Diocletian in AD 293 saw the end of Cyprus as an independent province, being placed under control of the Praetorian Prefect at Antioch. 81 Instead of an annual Proconsul, the island was now governed by a praeses or ὑπατικός with the rank of clarissimus. This official had a somewhat longer tenure than the Proconsul (between one and two years) and was appointed by the central government in Antioch, at the direction of (or at least with the approval of) the Emperor. 82

brought water ca. 11 kilometers from a source near the village of Sotira (Ypsimasikarka), (Fig. 2.4).87 However, this hypothesis not been archaeologically confirmed. Like the city site, the Sanctuary of Apollo underwent intense building activity during the 1st century AD that refined and upgraded pre-existing structures without altering the ritually fixed division of space. The majority of the work was likely undertaken during the latter half of the century, probably under Nero. This included construction of the second Temple of Apollo, final construction of the round building and its northwest ramp, probable rebuilding of the Northwest Building, construction of the temenos wall around the west enclosure, the initial construction of the palaestra, and possibly the first eastward extension of the archaic altar precinct.88

Based on the available evidence, life in Cyprus under Roman rule continued much as it had under the Ptolemies and its generally peaceful nature led to its mention only rarely in the ancient sources. Fortunately at Kourion, the archaeological record supplements this void, and permits the city’s development to be traced from the inception of Roman rule to its destruction and abandonment during the Arab raids on Cyprus in AD 647.

Despite the flurry of activity under Nero, the apex of building at Kourion occurred during the 2nd century AD, when all the major monuments previously mentioned on the acropolis (theater, nymphaeum, and baths) underwent renovation (in the case of the theater, probable significant renovation) under the emperor Trajan.89 Trajanic benefaction also extended to the Sanctuary of Apollo and included construction of the baths, the Kourion Gate, the

The Roman contribution to the archaeological record at Kourion is substantial. It is somewhat cyclical in nature, alternating between intensive and idle construction periods with the active phases closely connected to imperial benefactions. In the early 1st century AD, the theater underwent some restoration, which may have repaired damage caused by an earthquake known to have severely affected Paphos in AD 15.83 The theater was also the focus of a later renovation attested by an inscription that records the work was undertaken by “Q. Julius Cordus, proconsul under Nero,”84 which dates this work to AD 64/5. The extent of these renovations is disputed, ranging from rebuilding the stage and scene building to transforming the structure from its Hellenistic form to a Roman plan.85 Also in the early 1st century AD, the main bath complex of the city and a nymphaeum were built along the west side of the forum/agora (Fig. 2.3, E). Both structures were substantially altered and enlarged throughout their history, causing the details of their early phases to be largely obscured.86 It is probable that with the establishment of these two structures and the amount of water required for their daily function Kourion was first equipped with an aqueduct, the West Conduit, which 80

Strabo 17.3.25; Dio Cass. 54.4.1. This would later become Vicar of the Orient and after AD 331, Count of the Orient. Cf. Mitford 1980, 1375 ff. 82 The Theodosian Code and Digest of Justinian relate that among his duties, this official was responsible for all departments of administration, law and order as well as being judge of the first instance, and executing the commands of the central government. Unlike his predecessor the Proconsul, he was also responsible for revenue collection, maintenance and public works, and the supervision of the civic government. Cf. Mitford 1980, 1376. 83 Dio Cassius (54.23.7) records that Paphos was devastated by an earthquake, which Augustus alleviated. He also bestowed the title Augusta on the city. See Stillwell 1961, 77 for probable damage to the Kourion theater caused by this event. 84 For inscription see Stillwell 1961, 74 and Mitford 1971, 204-7 (#107). For commentary on Mitford see Bagnall and Drew-Bear 1973b, 230-31. 85 Stillwell (1961, 77) states that only the stage and scene building were renovated, whereas the guidebooks of Swiny (1982, 127-8) and Christou (2001, 22) state that the theater was transformed to a Roman plan at this time. 86 Christou 2001, 46, 50. 81

Figure 2.4: Map showing the water conduits that supplied Kourion. Dotted sections indicate probable routes that have not been confirmed archaeologically. After Last 1975, 40 Map 1. 87 Cf. Last 1975. It is probable that infrastructure associated with the water supply, including settling tanks, cisterns, and a small aqueduct to carry it into the city were also built at this time. 88 Soren 1987b, 39. 89 Stillwell (1961, 77) attributes the theater’s transformation from a Greek to a Roman plan to renovations under Trajan. It is possible that the structure was damaged by an earthquake in AD 76.

13

South Building,90 a remodeling of the stoa area east of the Archaic Precinct, and the final cobbled northwest approach to the Round Building. Following these additions, no further construction seems to have occurred at the sanctuary site except possibly the northwest corner of the Southeast Building, which may have been completed under the Antonines.91

is uncertain, but is believed to be sometime during the 2nd century AD.96 Little to no building activity appears to have taken place at the Sanctuary of Apollo during the course of the 3rd century AD (with the exception of smaller dedicatory monuments). At this time, however, many structures within the city reach their apex of development, most notably the nymphaeum, baths, and theater. The last of these went through two major renovations, the first ca. AD 214-17 during the reign of Caracalla when it was changed into a venue suitable for staging venationes (wild beast hunts); and the second in the latter half of the century, when it reverted back to its original function.97

The second half of the 2nd century AD at Kourion sees building activity inside and outside the city, including a number of new civic structures and improvements that continued its monumentalization. On the acropolis, three large colonnaded stoas formalized the north, west, and south boundaries of the forum/agora. These were remodeled during the late 2nd – early 3rd centuries AD,92 a period that also saw the replacement of the limestone decorative elements of the theater skene in marble and construction of the theater stair tower, most likely under Septimius Severus.93

New construction appears in the northwest part of the site along the main road leading to the city center. The remains belong to two large peristyle houses, although the possibility that they are a single structure cannot be discounted since the area between them remains unexcavated. The westernmost is the House of the Gladiators (Fig. 2.3, C), named for a mosaic in the southeast ambulatory of the central courtyard depicting two pairs of gladiators. East of this is the House of the Apsed Triclinium (Fig. 2.3, D), so-named for a large apsed space paved with elaborate mosaics flanked by three rooms on either side. Adjacent to this suite is a large peristyle courtyard with an Ionic colonnade. The entire complex appears to have been fronted by a public or semi-public paved street, possibly indicating that its nature was not domestic but served some type of public function.98 Construction of both the House of the Gladiators and the House of the Apsed Triclinium has been dated to the latter half of the 3rd century AD.99 While their roughly contemporaneous date may support the theory that they are a single monumental structure, it is equally possible that they are distinct from one another and represent a deliberate development (or redevelopment) of the northwestern acropolis undertaken at this time.

Outside the city, a monumental stadium measuring 229 X 24 meters was built between the acropolis and the Sanctuary of Apollo. This structure has been dated to the Antonine period (Fig. 2.1), although the site was probably utilized for athletic competitions prior to this formalization. Its footprint partially overlay the line of the West Conduit, which necessitated its relocation on a line south of and parallel to the stadium’s exterior wall. Disruption of the water supply for this construction, possibly for an extended period of time, may have necessitated the completion of a second water conduit to provide for the city. This line, known as the East Conduit (Fig. 2.4), originated from a spring near the village of Souni and followed a circuitous ca. 22 kilometer route before entering the city via an aqueduct 60 meters long with a maximum height of 5 m.94 Dating the construction of this aqueduct is crucial, as its foundations rest upon a set of walls belonging to the House of Achilles (Fig. 2.3, B) immediately outside the western city wall. Despite the name given to this structure, its function remains uncertain. It has not been fully excavated and additional remains were likely destroyed as a result of constructing the Limassol-Paphos road. Although possibly a domestic structure, its location at one of the principal entrances to the city combined with its opulent decoration have led some scholars to interpret it as being a public structure, possibly an apantitirion (reception hall) where officials or distinguished guests would be met before entering the city.95 Like its function, the date of the House of Achilles

Based on the archaeological and epigraphical evidence, Kourion was a thriving and prosperous city at the outset of the 4th century AD. By the end of this century, however, it would be an abandoned ruin, having been destroyed by a massive earthquake; one in a series of seismic events that affected Cyprus during this period. Cyprus has always been vulnerable to earthquakes because of its proximity to the convergence of the African and Anatolian plates. A number of these caused widespread damage severe enough to have warranted recording in the works of ancient authors. Like most of the major cities of ancient Cyprus, Kourion shows evidence of having been affected by seismic events

90

Mitford 1971, 207-11 (#108). The inscription states that Trajan constructed the two remaining exedrae for Apollo Caesar and Apollo Hylates under the supervision of the proconsul Q. Laberius Justus Cocceius Lepidus, who dedicated it. The probable date for the dedication is AD 101. However, as the South Building comprises five of these spaces, it seems that the structure was begun at an earlier time, and finished by Trajan. 91 Soren 1987b, 39. This may originally have been part of the Trajanic building program. 92 Christou 2001, 43-4; See also Christou 1983, 271-2 for initial excavation report. 93 Stillwell 1961, 77-8. 94 Last 1975, 48. 95 Swiny 1982, 100; Christou 2001, 58.

96 Mitford (1971, 360-1) attributes the mosaics to the late 3rd – early 4th c. AD, but these likely represent a renovation carried out on the structure at this time. 97 Stillwell 1961, 78. The dating of conversion back to a theater is based on numismatic evidence recovered under the restored seating areas. 98 Christou 2001, 54. 99 For the House of the Gladiators see Loulloupis 1971.

14

throughout its history. In most cases, however, they were not catastrophic, with a sufficient amount of time between events to permit damage to be repaired.100 This situation changes during the 4th century AD, when Cyprus, and much of the eastern Mediterranean, became the focus of earthquake “storms;” a series of rapidly occurring major earthquakes caused by the redistribution of stress on a fault, or system of faults in a particular region.101 Two major earthquakes at Salamis are recorded by the sources, the first in AD 332/3 and then in AD 342. These events were so devastating that following the second one, taxation was remitted for a period of four years and the city was re-founded on a smaller scale as the new capital under the name Constantia.102 While it is probable that these earthquakes also caused damage at Kourion, they were not responsible for its final destruction. Based on the coins recovered from his excavations on the acropolis, David Soren proposed that the Kourion earthquake is the one recorded by Ammianus Marcellinus, as occurring just after dawn on July 21, AD 365.103 Despite its initial acceptance, the hypothesis of Soren has been convincingly challenged and largely invalidated by recent archaeological and archaeoseismological research and discoveries. The former consists of arguments against Soren’s dating of the Valens “Split-S” type based on the lag time between the minting of new issues and their appearance in general circulation on the island.104 Combined with this, the archaeoseismologic data strongly supports locating the AD 365 event off the west coast of Crete, which destroyed the sites of Gortyn, Eleutherna, and Kisamos.105 Given the magnitude of this event, it is possible that it was felt at Kourion. However, considering its distance from the epicenter (ca. 860 kilometers), the damage inflicted on the city as a result of the earthquake itself was minimal, if any. The widespread damage recorded by Ammianus as having affected the entire eastern Mediterranean should rather be attributed to the tsunami generated by the event. This would have had far greater consequences for distant cities than the earthquake, particularly since many were located in low lying coastal areas.106 In the case of Kourion, since the city does not appear to have been equipped with a commercial harbor and the main settlement was high above the coastal plain, damage from the tsunami was also most likely minimal, especially when compared to sites such as Paphos and Alexandria.107 Based on the

evidence, it can only be proposed that the event that destroyed Kourion occurred sometime during the later 4th century AD, possibly ca. AD 370-380. Although no “smoking gun” exists in the archaeological or literary record that provides a definitive date for the event, all indications are that the destruction brought upon Kourion was sudden and complete. It caused massive architectural damage to all areas of the site as well as the deaths of a substantial number of inhabitants. In the aftermath, the site appears to have been largely abandoned for an unknown period of time before eventually being reoccupied during the early 5th century AD. 2.2.5: Early Christian – Medieval From the moment rebuilding began, Kourion was conceived of as a Christian city. This is demonstrated by the fact that only the acropolis was restored and occupied. In contrast, the Sanctuary of Apollo, once a major center of worship, was left in ruins and used periodically as “squatter” habitation from the late 4th – 6th centuries AD.108 On the acropolis, none of the structures destroyed by the earthquake were completely rebuilt and those that were rehabilitated did not retain their original function. The best illustration of this is the baths and nymphaeum on the western edge of the forum, public structures whose larger spaces were subdivided and utilized primarily as living quarters.109 Construction following the earthquake utilized materials salvaged from ruined buildings throughout the site, including the Sanctuary of Apollo. On the southeast side of the forum/agora, the main Christian basilica (Fig. 2.3, G), was oriented north-south so that its walls could utilize the foundations of the largely destroyed Roman civic basilica.110 The complex of buildings that formed the restored civic core was centered on the large basilica church with an apsed central nave and two flanking side aisles. This imposing structure was the seat of the Bishop of Kourion and appointed with mosaic and opus sectile flooring, marble wall revetment, and mosaic wall decoration. It was surrounded by a number of ancillary facilities including a large and ornate baptistery, a diakonikon, an atrium and courtyard, and a Bishop’s residence. Together, these elements form one of the earliest examples of an Early Christian cathedral precinct in Cyprus.111

100 An earthquake in AD 76 may have necessitated the Trajanic renovations of the theater, which appears to have been previously repaired after the AD 15 earthquake (cf. Christou 2001, 22). 101 Nur and Burgess 2008, 240-3. 102 Theophranes Chronographia PG 108, Col. 123; PG 108 Col. 133. Sozom. Hist. Eccl. 67.3.6. 103 Amm. Marc. 26.10.15-19. Soren and Davis 1985, 299; Molinari et al. 1988, 171; Soren 1987b, 43; Soren and James 1988, 86-7; Kelletat 1998; Jensen 1985. 104 Lichocka and Meyza 2001 185-94. 105 See Lichocka and Meyza 2001,185-94; Guidoboni 1994, 272-3; Stiros 2001, 558-9; Fokaefs and Papadopoulos 2007, 520-1. Kelly (2004 141-2) provides a good overview of the evidence. 106 Kelly 2004, 145. 107 Leonard (2005, 556-65) represents the most recent work on the harbor of Kourion. Evidence for any type of harbor installations is minimal, consisting mainly of the remains of a possible mole or jetty.

A second major post-earthquake structure dating to the late 4th – early 5th centuries AD is located on the eastern end of the acropolis northeast of the theater (Fig. 2.3, J). Any structures damaged by the AD 365 tsunami would most likely have been commercial in nature. 108 Soren (1987b, 42, 52) records what he interpreted as postearthquake “squatter” occupation in the Southeast Building, baths, and possibly the West Complex. 109 Christou 2001, 34-9. Post-earthquake structures were also found in the forum area, where the southeast part of the stoa was remodeled, and at the House of the Gladiators (cf. Loulloupis 1971). 110 Megaw 2007, 345. 111 Megaw 1985, 295

15

A mosaic inscription discovered during excavation records that it was built by Eustolios to help alleviate suffering endured by the inhabitants of Kourion, presumably in the aftermath of the earthquake.112 The plan of the structure is that of a large, opulent peristyle house with a small private bath complex, which led to it being named the “House of Eustolios.” Based on the inscriptional evidence, however, the structure served a more public function and was presumably intended to act as a (temporary) replacement for the ruined public structures around the forum/agora. The complex underwent a late renovation during or immediately following the reign of Leo I (AD 457-74) and remained in use, likely as a public guest house, until the Arab raids in the 7th century AD.113

likely functioned as a complement to the main basilica and although services would have been presided over by the Bishop of Kourion, it was probably only utilized on specific religious holidays celebrating an as yet unidentified patron saint.114 A second basilica church, similar in size and plan to the one At-Meydan, was built at the base of the Kourion acropolis sometime during the 6th century AD (Fig. 2.5). It is located in an area believed to have been near the as yet undiscovered harbor of the city and may have been intended to serve the religious needs of visiting foreigners, although this hypothesis is yet to be confirmed.115 Based on the available evidence, the picture of Kourion that emerges following its reconstruction and expansion during the 5th – 6th centuries AD is one of a prosperous, moderately-sized settlement that in a short period of time had quite literally brought itself back from the dead. It is this prosperity as well as its coastal location that made it a prime target for attack by Arab fleets from Tyre and Alexandria, which resulted in the final destruction of Kourion. The surviving population relocated to Episkopi, east and inland of the ancient city, which was less susceptible to raiders and, unlike the acropolis, had reliable access to an adequate fresh water supply.116 Following its abandonment, the ruins of Kourion continued to collapse, and although the memory of the city began to fade it was not completely forgotten. The

Constructions of the later 5th-6th centuries AD located outside the acropolis attest not only to the increasing influence of Christianity but also the prosperity of the revived Kourion. Between the city and the remains of the stadium, a three aisled basilica church was built at a site known as “At-Meydan” (Fig. 2.1). It was fronted by a paved courtyard with a large circular cistern in the center. The ancillary structures of the basilica, consisting of a small chapel, a vestibule that forms a second entrance, and a number of smaller spaces (most probably storerooms) along the north side of the basilica completed the complex. Given its size and location, this church

Figure 2.5: Kourion beach basilica. 114 Swiny 1982, 84. This structure and its associated finds remain unpublished. 115 Use of this church by visiting foreigners, most probably merchants, is largely dependent upon the Kourion harbor being commercial in nature. This has not been substantiated. This structure and its associated finds remain unpublished except for a brief description in Christou 2001, 74-5. 116 Swiny 1982, 154.

112 Mitford 1971, 356-8 (# 204). Mitford’s restorations for many of the mosaic inscriptions from this structure are controversial (cf. Bagnall and Drew-Bear 1973b, 237-43). 113 Swiny 1982, 134. The remains of the structure are largely unpublished, with the result that the dating of its initial construction and architectural development has not been firmly established.

16

presence of numerous circular kilns attest to the periodic use of the city’s monuments as raw materials for lime production from the 8th century through the Medieval period.117 After this, the ruins appear to have remained virtually undisturbed until the 19th century when they attracted the attention of antiquarians who began exploring the site intent on recovering its antiquities.

“hundreds of small mounds” that he believed marked locations of domestic structures, with larger mounds being “the debris of public buildings or palaces.”123 While investigating one of the larger mounds “where eight shafts of columns of a brownish granite lay imbedded in the ground,” Cesnola supposedly discovered a rock cut gallery, off of which opened four apsidal chambers. These were filled with objects in gold, silver, bronze, and iron, which he immediately designated “The Treasure of Kourion.”124 Initially, news of this discovery provided Cesnola the notoriety within the European archaeological community he desired, but subsequent scrutiny of the objects and account of their discovery revealed that the entire episode was a fabrication. In reality, the artifacts had been purchased and collected from native diggers throughout the island and the rock cut chambers on the acropolis where they were supposedly found did not exist at all. The ruse was created by Cesnola so that his discovery would rival the “Treasure of Priam” recovered by Schliemann at Troy, insuring that the collection could be sold promptly and thus secure his financial future.125

2.3: History of Excavation The first appearance of Kourion in the history of archaeology occurs in 1826 when Carlo Vidua noted in his work Inscriptiones antiquae that one of the inscriptions in his corpus was copied “among the ruins of the ancient hill of Curium.”118 A little over a decade later in 1839, the American missionary Lorenzo Warriner Pease recorded that during his travels he visited a place referred to by local people as “Apollonas” (of Apollo). The account of Pease provides the first record of a visitor to the sanctuary in the modern period and demonstrates the value of information preserved through local traditions. In this case the knowledge that this area was associated with Apollo and possibly the site of the extra urban sanctuary recorded by the ancient authors.119 In 1845, the first scholarly account of Kourion was written by the German archaeologist Ludwig Ross. He described the ruins of the sanctuary, recording a number of features including column drums, capitals, and statue bases with inscriptions that were visible on the surface.120

Following Cesnola, in 1895 the ruins of Kourion were investigated by H. B. Walters on behalf of the British Museum. His efforts were concentrated primarily on the necropoleis, with some limited excavation northwest of the city outside the fortification wall. It was during the latter that the dedicatory inscription to Demeter and Kore was recovered, leading to the hypothesis that a sanctuary of these two divinities was located in this area. This supposition was eventually disproved by excavations carried out from 1971-74 by A. Christodoulou of the CDOA. This work was undertaken in the hopes of locating this shrine but instead determined that the inscribed block was spolia utilized in constructing the small extra-mural basilica “At-Meydan.”126

It is not until the arrival of Luigi Palma di Cesnola, the American Consul to Cyprus in 1865 and his subsequent archaeological investigations that the excavation history of Kourion begins.121 Born to a noble Italian family, Cesnola immigrated to the United States. He served as a high ranking officer in the Union army during the Civil War, being appointed afterward to the consular position in Cyprus. Soon after his arrival, Cesnola became obsessed with augmenting his salary through commercial endeavors, the most lucrative of which was the collection and sale of antiquities found throughout the island. This pursuit eventually led him to the ruins of Kourion where he identified the remains of extensive necropoleis on the slopes and plain east of the acropolis, the three city gates, the theater, the stadium, the aqueduct, and the Sanctuary of Apollo Hylates.122 As recorded in his work Cyprus: Its Ancient Cities, Tombs, and Temples, Cesnola initially focused his excavations at Kourion on the unlooted tombs of a large necropolis surrounding the small chapel of Agios Ermogenis. Expanding onto the acropolis, he noted architectural elements protruding above the ground and

The “modern” archaeological era at Kourion did not occur until 1934 with a project sponsored by the University Museum of the University of Pennsylvania. Its purpose was to “trace the history of that ancient city in the different sites it occupied from the earliest times to the breakup of the Roman Empire.”127 While the project was technically under the direction of Bert Hodge Hill, the actual excavation was carried out by George McFadden, a vocational archaeologist and the expedition’s financial backer, and his assistant John Franklin Daniel.128 In order to achieve their goal of tracing the history of Kourion from beginning to end, the 123

Cesnola 1877, 300. The initial room explored by Cesnola was Room C (Cesnola 1877, 308). This held only gold objects and came to be known as the “Gold Room.” Room D (Cesnola 1877, 325) contained over 300 objects in silver and silver gilt. Room E contained objects of bronze, stone, and terracotta. Room F (Cesnola 1877, 334) was smaller than the others, and contained objects in bronze, copper, and iron. 125 Marangou 2000, 263. See also McFadden 1971. 126 The remains of this structure and its associated finds remain largely unpublished with the exception of the Swiny (1982, 80-5) and Christou (2001, 63-6) guidebooks. See also Nicolaou 1973a, 58; 1973b, 431; 1975, 132; 1976, 371-2 for yearly summaries of excavations. 127 McFadden 1938b, 3. 128 Dyson 1998, 236-7.

117

124

Christou 2001, 47. To the knowledge of the author, no evidence of lime kilns has been found at the Sanctuary of Apollo. It is possible that material could have been salvaged and taken to the city for processing. 118 Vidua di Conzano (1826, 36) records that Pl. 31, no. 4 had been copied "inter ruinas vetusti (sic) oppidi Curii." 119 See Buitron and Soren 1982, 59. 120 Ross 1852, 84. For a translation see Gunnis 1936, 231. 121 Mitford 1971, 3. Work carried out prior to Cesnola followed the antiquarian epigraphical tradition. Inscriptions recovered from the sanctuary site were recorded in 1845 by Ludwig Ross, in 1849 by Lt. Leycester, and 1862 by William Henry Waddington. 122 Cesnola 1877: Necropoleis: 294-7; Gates: 299; Theater: 299; Stadium: 337-8; Aqueduct: 341; Sanctuary of Apollo: 342-3.

17

team utilized total excavation methodology. An established practice in Near Eastern archaeology, this approach was based on the premise that an a site could only be fully understood by exposing it completely.129 This “big dig” mentality developed when archaeology was still largely in the hands of antiquarians, whose primary interest was recovering objects for museum collections. As a result, the premise of total excavation was based not so much on a desire to understand the historical and archaeological development of a site, but to insure that no artistically significant objects had been missed. Consequently, very often archaeological contexts were only cursorily described, if at all, with objects sometimes re-contextualized according to what the excavator believed they should be, rather than what their findspot indicated.

With the death of McFadden, the University of Pennsylvania abandoned the excavations at Kourion despite the fact that many structures remained only partially excavated. The most notable was the main basilica, discovered by Daniel’s first trial trench in 1934 and only cleared to its eastern end when work ceased. Completion and publication of this excavation was taken over by A.H.S (“Peter”) Megaw, who at that time was the first director of the CDOA. Work carried out from 195658 established the limits and plan of the basilica and cleared the entire structure except for the narthex. As the project progressed, Megaw noted a number of wellpreserved ancillary structures belonging to the ecclesiastical precinct but was unable to investigate them due to the political changeover in 1960. From 1974-79, Megaw returned to Kourion under the aegis of the Dumbarton Oaks Center for Byzantine Studies. These campaigns succeeded in exposing the remainder of the complex that formed the core around which the postearthquake settlement was oriented. 132

At Kourion, the University Museum excavations continued until 1954, although suspended from 1942-46 due to World War II. Work at the site was divided, with McFadden primarily at the sanctuary site where he excavated nearly all the major architectural features of the ancient complex. Daniel concentrated on the city site, where he discovered and excavated a number of structures including the theater, the main basilica, the House of Achilles, and the House of Eustolios, which he initially called “The Palace.” Ancillary excavations were undertaken by the Pennsylvania Mission at the stadium, the Bronze Age site of Bamboula, and the Kaloriziki and Agios Ermogenes necropoleis. The last two yielded a number of unplundered tombs ranging in date from the Cypro-Archaic to the Late Roman periods. 130

Following the excavations of Megaw, work at Kourion continued in the form of “clean up” projects and restoration of previously discovered monuments. In 1961 the CDOA with the assistance of John Travlos began restoring the cavea of the theater, which by 1963 had been completed up to the level of the diazoma. This same year, the CDOA resumed work on the stadium, which, like the basilica, had been partially cleared but not completely defined at the time of McFadden’s death. Completed in 1966, these works succeeded in exposing and consolidating the walls along the entire length of the structure as well as reconstructing one section of seating on its southwest side. 133

While short reports of preliminary findings were published, the work of the University Museum Mission at Kourion has ultimately suffered for two reasons. The first, and perhaps most significant, is the untimely deaths of the principal excavators (J. F. Daniel in 1948; George McFadden in 1953) before a final systematic publication of their work could be completed. Following McFadden’s death, fragments of the site were apportioned out to individuals associated with the project, who completed any necessary excavation before producing a final publication.131 The reports from these efforts are inconsistent and in many cases suffered due to the second problem – poor recording practices. In many cases the excavation notebooks do not document information such as basic stratigraphic relationships or contexts from which artifacts were recovered. Despite its shortcomings, however, the work carried out by the University Museum Mission provided numerous observations regarding the archaeological history of the site, perhaps the most significant being its apparent destruction by an earthquake during the later 4th century AD.

With the “loose ends” of the University Museum excavations tied up, the CDOA initiated work in unexplored areas of the city. The first began in 1967 under the direction of M. Loulloupis, who focused on the area southeast of the House of Achilles immediately inside the fortification wall. In a series of campaigns that lasted until 1974, Loulloupis uncovered the remains of the House of the Gladiators and the House of the Apsed Triclinium; elite structures that preserved sophisticated architecture, marble wall revetment, and elaborate mosaic pavements.134 It was at this time that the first synthetic work on Kourion was produced in the form of T.B. Mitford’s The Inscriptions of Kourion. This volume provided a comprehensive examination of the epigraphic corpus from the sanctuary and city sites, which range in date from the 7th century BC – 6th century AD.135 132 For basilica excavation reports see Megaw 1951, 1953, 1956; 1976, 1979, 1985. Final publication Megaw 2007. 133 For reconstruction of the Kourion theater by the CDOA see Dikaios 1961-62, 43; Karageorghis 1962, 414; 1963, 386; 1964, 378; 1966, 389. For reconstruction of the stadium by the CDOA see Karageorghis 1966, 386; 1970, 300. 134 The architectural remains and sculpture from the structure are published in Loulloupis 1973. A very small number of the smaller portable finds are mentioned, but described only cursorily. 135 The corpus is predominantly composed of objects recovered during the University Museum Mission, but includes material from the work of Cesnola and Walters as well.

129

Davis 1989, 165. For reports on the University Museum Mission excavations see especially McFadden 1935, 1938a, 1938b, 1940, 1954; Daniel 1937, 1938; Fales 1950. 131 Theater: Stillwell 1961; Sanctuary: Scranton 1967; Water Supply: Last 1975. 130

18

Although largely superseded by the critique and commentary of Bagnall and Drew-Bear, Mitford’s work remains a landmark attempt to collect and study a significant component of the material record produced throughout Kourion’s long history.

Overlapping slightly with the excavations of Christou, which concluded in 1998, were the investigations of the Amathus Gate cemetery beginning in 1995 directed by Danielle Parks. Over five field seasons, Parks and her team examined the layout and organization of the cemetery in order to define its relationship with the city. The project distinguished three phases of use; the first characterized by large chamber tombs dating from the Hellenistic and Roman periods that remain in-use until at least the early 3rd century AD. In its second phase the cemetery appears to have been used as a quarry for building stone, probably for rehabilitation and/or new building projects undertaken following the destruction of the city. After this, the area reverted to a necropolis utilizing cist graves until its final abandonment during the 6th century AD.139

In 1975, the CDOA began a series of campaigns under the direction of Demos Christou that exposed the heart of the ancient city including the forum/agora, the baths and nymphaeum, post-earthquake domestic structures, and elements of the city water supply. Three years later in 1978, work resumed at the Sanctuary of Apollo with a joint project sponsored by the Walters Art Gallery and the University of Missouri co-directed by Diana BuitronOliver and David Soren. Its purpose was to continue and expand the University Museum excavations and explore questions concerning the chronology and architectural development of the complex. The project split in 1980, with Buitron-Oliver focusing on the archaic precinct of the sanctuary while Soren concentrated on the Temple of Apollo and the area to its southwest in the hopes of better understanding their expansion and development. Soren successfully clarified questions concerning the sanctuary’s chronological development and discovered a new structure, the “Round Building,” which he believed was the location of a sacred grove where rituals connected to Apollo Hylates were conducted. 136 During the 1980 field season Soren incorporated a small survey project around the base of the Kourion acropolis and the neighboring hill of Yerokarka, which identified a number of previously unknown features.137 This appears to have been the first systematic survey ever done of the area and represents an initial attempt, although on a small scale, to contextualize the city within the wider chora it controlled.

It was not until 1997 that the first large-scale pedestrian survey of the Kourion chora was undertaken by Stuart Swiny as part of the Sotira Archaeological Project (SAP). The goal of the SAP survey was to increase understanding of how the land around the city was organized and contributed to its economic development. While the borders of the chora controlled by Kourion are a matter of some debate, it is generally believed to have been bounded on the north and south by the Troodos massif and sea respectively, on the west by the gorge of the Pharkania river west of Paramali, and to the east as far as Neapolis (modern Limassol). 140 Of this territory, the 1997 survey covered ca. 154 km2, from the Kouris River Valley in the east to Paramali in the west. Perhaps the most significant finding to emerge from this project is the apparent lack of development in the Kourion hinterland during the early periods of its history. It is only during the Roman and especially the Late Roman period that exploitation of the countryside reaches its apex. This was followed by a period of decline that sets in after the mid7th century AD.141

In 1984, Soren divided his team and began excavating on the acropolis in the hopes of confirming his hypothesis that the July 21, AD 365 event recorded by Ammianus Marcellinus was the earthquake that destroyed the city. The remains and their excavation unearthed over four field seasons are the focus of the present work and will be outlined in the following section (Ch. 2.4).

Following completion of the Amathous Gate Cemetery project, Parks began the Kourion Mapping Project from 2004-06. The goal of this project was to utilize both GIS and GPS technology to create a site master plan that included the physical topography and all excavated remains on the acropolis. The plan would provide an accurate base reference for future scholars working at Kourion, allowing the urban fabric of the city as it exists to be visualized and analyzed much more readily than is presently possible. Its electronic format will, in theory, permit the findings of subsequent excavations to be added

Buitron and Soren concluded their investigations at the Sanctuary of Apollo in 1985. With the exception of reconstructing and consolidating previously discovered monuments, no further excavation has occurred at the sanctuary with subsequent work focusing on the acropolis and its environs. In conjunction with his ongoing excavations in the forum/ agora of the city, from 1989-90 Demos Christou investigated the remains of a large monumental tomb constructed during the Cypro-Archaic II period (500-475 BC) in the area of the Agios Ermogenes necropolis. In 1993, Christou also began exposing the Christian basilica located on the beach below the south face of the acropolis.138

during this time. The tomb and the basilica as well as their associated artifact assemblages remain almost completely unpublished. 139 See Parks 1996; 1997; Parks et al. 1998; 1999; and 2000 for preliminary reports of these excavations. Final publication in progress. 140 Cf. Mitford 1966, 94-5; 1971, 1; Bagnall and Drew-Bear 1973a, 100-2. 141 Swiny and Mavromatis 2000, 438. Of the 22 sites registered, the majority appear to have been small farmsteads. Based on the ceramic evidence they were active until the time of the Arab raids (ca. AD 649).

136

See Soren 1987b. For the results see Leonard 1987 in Soren 1987a. 138 Christou 2001, 75-8; Herscher 1995, 277-8. See also Christou 1994 and 2001 for reports and results of work in the forum carried out 137

19

quickly, insuring that the plan can be frequently and perpetually updated.142

uncovered a number of walls, which he believed belonged to three distinct structures: House 1, House 2, and House 3. Among the walls he discovered a significant number of objects including complete or nearly complete pots, bronze vessels, coins, and lamps that appeared to have been preserved in situ from the time the structures were destroyed.

Summer 2012 saw the inaugural field season of the Kourion Urban Space Project (KUSP), the most recent work undertaken on the acropolis. One of the primary questions that KUSP seeks to answer is what the urban context was in the part of the site where the Earthquake House and Building B are located. The project will also investigate to what extent the Earthquake House and its assemblage are typical of domestic structures of 4th century AD Kourion, whether it is an isolated domestic structure or part of a “neighborhood,” and whether or not the cultural change on Cyprus from paganism to Christianity can be identified in the material record.

On June 5, 1934 Daniel’s notebook records the discovery of human remains apparently resting upon a floor in House 1. This resulted in widening the trench to ca. 4 m between meters 28-36 to facilitate their excavation.145 Within this expanded area, designated Room 1, Daniel uncovered two skeletons in the center of the room surrounded by a number of largely intact objects consistent with those found elsewhere in the trench (Fig. 2.7). The position of the victims and the nature of the objects surrounding them led Daniel to conclude that they did not belong to a burial, but represented individuals and objects in the space when the structure collapsed. Since he observed no evidence of fire in the structure, Daniel believed that an earthquake was most likely responsible for the destruction.146

The results from the initial season of excavation have produced data that has altered previous hypotheses about the size, layout, and development of the ancient city as it existed during the later 4th – 6th centuries AD. In the future, KUSP will also investigate areas further afield of the Earthquake House through geophysical prospection in order to reveal the overall plan of Kourion without excavation. The potential of the Kourion Urban Space Project for providing comparative evidence to the Earthquake House and elucidating daily life at Kourion during the 4th century AD is enormous, and will contribute significantly to our understanding of life at the site during the Late Roman period. 2.4: The Earthquake Complex Excavation Prior to the University Museum mission excavations in 1934, the Kourion acropolis had not been extensively explored. This is likely due to the fact that few of its monuments were visible on the surface. As a result, the initial investigations by J.F. Daniel consisted of a series of trenches along the southwestern and southeastern edges of the site, basically “prospecting” for the remains of the ancient city. In a shallow hollow on the east end of the acropolis, along the main road from Amathous ca. 200 m west of the theater, Daniel placed a trench where he observed large architectural blocks protruding above the ground surface (Fig. 2.3, H). This excavation, known as Trench III, is recorded by Daniel in his field notebook on May 29, 1934 as measuring 24 m long, 1.5 m wide, and oriented roughly north-south.143 Two days later he extended the trench an additional 12 m south (a total length of 36 m) and widened it to the west between meters 1-8 and 12-16. This created two small triangular shaped extensions that are recorded as being “about 3 meters deep” (Fig. 2.6).144 During this excavation, Daniel

Figure 2.6: Plan of Trench III showing architectural remains uncovered during excavation. 145 The widening of Trench III in this area is not recorded in the field notebook with the exception of sketches, which are generally not to scale. It appears that the trench was widened twice, one to the east approximately 1 m wide (2.5 m total), and another to the west approximately 1.5 m wide (4.0 m total). A third extension to the west of ca. 2 m to include the southwest corner of the room created a maximum width for Trench III of 6 m. The June 8, 1935 notebook entry (p. 31) records “Clear away dump at south end in order to extend excavation of house no. 1 to east and west.” The drawings preserved in the notebook appear somewhat contradictory and do not record the process as to how Trench III was enlarged. See DFN: 1934, pp. 21, 25, 27, and 38 for relevant plans. 146 DFN: 6/5/1934, p. 25. “Find clay pot, bronze pitcher, broken lamp, and part of another pot directly on the floor. Next to these,also on the floor, a human skeleton. This was not a burial. The person seems to have been standing when the house fell, and crushed where he stood. As there are no signs of fire an earthquake may have caused the disaster.”

142

Sadly, Danielle Parks passed away shortly before the completion of the final mapping season. Prior to her passing, Dr. Parks arranged for Tom Davis, Director of the Cyprus American Archaeological Research Institute to take over direction of the project. See Buell et al. 2009 for publication. 143 DFN: 1934, p. 6. 144 DFN: 5/31/1934, p. 12. A plan on p. 16 of the notebook indicates the expansion of Trench III occurred between meters 12 and 20. No account of a second widening of this area of the trench, or for the discrepancy between the plan and the written record is made in the notes.

20

Despite the discovery of an apparently undisturbed earthquake debris deposit, Trench III was abandoned in favor of monumental structures such as the basilica, theater, and “palace,” all of which could potentially provide objects for the University Museum collection.147 A brief account of the Trench III excavation was published in the Museum Bulletin of the University of Pennsylvania in 1935 and McFadden mentioned it in a 1940 article regarding his conclusions about earthquake collapse at the Sanctuary of Apollo.148 Following a brief “clean up” of Room 1 that occurred during the winter of 1935, Trench III was largely backfilled and quickly faded into obscurity, especially following the untimely deaths of the two principal investigators. The area of Trench III remained unexplored until 1984 when it was reopened by David Soren in the hope it would provide evidence supporting his theory about the earthquake that destroyed the site. With the help of Roger Edwards of the University Museum, Soren obtained Daniel’s field notebooks along with a photograph of the Room 1 victims preserved in the museum archive. Although the notebooks did not contain the precise location of Trench III, a plan produced by J.S. Last (the architect of George McFadden) and published in T.B. Mitford’s Inscriptions of Kourion recorded the locations of all of the trenches and structures excavated by the University of Pennsylvania mission. This enabled the trench, whose remains were still visible on the surface, to be located and reopened by Soren and his team from the University of Arizona on the 50th anniversary to the day of Daniel’s excavation.149

Figure 2.7: Victims recovered in House 1, Room 1 by J.F. Daniel during the Trench III excavation. Courtesy of the Penn Museum, image # 100369. encompassed ca. 673.58 m2 150 and revealed the remains of two structures: Building A known generally as the “Earthquake House;” and Building B, the so called “Market” building. These were separated by what has been interpreted as a small street or alleyway (Fig. 2.8). Of these two structures, the present study is concerned only with the Earthquake House and the artifact assemblage recovered during the course of its excavation. This is because unlike Building B, the Earthquake House was excavated fully, and thus the data set recovered from it is virtually complete. To this day, Building B remains only partially excavated with its full plan and assemblage awaiting recovery and study.

Initially, work focused on the area of Daniel’s House 1, Room 1 in an attempt to locate the earthquake debris recorded in his notes. Although the walls of the room had been previously exposed, the fill along its southern side had not been cleared in 1934. When excavated, this area produced a number of finds that justified expanding the scope of the investigation. The trench was extended westward, which quickly revealed the presence of two additional rooms of the structure as well as a third human casualty of the disaster. Based on the results obtained during this initial field season, the following year Soren accelerated the pace of excavation continuing west before shifting south and east. This resulted in the discovery of four additional rooms of the structure and one additional victim. The years 1986 1987 saw the primary area under investigation shift north of the initial 1984 excavation. This revealed that the features identified by Daniel as “House 1” and “House 2” in Trench III belonged to a single structure, with “House 3” forming part of a separate building located further to the north. At the close of the 1987 field season, the area excavated by the University of Arizona mission

147

This was one of the primary goals of the University Museum Mission from its inception. 148 See McFadden 1935; McFadden 1940. 149 Soren & James 1988. 80-3.

150

21

1984: 42.38 m2; 1985: 196.92 m2; 1986: 289.06 m2; 145.22 m2.

Figure 2.8: Earthquake complex state plan. After Buell et al. 2009, fig. 1.

22

Chapter 3: Methodology objects exist.153 Horizontal control over the length of Trench III was achieved by setting stakes along both sides of the cut every two meters. Each was labeled using consecutive letters of the Greek alphabet, with A (alpha) designating the starting point.154 The width of the trench is rarely indicated on sketch plans in the field notebook except where it deviated from the standard 1.5 m. In these instances it is normally recorded in half meter increments using numerical values along the top of the plan (Fig. 3.1).155

As discussed previously, the architecture and artifact assemblage of the Earthquake House were recovered during two distinct excavations: the University of Pennsylvania Museum Mission from 1934-35; and the University of Arizona Mission from 1984-87. During the fifty year interval between the two, the discipline of archaeology developed and refined the techniques for excavating, recovering, and recording objects in the field and laboratory. This resulted in a significantly more sophisticated methodology being implemented by the latter project. The disparate nature of these approaches to the archaeological process as well as their execution had differing impacts on the artifact assemblage, which need to be recognized and accounted for to mitigate any effects they may have on the present study. This chapter examines the methodologies employed by these earlier investigations and their potential impacts prior to outlining the procedures used to collect and interpret the data this work is founded upon.

Artifact findspots were recorded in the field notebook in relation to the closest trench stake, or in situations where a cluster of objects was encountered, the stakes it was located between. No measurements attempting to triangulate more precise locations along the horizontal axis appear to have been made. The vertical position at which objects were found was recorded in relation to their depth below the modern ground surface. However, no record of the starting elevations from which these measurements were taken appears to have been made. Without these starting elevations, reconstructing the vertical context for the Trench III artifacts cannot be done with any degree of confidence.

3.1: The University Museum Mission Excavations: 1934-35 While considered primitive by today’s standards, the excavation methodology of J.F. Daniel was acceptable for the time, being based on established practices developed at sites throughout the Mediterranean and Near East. Because one of the main goals of the University of Pennsylvania project was to recover artistically significant objects for the University Museum collections, its focus was on locating and excavating public spaces where such works were likely to be found. To expose these areas quickly, Bert Hodge Hill, the titular director of the University Museum Mission at Kourion, employed a methodology known as “trench and peel.” This technique was developed by German teams at Olympia in the latter part of the 19th century and had been used by Hill on projects in Athens and Corinth.151 The process began by excavating a trench, normally long and narrow, to reveal subsurface features. At these points, the edges of the trench would be “peeled” back to expose the remains below as fully as possible.152 In the case of Trench III, the remains did not merit a great deal of “peeling” except on the southern end where it was widened from 1.5 to 6 m so that the human remains in this area could be examined.

The imprecise nature of Daniel’s recording is compounded by the fact that his plans were general, stylized, rough sketches. They primarily depict the architecture and were not drawn to scale. This lack of detail also affects the single stratigraphic profile recorded for Trench III, which represents only five of the 72 m face exposed by both baulks.156 Like the plans, the section drawing is rough, using stake designations to fix its 153 This problem is not exclusive to Trench III, as David Rupp encountered a similar situation attempting to examine material from the House of Eustolios (Personal Communication). 154 This system is employed inconsistently by Daniel throughout the rough plans drawn in his notebook. Some plans utilize only letters, others only the meter measurements. A few plans utilize both. For Trench III, it appears that the letter I (iota) was omitted from the sequence. 155 A single exception to this Trench III recording methodology is found (DFN: 6/14/1934, p.38). This plan shows the widening of Trench III for the excavation of Room 1, House 1, from stakes O – Υ (meters 28-38). Along the top of the plan are a series of five divisions, the center of which is labeled “O – Y.” The two divisions to the right of the central one are labeled with the same designation and a sequential numerical subscript (O1 – Y1; O2 – Y2). The two divisions to the left of the central one are the same format, with the exception that the letters are in lower case and the sequence is reversed (ο2 – υ2; ο1 – υ1). These labels appear to reflect a new recording system that Daniel instituted on June 1, 1934. An entry on page 17 of his notebook documents the opening of Trench IV and states that it was done “according to NEW SYSTEM (original emphasis), 20 meters long, two wide.” A sketch plan of the new trench shows that Daniel began dividing the trench into units measuring 2 X 2 m. The length of the units remained the same as in the previous trenches (two meters) and retained their designation using Greek letters. The trench width of two meters was split into two halves each one meter wide and designated by a numerical subscript related to the position it occupied along the length of the trench (e.g. A1; A2; B1; B2; etc.). 156 DFN: 6/1/1934, p. 18. The original section drawing of this area was first sketched on page 13 of the notebook, but was crossed out with the note “See p. 18 for accurate cross section.” The drawing records the profile of the east baulk between stakes Λ (lambda) and N (nu) + 1 m, which equates to meters 20-25 of the trench. This area would be

The procedures used by Daniel to document the context of artifacts in his trenches were very crude. Because of this, most object findspots from his excavations cannot be reconstructed, except in cases where detailed descriptions, or more rarely, photographs of the in situ

151

Tom Davis (personal communication). Davis 2003, 55-6. During the 1920s this methodology was largely replaced by that developed at the University of Pennsylvania Beth-shan excavations by Clarence Stanley Fisher. This advocated excavating areas rather than trenches, combined with systematic recording. The latter technique appears to have been utilized by Daniel in his excavation of the “Palace” (House of Eustolios). This change in excavation methodology is recorded on page 24 of Daniel’s notebook dated 30-3-35. 152

23

documented, but these have either degraded to the point of illegibility or been destroyed by insects. Despite the inherent problems with Daniel’s Trench III methodology, some useful information regarding the Earthquake House and its artifact assemblage can be salvaged. Where Daniel identified in situ occupation deposits, particularly in Room 1, House 1,158 his descriptions of the material and its location within the room are quite detailed and in two cases were photographed. In these instances, the general type and shape of the objects as well as their findspots were able to be reconstructed accurately, augmenting the larger body of data recovered during the subsequent University of Arizona excavation. 3.2: The University of Arizona Excavations: 1984-87 While his excavation methods have proven challenging for posterity, nevertheless, the earthquake deposits recorded by Daniel provided the foundation upon which the University of Arizona investigation was based. Unlike the previous project, the Arizona Mission was not concerned with recovering museum quality works of art, but investigating the presence of an undisturbed earthquake debris layer that had also been observed at the Sanctuary of Apollo. Reopening Daniel’s Trench III, the University of Arizona team excavated the remains of the Earthquake House in a series of campaigns from 1984-1987. Results of the first three Arizona seasons were published as preliminary reports in the Report of the Department of Antiquities, Cyprus (RDAC),159 which were supplemented by short articles focusing on selected finds such as the lamps and amphorae.160 Although an edited volume containing the results of the Arizona Mission’s work at the Sanctuary of Apollo was published in 1987,161 no accompanying volume was produced for the acropolis excavations. A single book, Kourion: The Search for a Lost Roman City was published in 1988 and remains the only treatment of the entire Earthquake House excavation. This text was written for a popular audience and lacks basic elements required for serious scholarly research. Since this publication, no work on this material, except for a few specific objects, has been undertaken.

Figure 3.1: Trench III excavation notebook page illustrating recording protocols discussed in text. DFN 6/14/1934, 38. Courtesy of the Penn Museum. location in the trench. Although differentiation between the stratigraphic layers is illustrated and vertical measurements indicate the trench depth in this area (ca. three meters), descriptions of the strata are very general with no mention of amounts or types of artifacts recovered from each one. An additional factor that impacts the archaeological value of the Trench III material for the present study is that a significant amount was discarded following excavation, with only diagnostic or artistically valuable objects retained.157 Assessment of the Trench III material in the Kourion Museum storerooms confirms this, as its quantity appears substantially less than would be expected given the excavation size and richness of the deposit. The majority of the material saved is of little use, due not only to the poor recording outlined above but also a lack of labeling once placed in storage. Trays containing Trench III objects are simply marked “Tr. III” on the exterior, with no indication as to where the contents originated. Scraps of paper found amongst the artifacts likely held what provenience information was

The four seasons of work by the Arizona Mission resulted in the recovery of a significant in situ artifact assemblage using the latest methodological and technological techniques available. The following sections outline and discuss the field and finds processing procedures employed by this project. 3.2.1: Field Procedures During the initial season (1984), the excavation methodology on the acropolis was basically the same as that employed during the Arizona Mission’s work at the

identified by subsequent excavations as a probable small street or alleyway that separated Building A from Building B. 157 Although only a single notebook entry on page 12 attests to this activity in Trench III, it was probably not an isolated occurrence but a standard practice employed during the excavation. The entry reads: “Discard from fill (rocky, 40 cm. to 1-2 meters) 5 pails of sherds, 15 of tiles. Keep 1½ buckets of sherds; assorted tiles; and pieces of marble; etc.”

158

E.g. DFN: 6/5/1934, p. 25, 27. Davis and Soren 1985; Soren et al. 1986; Molinari et al. 1988. 160 Williams 1987; Dias et al. 1988. 161 Soren 1987a. 159

24

Sanctuary of Apollo.162 Horizontal control was maintained by a 10 X 10 m grid that had previously been established over the entire site.163 Each grid square was identified using alphanumeric coordinates (e.g. M9); letters and numerals designating the north – south and east – west axes respectively. These were subdivided into four 5 X 5 m squares that served as the primary unit of excavation. Initially, the Arizona Mission utilized “baulk and debris” methodology, but following the first season this system was abandoned in favor of an open excavation with each unit described by the architectural divisions of the structure.164 This alteration made sense since it permitted material from a single architectural space to be collected as an intact assemblage without having to cross reference between different excavation units. One anomaly with this open procedure is that except in a few cases, rooms were not bisected with each half excavated separately. This poses several problems for the present study because it did not permit the room stratigraphy to be visualized and recorded, making it difficult to assess the depositional processes that led to their formation.165

and third pages provided space for recording finds, excavation strategies, and daily progress from beginning to end. Use of these forms allowed the consistent documentation of each locus, reducing the probability of oversights and facilitating data retrieval for analysis and synthesis. In order to maximize the data recovered from the excavation, all material was retained and archived for future study. This is one of the great achievements of the project and a testament to its foresight in recognizing the assemblage’s potential to answer questions regarding daily life. As at the Sanctuary, the Arizona Mission cast the widest net possible, diligently collecting not only standard artifacts such as ceramic, metal, and glass, but also ecofacts such as bone, shells, botanical remains, and charcoal.168 Except where noted in the locus sheets and/or field notebooks, all fill appears to have been sieved using a ¼ inch wire mesh screen to prevent the loss of smaller objects.169 Flotation samples were taken from loci likely to yield floral and faunal evidence, which were analyzed toward the end of the field season.170

Each archaeological stratum or feature (e.g. wall, pit, etc.) was identified as an individual “locus” and given a numerical designation that distinguished it from others excavated in the same space/unit.166 Data concerning individual loci were recorded on pre-printed forms known as “locus sheets.” The first page contained standardized questions ranging from general observations about the physical attributes of the locus (e.g. general composition, particle size, color, etc.) to more interpretive questions such as how the locus was formed and its stratigraphic relationship to other loci in the same unit.167 The second

Artifacts and ecofacts were recovered by locus. The basic collection unit was the basket, which could be used for a specific class of material (e.g. pottery sherds) or single objects such as completely or largely reconstructible vessels. Top and bottom elevations were normally recorded at the collection limits of each basket to provide vertical context. Horizontal context was usually recorded with a rough outline of where the material originated on the unit/room plan. The latter procedure does not appear to have been regularly followed and in many cases the locations where general collections originated cannot be conclusively determined.171 Significant objects such as intact or reconstructible vessels were usually point provenienced using a surveying instrument. In cases where this option was not available, their location was triangulated from fixed, surveyed points on the architecture. Elevations were normally taken by area supervisors using a construction (“dumpy”) level. The provenience data for artifacts recovered during screening were tied to the corresponding general pottery basket for the collection area and its elevations.

162 This is the Wheeler-Kenyon method of excavation, adopted by American excavators in the Near East working at Samaria (Palestine), Pella (Jordan), and Gezer (Israel) during the 1960’s. See Dever et al. 1978; 1970; and Toombs n.d. for an outline of this methodology. Greene (1987) provides a detailed account of the excavation and recording procedures used at the Sanctuary of Apollo. For the most part, these remained in place during the Earthquake House excavations, although they were slightly modified based on the conditions encountered as the work progressed. 163 This grid was not oriented to the major points of the compass, but appears to have been laid out with respect to the long and short axes of the acropolis topography. 164 During the 1984 season, the backfill of Trench III and overburden above the debris layers to the west was removed manually. As the upper layers were composed of material deposited largely through natural formation processes containing little cultural material, it was decided at the start of the 1985 season to strip this layer mechanically to where the remains of the architecture and/or earthquake debris layer were visible. Manual excavation of the overburden was resumed for the 1986 and 1987 seasons. 165 Only in two cases, where the space being investigated was exceptionally large (Room 8, Room 11) was the area divided into multiple units for excavation. 166 The assignment of locus numbers, particularly regarding stratigraphy, was not uniform over the entire site. Rather, they were assigned based on the order in which strata and/or features were encountered during the excavation of the room. For example, the earthquake destruction layer in one room may be assigned Locus 002, whereas the same layer in another room may be assigned Locus 004. 167 Greene 1987, 7; 10-2. The format used at Kourion was similar to that developed for the American Schools of Oriental Research project at Idalion, Cyprus and used subsequently by the ASOR Punic Project at Carthage, Tunisia.

Each basket was identified by a unique number assigned by the area supervisor, which was logged into the field notebook registry and/or locus sheets before being sent for processing. In the lab, each basket was recorded in a separate registry according to the material it contained 168

Greene 1987, 4. The rapidity at which the excavation was carried out is attested by the number and size of objects recovered in the screen. These included the majority of the coins as well as a large iron tool head. Mechanically removed fill was not subjected to screening. 170 Soren and James 1987, 126. These samples were analyzed by archaeobotanist Karen Adams of the Crow Canyon Archaeological Center. The results of these analyses were not formally published. 171 The inconsistency of recording areas to which general artifact baskets belonged presents something of a difficulty for the present study. This is due to the fact that the location, size, and disposition of significant pottery fragments could help elucidate aspects of artifact use/reuse and formation processes relating to the assemblage. 169

25

cleaning, conservation, examination, and registration.175 Pottery was washed and sorted before being counted and described in terms of form and ware. This information was recorded in the artifact registry and occasionally in the pottery record section of the appropriate locus sheets. Certain diagnostic sherds (primarily finewares) and reconstructible vessels were selected for individual registration, which involved removing them from the collection basket (if necessary) and assigning them a unique pottery number and new basket number.176 This identification was inked on the object along with the excavation code, year, grid, and locus number. Registered objects were then bagged and tagged individually. In terms of the present study, this standard archaeological practice posed a challenge since it fragmented the assemblage, which had to be reassembled in order to be properly assessed. Artifacts belonging to other categories of material culture (e.g. metal, glass, stone, etc.) were sorted, cleaned, conserved (when necessary), catalogued, and labeled in the same manner as the pottery prior to being examined by the appropriate staff specialist.177 Registry numbers assigned to objects were also recorded on the locus sheets by the area supervisors.

(e.g. ceramic, metal, glass, etc.). Periodically the field and lab records were reconciled to insure that no errors had been made on either end of the process. Initially, the lab registry entries were recorded manually and assembled into notebooks for each excavation season. Beginning in 1986, however, baskets were entered directly into an electronic database that generated the registries and expedited accessing a specific artifact or set of artifacts for future study. During the initial season, architectural plans with artifact findspots were normally hand-drawn by the area supervisors. In subsequent years, this procedure was enhanced by computer plans generated from data collected by on-site surveying instruments.172 This use of computer technology in the field was beneficial, enabling plans to be produced quickly and the spatial distribution of artifacts within particular areas to be visualized. A significant drawback to the computer mapping/plotting used by the Arizona team is that a number of area supervisors appear to have discontinued manually recording plans and artifact provenience data in their excavation notebooks. In some cases the supervisors relied on the mapping team to record all provenience data, including basic elevations in the belief that they could be retrieved later, when in fact they could not.173

All significant small finds were photographed utilizing 35 mm black and white film with a scale and information card containing the registration number and provenience data of the object. In a few instances, significant fragments of diagnostic pottery (particularly rims) were drawn to scale, detailing any exterior decoration and the profile cross section. Because the Earthquake House was destroyed under exceptional circumstances its excavation produced a staggering amount of in situ artifacts and ecofacts, which required a special emphasis be placed on conservation and reconstruction.178 One of the primary tasks for conservators was reassembling the considerable number and variety of complete and nearly complete ceramic vessels broken during the earthquake.179 Their efforts resulted in a significant number of successfully restored vessels, which permitted their form to be completely visualized. In certain cases vessels were returned to the field and temporarily replaced in the room they were found so the relationships between them could be documented photographically.180 The practice of reconstructing vessels was beneficial for the present study, helping to identify individual vessels whose levels

Using computers to create an electronic artifact database and generate architectural plans were some of the most progressive aspects of the work performed by the Arizona Mission at Kourion. They are, however, also perhaps the most detrimental in terms of the present study since neither the artifact registry nor the surveying/architectural databases were technologically updated. In the case of the artifact registry this poses little problem since hand written and hard copy printouts of the finds lists were present in the excavation archive. In contrast, all the original survey and artifact provenience data from which the computer plans were generated appears to have been irretrievably lost.174 As a result, it was necessary to reconstruct much of this data based on measurements recorded in the field notebooks and surviving plans (both manually and computer generated) present in the archives. 3.2.2: Finds Processing Following recovery in the field, artifacts and ecofacts were brought to the Kourion Museum in Episkopi for

175 The Episkopi Museum, located in the former residence of George McFadden, served as the base of operations for the Arizona Mission throughout the project. 176 Each class of material had its own set of consecutive registration numbers that were continued from year to year. 177 Material culture categories for the Earthquake House are based on those used at the Sanctuary of Apollo (see Greene 1987, 20-1 for a list). 178 See Soren and James (1987, 110-114) for the reasoning that led to the emphasis on conservation efforts undertaken during the project and the problems faced by the conservation team. 179 The conservation team consisted of five full time conservators including Terry Weisser of the Walters Art Gallery, Jane Carpenter of the Brooklyn Museum, and Marian Kaminitz of the American Museum of Natural History (Soren and James 1987, 112) as well as volunteers from the field crew when necessary. 180 In only one instance were all the reconstructed vessels from a single room replaced and photographed in their findspots. In other rooms, only selected vessels were replaced and photographed in situ.

172 Soren and James 1987, 104-5. The computer mapping team consisted of John Sanders, now of the Oriental Institute of the University of Chicago, and his wife Peggy Sanders. 173 On some of the locus sheets, in the section for the recording of elevations a note of “See Sanders,” is written. 174 Personal communication by the author with John Sanders revealed that no original data used to create the computer plans survived and that the program used to create them is no longer supported on modern equipment. The only computer generated plans available for examination were the in hard copies present in the archives of David Soren at the University of Arizona. This provides a textbook example of the kind of information that can be lost when electronic data is not upgraded to be compatible with current technology. Fortunately, in this case, a significant portion of the compromised data could be reconstructed.

26

of completeness indicated they were likely in use when the structure was destroyed. In a few instances, however, these reconstructions posed an obstacle for the present study. During the first season, detailed notes were kept as to what pottery basket(s) contained fragments of a particular reconstructed vessel. In subsequent years, the rapidity of excavation and sheer amount of material requiring reconstruction were such that fragment provenience records do not appear to have been kept diligently, if at all. While the majority of the reconstructible vessels were collected as either a single pottery basket or multiple baskets with consecutive numbers, a few contain fragments from multiple baskets collected throughout a single room. In such cases where fragment origin documentation is lacking, the provenience of all sherds from a reconstructed vessel could not be definitively determined. This is unfortunate, since fragment findspots can be used to establish whether or not a vessel was likely intact, or broken and dispersed in the room at the time of the earthquake, which affects the interpretation of its use status. The location of sherd concentrations, particularly those containing diagnostic elements, may also indicate whether specific parts of a vessel were deliberately segregated and possibly retained for some type of reuse or recycling behavior.

were in agreement, a new examination of the structure’s wall junctions was conducted. The basis for this assessment was the latest 1:25 Earthquake House state plan produced by the Kourion Mapping Project in 2005.182 This was digitized using AutoCAD™, which enabled detailed drawings of the structure to be created at whatever scale required. Hard copies of the plan were taken into the field where a visual examination of the standing architecture was carried out over several days. This inspection focused on investigating and recording the relationships between the joins of the walls, general construction characteristics, and any evidence for remodeling and rebuilding that might have taken place.183 These observations were used to create a block plan that permitted the alterations and additions observed throughout the structure to be differentiated. The architectural evaluation forms an integral component of this study. It was essential to fully understand how the structure developed and changed over time in order to accurately interpret the objects and understand the behavioral dynamics present during its terminal phase. 3.3.2: Artifact Assemblage The corpus of artifacts from the Earthquake House forms the largest and most complex set of data both in its diversity and the sheer quantity of material. Dealing with such a staggering amount of information quickly became overwhelming and required differentiating objects relevant to the study (the in use and associated assemblages) from the total corpus of material unearthed during the excavations. Although the Arizona Mission computerized artifact database was no longer viable, the original registry sheets in the archive allowed a new database to be created using Microsoft Access™. A separate database was created for each class of material (e.g. ceramic, stone, glass, etc.). Each entry recorded the grid, room, locus, basket number, and a brief description of the object(s). Using multiple databases permitted criteria unique to individual artifact classes to be added, enabling searches to be carried out using specific parameters.184 The creation of these new databases, collectively containing 6,204 records,185 allowed assemblages from specific loci within the Earthquake

Despite the challenging and at times problematic nature of certain aspects of the excavation, the work carried out by the Arizona Mission was truly visionary in the way it approached the artifact assemblage. It is only because of the staff’s diligence in its collection and documentation that the data set used for this study is so rich. 3.3: The Present Study The study of the Earthquake House presented here draws upon processual and post-processual methodologies in its analysis, focusing predominantly on two components: 1) an examination and assessment of the architecture; and 2) an analysis and interpretation of the artifact assemblage. Both are complementary to one another and when combined, permit a better understanding of the structure’s complexities at the time it was destroyed. Each component utilized a distinct data collection methodology, which will be outlined in the following sections. 3.3.1: Architecture The first step in assessing the architecture was examining the previously published plans as well as unpublished versions present in the excavation archives. Although they are consistent in recording the layout of the structure, these plans vary concerning the relationships among the walls as to whether they bond or abut one another.181 The question of how walls interface is crucial for understanding how a particular structure developed over time and their evaluation is a standard methodology employed in architectural studies. Since none of the plans

182 Thanks are due to the late Dr. Danielle Parks for allowing me to use this plan for my study. This plan was subsequently published in Buell et al. 2009, fig. 1. 183 Ceramic material recovered from the walls during the Arizona excavation was examined in the hope it could provide some type of chronological indicator for construction. Unfortunately, the diagnostic fragments collected were cooking and plain wares that cannot be dated precisely. Although noted as being collected from the walls, no indication is given as to the exact location, thus the precise context is unknown. 184 For example, pottery records include categories for type (fineware, cookware, etc.), ware (African Red Slip, Cypriot Red Slip Ware, etc.) and form (plate, bowl, etc.). These criteria are not generally applicable to other categories and thus excluded from the other databases. The inclusion of category specific parameters greatly increases the flexibility and ease by which the records can be searched, enhancing its usefulness. 185 Some of the pottery baskets, although entered as a single record in the database, contain hundreds of sherds. The actual number of artifacts evaluated by the study is far greater than the number of database entries.

181

This variance in interpretation is also present in the locus sheets and season final reports. The cause for these discrepancies is the nature of the construction at Kourion, which in many cases does not allow the bonding or abutting nature of the walls to be readily discerned.

27

House to be reconstructed accurately and rapidly. In examining the various artifact classes, a number of objects were discovered that were not recorded in the official excavation registry. When encountered, missing provenience data for these objects was entered into the database, resulting in reconciliation between the material and paper records.

identify different species, estimate a minimum number of individuals, or discern patterns of dismemberment. Following analysis, the results of the room assemblage examinations were brought together and synthesized to discern patterns indicative of activities that occurred in each space. These interpretations were used to form general conclusions concerning activity areas and behaviors that reflect the use of space throughout the structure at the time it was destroyed.

After reconstructing the artifact database, an examination of the notes and locus sheets from individual rooms was undertaken to identify loci that potentially contained material relevant to the study. Once these had been identified, the artifact databases were searched to compile lists of baskets from each room, which were then assembled into an overall catalogue of material to be examined. Rather than reconstituting the total assemblage, the artifacts from each room were examined and evaluated separately according to class.186

In order to visualize the vertical contexts of artifacts within a particular room, whenever possible, section drawings created during the excavation were consulted. However, these were exceptional as a search of the archives revealed that the majority of rooms did not include such drawings as part of their documentation. As a result, it was necessary to reconstruct sections utilizing the maximum and minimum elevations for each stratigraphic unit recorded on the locus sheets.

The bulk of the assemblage, including ceramic, glass, stone, metal, and worked bone from the Earthquake House is located in the storerooms of the Kourion Museum and was available for detailed examination. A small number of objects (ca. < 5%) either in permanent museum displays or in the storerooms of the Cyprus Archaeological Museum in Nicosia were unavailable for direct inspection. In these cases, the objects were only visually examined, with relevant data (e.g. rim diameter, base diameter, height, etc.) being estimated. Fortunately, for the majority of these objects, particularly those in metal, previous publications containing accurate dimensions and descriptions could be referenced.

As the room assemblages were evaluated, findspots of individual artifacts were added on the AutoCAD™ state plan. The original computer plans generated by the Arizona Mission recorded some object findspots and whenever possible, these were consulted to insure that artifacts were located accurately on the new plan. However, not all the objects relevant to the present study were depicted on the computer generated or field drawn plans. In these cases, the artifact findspot was reconstructed using triangulation data recorded in the excavation notebooks. If such measurements were not available, or if the object was recovered in the sifting screen, its findspot had to be approximated. In some instances, objects were only able to be placed within a particular quadrant or half of a room and are differentiated on the plan by enclosure within a square or rectangular frame.

Only one group of artifacts, the coins, was not subjected to detailed examination by this study. At least a sample of these had been evaluated and identified by Eugene Lane during the Earthquake House excavation. Unfortunately, no formal publication of this analysis was made and attempts to locate Lane’s notes on the coins were unsuccessful. The coins are significant in their potential to illuminate aspects of the island’s monetary economy during the 4th century AD. However, for the purposes of this study it was decided that the examination and identification of individual coins would be of nominal value and therefore not undertaken at this time. Instead, the coins were evaluated as a single component of material culture to illuminate aspects of household storage and economics, and provide evidence for site formation processes.

Incorporating artifacts into the AutoCAD™ plan permitted relationships between the objects to be visualized readily and rapidly. In previous plans, a variety of standardized symbols were used to represent individual artifacts and/or artifact classes. These gave no indication of the object’s size in relation to others or how much space it occupied in the room. The plans created for this study feature artifacts drawn to scale based on measurements obtained during their assessment. This permits the size and spatial distribution of the objects found throughout the room to be realistically depicted. This format also allows relationships between objects and architectural space to be readily appreciated, which aided in identifying use patterns within each room and the structure as a whole.

Ecofacts recovered during the excavation (animal bone, botanical remains, and shell) were not subjected to any type of systematic study by this project. Examination was limited to animal bone and only undertaken in cases where elevated concentrations, possibly indicating waste disposal loci, were encountered. In such instances, this material was checked only to verify its provenience and the exact quantity present. No attempt was made to

The procedure used by the study to evaluate and catalog the artifacts was relatively straightforward. All relevant data for each object were recorded including its material, dimensions, provenience, description, assessment of function, and condition at the time the structure was

186 The aim of this task was to produce an inventory of the artifacts from each room of the structure rather than a detailed catalogue with typological identifications.

28

destroyed (e.g. “worn,” “appears new,” “broken,” etc.).187 In addition to sketches that accompanied the written descriptions, each object was photographed from multiple angles using a digital camera.



A specialized aspect of analyzing the ceramic and glass assemblages was their quantification to establish the probable number of vessels represented by the material from each room. The same general procedure was utilized for both classes of artifacts. In addition to recording the standard data, each specimen was also evaluated in terms of its levels of completeness (how much of the vessel is present) and brokenness (how many pieces it has broken into).188 These observations were used to provide insight into the object’s probable condition (i.e. intact or broken) at the time the structure was destroyed. Because of the significant number of largely or completely reconstructible vessels, the assemblage was quantified by determining the estimated number of vessels represented (EVREP).189 The methodology used for calculating EVREP is as follows:190  

All sherds collected from the relevant loci of a single room were laid out, and joins sought among them. The primary criterion for identifying an individual vessel was the presence of a rim since all forms in the Cypriot ceramic repertoire have them. Not all forms possess bases and/or handles. o Each rim fragment, any joining fragments, as well as non-joining sherds that could be attributed to the vessel based on form or fabric type were grouped together to create an individual “nuclear sherd family.”191 o Each rim and associated sherd family was assigned a Minimum and Maximum EVREP value of

1, which numerically indicates a distinct vessel. After all rims were accounted for, other diagnostic fragments (bases and handles) were examined and joins also sought among the recovered material. o Unless these fragments had a characteristic (e.g. fabric, decoration, etc.) that distinguished them as distinct from the rim-identified vessels, they were assigned a Minimum EVREP value of 0, and a Maximum EVREP value of 1. o These values allow for the possibility that these diagnostic elements could belong to one of the previously identified vessels, but cannot be conclusively linked to them (Minimum EVREP: 0), or could belong to a distinct vessel where the main diagnostic feature (the rim) has not been preserved (Maximum EVREP: 1).

Through this process, both a Minimum EVREP, equivalent to the minimum number of vessels (MNV), and a Maximum EVREP was calculated and averaged, which provides a more accurate estimate of the number of individual vessels that the recovered material actually represents.192 This average was used as a standardized baseline through which assemblages from different rooms of the structure could be analyzed and compared. Although the general value of EVREP as a quantification method (and the value in quantifying assemblages in general) is a topic of some debate, the establishment of levels of completeness and brokenness of each vessel is even more difficult. To the knowledge of the author, no studies have been carried out to determine definitive criteria for objectively evaluating levels of vessel completeness and brokenness.193 Due to the nature of the Earthquake House assemblage, levels of completeness and brokenness were assessed against the standard of intact or completely reconstructible specimens (completeness 100% to ca. 97%). Vessels with a high level of completeness were normally > 90% reconstructible. A very low level of brokenness was only given to vessels recovered virtually intact, with other

187

Form was also a criterion in examining the ceramic assemblage, as well as general (fineware, plainware, cookware, etc.) and specific (African Red Slip, Cypriot Red Slip Ware, etc.) fabric identifications where possible. Similarly, a general evaluation of the glass included a description of the probable form of the vessel where it was able to be determined. 188 For the glass assemblage, the overall level of brokenness was not generally considered as a significant factor in evaluating the status of the vessel. 189 Additional data used in other pottery quantification methods were collected including count, weight, and estimated vessel equivalency (EVE). This final method involves measuring a diagnostic element that is used to represent a percentage of the whole vessel (See Orton 1993, 272). EVE data for this study was collected using the rim, which was measured using a diameter chart. For additional discussion of quantification see Orton and Tyres 1992; Chase 1985; Vince 1977. 190 This process is a slightly modified version of the methodology outlined in Orton (1989, 94). 191 Orton et al. 1993, 172. This is a refinement of the term “sherd family” which is attributed by Orton to Smith (1983). A sherd family is defined as “all the sherds from the same pot,” a nuclear sherd family is defined as “all the sherds from the same pot in the same context.” This is opposed to the extended sherd family which constitutes “all the sherds from the same pot in the same collection” [e.g. a structure, a site, etc.].

192 Orton 1982, 1; Orton et al. 1993, 172. The problem is that the Minimum EVREP tends to under-represent and Maximum EVREP tends to over-represent the number of vessels present in the assemblage. By averaging these two numbers, a closer estimate of the actual number of vessels that the assemblage represents can be obtained. 193 Although Orton (1989, 97) provides a formula for calculating brokenness (Brokenness = Sherds / EVE), he does not provide any definitive numerical limits for determining various levels of brokenness (e.g. X = “Very Broken;” Y = “Not So Broken;” etc.). One possible benefit that continued examination of the Earthquake House ceramic assemblage may provide is the calculation of a definitive set of numerical values in terms of sherd size and/or weight from which levels of brokenness can be assigned to vessels.

29

levels assigned based on the character of the sherds. In the case of highly reconstructible vessels sherd size tended to be large, thus a low level of brokenness. Based on the condition of the vessel, assessed primarily on its level of completeness and secondarily on its brokenness, it was judged whether or not it was capable of being employed as a container when the structure was destroyed, thus forming part of the life assemblage. Vessels with a completeness of ≥ 90% were placed into this category.194 If completeness was below this level, the vessel was not assigned directly to the life assemblage. Instead, it was considered to have been “broken" material with a completeness level that permitted continued use, formed pre-destruction refuse, or was being cached for reuse and/or recycling in the structure. Placing material into these three groups represents the initial effort to identify sub-assemblages for examination based on the characteristics of the objects themselves. Although this process appears straightforward, in reality assigning objects to the life or death assemblage is far more complicated. This initial assessment sought only to answer questions of whether or not the object could have functioned at the time the structure was destroyed. It does not take into account the context of the object, a factor that can have a tremendous impact on its interpretation. Such contextual questions and their implications for interpreting the assemblage will be addressed in Chapter 6. While a very time consuming process, especially regarding the ceramic evidence, this detailed autopsy of the entire assemblage was invaluable. It has provided hard archaeological evidence that illuminates the organization of activities in the Earthquake House during its terminal phase. The evidence also reveals behaviors that contributed to the formation of the assemblage, not only while the structure was in use but also following its demise when the site was reoccupied on a much smaller scale.

194 This figure is designed to take into account vessels such as amphorae, which may have had their tops removed, but could technically still serve as storage receptacles.

30

Chapter 4: Architecture Accurate analysis of the Earthquake House artifacts is dependent upon an investigation of the architectural space that provided the setting where they were used. The excavator’s reading of the Earthquake House was influenced by the traditional textual approach in interpreting domestic architecture.195 This led to labeling some spaces in the structure with specific nomenclature impregnated with implications about the types of activities and behaviors that occurred within them.196 As recent scholarship has demonstrated,197 a significant flaw in this approach is that descriptions by authors such as Xenophon, Vitruvius, and Pliny the Younger focus predominantly on idealized models and elite domestic architecture. These sources do not mention the fluidity and diversity of domestic space in terms of use, a trait that has been widely documented in the archaeological record. The present work does not attempt to label architectural spaces, examining them without presupposition regarding their function. This approach is particularly applicable to non-elite domestic architecture, which the Earthquake House represents, whose elements are likely to have been arranged pragmatically rather than in adherence to specific architectural canons. Although no close parallels for the plan of the Earthquake House have been located, certain elements have been observed in other domestic structures in Cyprus and the Near East.198 The chapter begins with a discussion of construction techniques followed by an architectural analysis based on the building phases and alterations observed in the remains. The final section offers remarks and conclusions based on a comparison of the Earthquake House to other known domestic structures on the Kourion acropolis and a discussion of factors that inhibit contextualizing the structure within the urban fabric of the ancient city.

Figure 4.1: Stone foundations of the Earthquake House showing construction techniques. varying sizes and shapes.199 To provide additional structural support, fragments of limestone and in some cases pottery, were used to fill the interstices between the larger stones.200 This type of construction complicates identifying additions and changes made to the structure since its irregular nature can make it appear that walls bond and abut at the same juncture. These ambiguities likely account for inconsistencies observed in previous architectural plans.

4.1: Construction The architectural remains consist of foundations ca. 30-60 cm thick, constructed from large- to moderately-sized dressed and undressed limestone blocks (Fig. 4.1). In some cases these blocks represent reused materials salvaged from previously destroyed or damaged structures. Such recycling appears to have been widespread at Kourion, since the utilization of architectural elements, particularly columns and smaller stone objects (e.g. querns, small basins, tabletops, etc.) as components of wall construction can be observed throughout the site. The blocks were bonded together with a mud mortar in irregular courses due to their

In addition to stone, a significant amount of decayed mudbrick was recorded in the debris layers throughout the structure.201 The excavation notebooks make it clear that neither material (stone or mudbrick) was present in quantities sufficient enough to permit complete reconstruction of the walls. This indicates the Earthquake House was a hybrid construction with stone foundations supporting a mudbrick superstructure, a technique that continues to be used in Cyprus today.202 Evidence

195 Allison 1995, 146-7; 1992, 56; Cahill 2002, 150; Ault and Nevett 1999, 44-5. 196 Soren 1988, 41; Soren and James 1988, 119. One room was identified as a tablinum and another as a possible triclinium. It is acknowledged that these spaces had ceased functioning in these capacities when the structure was destroyed. However, no archaeological evidence was recovered that supports these initial identifications. 197 Most notably Allison 1999, Allison 2004, Ault and Nevett 1999, Ault 2005, Berry 1997, Cahill 2002, and Nevett 1999. 198 Soren and James 1988, 117; Soren et al. 1986, 206; Fejfer 1995; Meyers et al. 1981, 50-61; Walmsley 2007.

199

Wright 1992, 416. This type of construction has been referred to as “snecked masonry” by Connelly (Connelly and Wilson 2002, 284). 201 The use of mudbrick is acknowledged as a component of the wall construction of the Earthquake House (Soren 1988, 32-3). However, the only published reconstructions (Soren 1988, Soren and James 1988) depict the structure as being all stone. 202 Wright 1992, 414. The use of mudbrick with a stone socle appears on Cyprus as early as the Neolithic period at the site of Khirokitia. It continues to be used as a method of wall construction down to the modern period, particularly outside large urban centers. 200

31

traditional Cypriot construction techniques.205 Primary support came from one or more large wooden bearing beam(s) that ran the length of the room whose ends were embedded in the walls. Once in place, a series of smaller beams placed perpendicular to the main one(s) created a framework that was covered with smaller branches, sticks, reeds, and matting. This was then packed with mud to create the surface of the roof (Fig. 4.3). Based on the field notebooks, it appears that the majority of the structure would have been covered by flat or gently sloping roofs of this type. However, over the largest room this same wooden framework supported a pitched tile roof, the evidence for which will be discussed below.

recovered during the excavation indicates that the wall surfaces were finished with mud plaster, although some small areas with traces of degraded lime plaster were also noted. None of the evidence indicates that any wall of the structure was decorated with painted plaster (fresco). Construction of certain architectural elements, namely the doorways, appears to have been standardized. These consist of a large threshold block or blocks, with at least the lower jambs created using either dressed ashlars, or large stone slabs (Fig. 4.2). Occupation surfaces (floors) were normally of plain compacted earth although in a few instances traces of lime plaster were noted. These, however, appear to be exceptions and were difficult to identify during excavation. If the floors were plastered, the vestigial nature of the surface indicates that by the time of the earthquake they had undergone significant degradation.203 Generally, the Earthquake House floors were uneven and in many cases significantly lower than the entrance threshold. This necessitated placing a stone block (or blocks) to use a step facilitating access to the room.204

4.2: Architectural Phases/Structure Development As discussed earlier, previous Earthquake House plans vary as to what walls bond or abut; a significant factor in understanding the structure’s architectural development. Because of this, the present study reexamined the relationships between all the walls to identify discrete construction events that could be used to trace how the structure evolved over time. While it is not possible to place these architectural changes into an absolute chronological sequence at present, further excavation and study of ceramic from the wall foundations may help to

Like many of the architectural elements used in the Earthquake House, roof construction was also based on

Figure 4.2: Room 14 doorway illustrating the monolithic blocks used for the threshold and jambs with a stone block as a step. 203 Floors of beaten/packed earth are standard in village construction in ancient Cyprus (Wright 1992, 467). This technique continues to be used into the modern period for traditional village houses (Ionas 1988, 79). 204 Wright 1992, 490. The presence of sunken floors with steps leading up to the doorway is also a long standing architectural tradition in Cyprus, first observed at the Neolithic structures of Khirokitia.

Figure 4.3: Pano Kivides, Cyprus. Traditional roof construction that would have been used extensively in the Earthquake House. 205

32

Wright 1992, 494.

answer such questions. An analysis of the surviving architecture does, however, suggest a logical relative sequence that provides some insight into how the Earthquake House expanded and transformed during its lifetime. For simplicity, in the following discussion north is located at the top of the plans. The dimensions, areas, flooring, roofing, and fixtures presented in the room descriptions are summarized in Appendix C, Table 1.

architecture cannot be evaluated since they appear to have been either removed or backfilled and only cursorily described in the excavation notebooks. Room 15 and Room 18: The courtyard and its portico are the central feature of the initial construction. Although excavated as two units (Room 15 and Room 18), they form a single L-shaped space that linked the other elements of the structure. The space is bounded on the north by Wall T, the east by Wall V, and on the south by Wall B and Wall E, which form the southern sides of Room 15 and Room 18 respectively. The west side of Room 15 is defined by Wall Z and a built staircase in its northwest corner that provided access to two of the three spaces (Rooms 12 & 14) forming the west suite. The western side of Room 18 was formed by Wall A. Room 15 measures 4.1 m N-S by 7.6 m E-W; an area of 31.16 m2. Room 18 measures 4.0 m N-S by 2.6 m E-W; an area of 10.4 m2. Their combined area is 41.56 m2. In addition to the primary entrance in Wall T, secondary entrances may have existed in Wall V and Wall E. However, the possibility that these may represent

4.2.1: Initial Construction The initial phase/architectural core of the Earthquake House is composed of three elements: the courtyard and colonnade (Rooms 15 and 18), two rooms south of the courtyard (Rooms 1 and 8), and a suite of three rooms west of the courtyard (Rooms 12, 14, and 19) (Fig. 4.4). On the north side an alleyway provided access to what was likely the structure’s primary entrance that opened onto the paved courtyard and colonnade (Room 15).206 Additionally, a small space formed by walls south of Room 18 may have also been a component of the original construction, perhaps a vestibule to the doorway in Wall E. The relationship of these two walls to the surviving

Figure 4.4: Earthquake House – Initial Construction. Adapted from Buell et al.2009, fig. 1. 206

Soren 1988, 41; Soren and James 1988, 130.

33

later modifications cannot be discounted.

doorway that may have been the primary entrance to Room 1 was located in the western part of Wall B.210 Three flat stones placed inside this doorway functioned identically to those situated inside the doorway in Wall A.

The paved courtyard surface is composed of well squared and dressed limestone blocks. Some preserve angular cuttings indicating use prior to that in the Earthquake House. Along its southern edge, this paved surface is delineated by a slightly raised stylobate, upon which stood six unfluted Doric columns whose spacing is recorded in the excavation notes as being “noticeably irregular.”207 Like the paving, some of the drums and capitals appear to have been previously used. The floor of the portico and short corridor (Room 18) were composed of compacted earth. Based on the lack of tiles in this part of the structure, the colonnade roof was probably of wood and mud, sloping north towards the courtyard paving and east toward Wall V. The location of the column components at the time of excavation indicates that the colonnade and its roof were not in place at the time of the earthquake, having collapsed previously at some point.

In addition to those opening onto courtyard/portico area, Room 1 also possessed two internal doorways. One was located on the southeast side in Wall E, providing access to the adjacent Room 8. Its threshold appears to have been elevated significantly above the occupation surfaces in Room 1 and Room 8, requiring the placement of stone steps on both sides. The second interior doorway was located in Wall D, which at the time of the earthquake was used to access a space created by a later architectural addition.211 Although the nature of the original space accessed by this doorway cannot presently be determined, limited excavation west of the Earthquake House revealed sections of two walls.212 These remains align with Walls B and E to the east and may represent vestiges of this original space. A possible third internal doorway was identified by the excavators toward the west end of Wall E, which would have provided a second entrance to Room 8. The identification of a doorway here is problematic since its construction would be different than every other example in the structure (Fig. 4.5). Specifically, the “threshold” is much thinner than in other doorways and it does not span the entire width of the wall. Also, no jambs in the form of monolithic slabs or ashlar blocks can be identified, which is inconsistent with the standardization observable elsewhere.213 If some type of access point between Room 1 and Room 8 did exist at this location, it was not part of the initial construction, but a modification undertaken during the latter part of the structure’s life.

In the northwest corner of Room 15 four well-built stone steps rise westward to a nearly square landing that provides access to Room 14. From here the staircase turns roughly 90˚ southward with four stairs continuing up that provided access to other areas. Parallels for such Lshaped staircases can be observed in the 4th century AD “Patrician House” at Meiron, Israel, Cypriot architecture of the early-middle Byzantine period from Agios Konanos on the Akamas Peninsula, and a Justinianic period house (AD 527-565) excavated at Pella, Jordan. In all of these cases, the staircase provided access to upper floors.208 An analogous situation appears to exist at the Earthquake House, where both the architecture and material evidence indicate the presence of a second story over at least a portion of the structure. Arguments in favor of an upper floor will be presented and discussed in subsequent chapters.

Room 8: Adjacent to Room 1 and connecting via the doorway in Wall E was Room 8, a rectangular space measuring 7.65 m N-S by 6.40 m E-W; an area of 48.96 m2. It was bounded on the north by Wall E, the south by Wall J, and on the west and east by Wall K and Wall N respectively. This room is the largest covered space of the structure and the only one possessing a tile roof at the time of destruction. Because of its size, Room 8 was equipped with a central support consisting of six flat limestone slabs of varying thickness measuring ca. 52 X 52 cm that were stacked to a total height of 1.14 m. Above this stood

Room 1: South of the courtyard/portico is Room 1. It is defined by Wall B on the north, Wall E to the south, and Walls A and D to the east and west respectively. These form a rectangular space measuring 3.25 m N-S by 6.25 m E-W; an area of 21.12 m2. One entrance was via the short portico corridor (Room 18) through a door in Wall A. The threshold of this doorway was ca. 36 cm higher than the compacted earth occupational surface, which may have been lime plastered at one time. The elevation difference between the threshold and the floor necessitated placing a line of stones adjacent to Wall A that form a step down into the room.209 A second

when the floor of the room was likely lower, this feature would have been essential to facilitate room access. 210 This hypothesis is based on the Wall T doorway being the main entrance to the structure. If it was, then the Wall B entrance would be the nearest entrance to the structure proper. However, if the main entrance was via the doorway in either Wall V or Wall E, the Wall A doorway was closest. 211 The bonding of the northern part of Wall D to Wall B indicates that at least this section is contemporaneous with the initial construction. 212 Wall P and Wall H. 213 In his final report for the 1985 field season, assistant field director Ron Gardner disputes the assertion of a second doorway linking Room 2 and Room 8. He states: “I think the possible door of Room 2 in Wall E – apparently deliberately blocked, is not a door at all, but a frame to hold a wooden frame of some sort of shelving or storage. This frame projects out 12 cm. into Room 2…It has supporting blocks underneath…This is not a threshold…No door here.” (GFR 1985, 13).

207 Five capitals were recovered by the Arizona excavations. Daniel records finding one capital during the Trench III excavations (DFN: 6/14/1934, 38). Several are described as having “cuttings.” The column drums varied in length from 0.22 – 1.66 m. An analysis of the column dimensions and diameters allowed for an estimated maximum height of 2.76 m for each column. 208 Fejfer 1995, 85; Meyers et al. 1981, 50-61; Walmsley 2007, 257. 209 AZLS Room 1. The threshold elevation is recorded as 68.05 ASL, and the top elevation for the stone step 67.81 ASL. The degree to which this feature was necessary during the final phase of the structure is debatable, since the level of the packed earth floor was only slightly lower (12 cm) than the top level of the steps. During earlier periods,

34

possibly some of the tegulae, were secured in place using this method. Corinthian type imbrices were then cemented to the underlying tegulae with lime plaster containing small river gravel aggregate.216 The occupation surface in Room 8 was of beaten earth whose matrix displayed little or no preparation, although multiple resurfacings were noted.217 It is described as being uneven and sloping gently to the north. As was the case in Room 1, the floor of Room 8 was significantly lower than the doorways that provided access to it, in some cases by as much as 43 cm.218 This required placing blocks as steps in these locations, one of which was an upside down pilaster capital adjacent to the Wall N doorway.219 Along with the previously discussed entrance from Room 1 in Wall E, Room 8 possessed two additional doorways; one on the east side through Wall N, and the other on the opposite side of the room in Wall K. Because the Earthquake House underwent significant structural changes during its lifetime, it is unknown whether one or both provided exterior access or linked internal spaces. Given the probable existence of a room belonging to the initial construction west of Room 1, it is plausible that the Wall K doorway, directly to the south of the one in Wall D, may have originally connected two interior spaces. With the exception of the central column, no built features (e.g. benches, basins, etc.) were discovered in Room 8. However, it was equipped with two “semipermanent” fixtures220 both of which appear to have been fashioned from salvaged/reused material (Fig. 4.6). The first is a large stone table on the west side of the central column. It was composed of a rectangular alabaster slab (S88), measuring 86 X 80 X 7 cm and estimated to weigh ca. 110 kg.221 Three legs that appear to be carved from local limestone (S89, S90, S91) were discovered in association with this piece. Although the circular legs are roughly the same size with a diameter of ca. 14 cm, each is different and only one (S90) is complete, the other two being broken on one end.222 While it is possible that the tabletop was supported only by three legs, based on its size and weight a fourth leg (probably in wood) was almost certainly used as well.

Figure 4.5: Area of previously hypothesized “second doorway” in Wall E between Room 1 and Room 8. Top: North face of Wall E looking south; Bottom: South face of Wall E looking north. two column drums measuring 0.96 and 1.14 m in height with a diameter of 42.5 cm. Combined, these elements gave the feature a total height of 3.24 m.214 The exact arrangement of the roof timbers cannot be determined, but based on archaeological and ethnographic data from Cyprus, it probably consisted of a central north-south oriented ridge that sloped to the east and west at an angle of ca. 22.5˚.215 Considering the number of iron nails recovered from Room 8 (ca. 250), it appears that a significant portion of the wooden elements, as well as

216 GFR 1985, 34. A minimum of 228 nails were recovered from the final occupation surface and earthquake debris layer of Room 8. 217 GFR 1985, 31. 218 In Room 8, the difference between the threshold and floor level for the doorway on the east side of the room in Wall N is ca. 43 cm. On the west side, the doorway in Wall K has an average difference of ca. 33 cm (46 cm max; 20 cm min). See plans in GFR 1985, 32; and AZLS Room 8. 219 GFR 1985, 33; AZLS Room 8. This element could not be located on site or in the storerooms for examination. 220 These are defined as large objects that were not easily portable and made of heavy materials. Although they could be moved, significant effort by at least two individuals would have been required. 221 The tabletop is heavily fragmented, so an accurate weight could not be obtained from the object itself. Based on its dimensions (GFR 1985, 33), the weight was calculated using a specific gravity of 2.3. This is dependent on the identification of the stone as alabaster being correct. 222 AZLS Room 8. S89: 24 cm; S90: 30 cm; S91: 33 cm.

214

GFR 1985, 31. Such features are common in traditional Cypriot architecture and serve as bearing elements for the main roof beam. The central support often coincides with a join between two beams that formed the main roof timber. It is unknown whether the column of Room 8 was used in this capacity or composed of a single beam (c.f. Wright Vol. 2, Fig. 329). A possible base for a second support may have been located adjacent to Wall J, where a squared stone block was discovered during excavation. It was postulated that this served as a base for a wooden support beam. Given the nature of Cypriot roof construction, where beam ends are typically set into the walls themselves, it seems unlikely that this element would have required additional support in this location. It is possible however, that this support was necessary since the wooden beams are reported to have been weakened by insects (Soren et al. 1986, 202; Soren and James 1988, 115) and may have been in danger of collapsing. 215 SFN 1985, 57; GFR 1985, 32. This calculation is based on the number of tiles recovered, as well as the impressions of overlapping tiles preserved in the plaster used to bond them together.

35

adjacent Room 12 as well as the upper room(s). As in other spaces, the floor of Room 14 was significantly lower (73 cm) than its entrance, necessitating a high step inside the room to facilitate access. A medium-sized hemispherical mudbrick oven, similar to those still used in Cyprus today, was located in the room’s southeast quadrant. The height of this feature cannot be reconstructed, but based on the preserved dimensions it occupied roughly 1.3 m2;226 ca. 15% of the total room area. The placement of this feature in the southeast quadrant close to the intersection of Walls D and Q created two narrow spaces along its eastern and southern sides. Chinking made from recycled pottery forms a major construction component in Wall D and Wall T (Fig. 4.7). This is striking and anomalous, especially when compared to other walls of the structure including the other walls of Room 14. While the reason for the predominance of ceramic chinking in these walls is unclear, it may represent repairs to damage that occurred during a previous seismic event. If this is the case, recycled broken ceramic (which would have been ubiquitous throughout the site seismic event or not) may have been viewed as a better option than the usual limestone chinking (at least in the short term) since it could be produced with a minimum amount of effort and/or tools. It could also be modified quickly to fit into any space where needed, which would have assisted in making Room 14 functional as quickly as possible. It is also noteworthy that the Room 14 chinking is very homogenous, composed predominantly of amphora body fragments with only minor amounts of cookware and plainware – a preference that may be reflected in ceramic evidence discovered elsewhere in the structure and discussed in future chapters.

Figure 4.6: Room 8 plan showing locations of semi-permanent fixtures. Reconstructed by the author from AZLS 8. The second fixture was located in the northeast corner of the room and consists of an upside down column base that measures 68 cm in diameter. On top of this, a “well worn” Gabbro groundstone, likely originating from the Troodos massif to the north, was discovered.223 Based on descriptions in the notes, these two objects seem to form a small workspace that would have been utilized predominantly for the preparation/grinding of foodstuffs.224

Room 12: Immediately south of Room 14, Room 12 is bounded by Wall Q to the north, with its east and west sides formed by Walls D and F respectively. The south side of the room is formed by Wall B. A north-south cross wall, Wall AA, that does not bond to Wall B or Wall Q roughly bisects the space into west (12A) and east (12B) halves. The western part (Room 12A) was a cistern measuring 2.2 m N-S by 1.25 m E-W; an area of 2.75 m2. The interior wall surfaces were prepared with a layer of hydraulic plaster that is preserved to a height of 1.79 m, providing a minimum volume of 5.00 m3.227 Since no drain was found in the lower walls or bottom, access to the cistern had to be from the top by means of the stairway. The eastern half (Room 12B) was filled in with earth that buttressed Wall AA against the hydraulic pressure exerted against it.228 While the majority of the rooms utilized traditional stone foundation / mudbrick

Room 14: Room 14 is bounded by Wall T on the north, Wall Q to the south, and Walls D and F to the east and west respectively. These form a roughly square space measuring 3.16 m N-S by 2.82 m E-W; an area of 8.85 m2.225 This space was accessed from the paved courtyard (Room 15) via the staircase and small landing fronting the Wall D doorway. From here, the staircase turns 90° to the south and continues upward, providing access to the 223

GFR 1985, 33. Stone identified as Gabbro by Reuben Bullard. Gardner (GFR 1985, 33) also records a “rough little construction of small stones with an ashlar block on top projecting 55 cm into the room from Wall N” located at the intersection of Walls E and N. It is possible that this feature was part of the same installation (possibly a makeshift seat of some sort). No detail drawings or photographs of these two features could be located in the excavation archives. 225 Molinari et al. 1988, 171. 224

226 Molinari et al. 1988, 174. Because the perimeter of the oven is irregular, its area was calculated using the formula for an oval (A = (L X W) X 0.8) with the published dimensions of 1.37 and 1.18 m. 227 Molinari et al. 1988, 171-3. This likely accounts for only a fraction of its original capacity. 228 See Molinari et al. (1988, 177-8) where it is concluded that Wall AA was not strong enough to resist the hydraulic force exerted on it by

36

interpreted by the excavator as a storage area for something accessed frequently but left no significant remains in the archaeological record.230 4.2.2: Architectural Additions: Two major additions were identified based on the abutting relationships between their walls and those of the initial construction (Fig. 4.8). At present, the order they were constructed as well as their date cannot be determined. It is possible that they represent roughly contemporaneous additions carried out during a period of general expansion and/or refurbishment. Addition A: This is located on the west side of the structure and consists of Room 3, Room 6, and Room 7. The basic space was created by extending Wall J to the west and Wall F to the south. Division of this space was accomplished by constructing an abutting full length cross wall west of Wall E that created Room 3; and an abutting partial cross wall that partitioned the remaining area into the adjoining Room 6 and Room 7.

Room 19: This small space lies immediately east of Room 12. It is bounded on the west by Wall D and on the south by Wall B. The north side of the room is formed by the foundation of the stone staircase, which continued in wood over this space. This wooden portion was supported on the east side by Wall Z, through which a doorway provided access from the colonnaded portico and paved courtyard (Room 15) to Room 19. The space formed by the walls and staircase measures 2.96 m N-S by 1.16 m E-W; an area of 3.43 m2. Its floor was significantly lower (ca. 40 cm) than the adjoining courtyard, but unlike other rooms, no step was present to aid with access. The absence of such an element is notable and has implications as to how the space was used. This will be discussed in detail in Chapter 7. The threshold block of the entrance to Room 19 was noted as being “well worn,” which according to the excavator indicated “frequent passage into and out of the room.”229 Combined with its rather cramped nature, the condition of the threshold led to this space being

Room 3: This is located west of and adjacent to Room 1, with the two communicating via a doorway in the common Wall D. The remaining walls consist of Wall F to the west and Walls B and E to the north and south respectively. These create a rectangular space measuring 3.5 m N-S by 2.75 m E-W; an area of 9.62 m2. Based on the wall vestiges discovered west of the Earthquake House, it is possible that this space replaced an earlier one belonging to the initial construction. Unlike the rooms previously discussed, the compacted earth floor in Room 3 was level with or slightly higher than the entrance. During excavation, the doorway in Wall D was the subject of debate because it appeared to have two threshold blocks one on top of the other forming a step (Fig. 4.9), with the upper stone rotated ca. 90˚ from the lower. This “displacement” was used by the excavator as evidence for the power of the earthquake, which supposedly caused this block to be lifted and twisted from its original position.231 A double threshold would be unique in the Earthquake House, but an examination of this feature’s remains and associated stratigraphy offers an alternate interpretation. The upper block is slightly smaller than the lower and irregular in shape, which is very different from other thresholds in the structure. Between the upper and lower blocks is a layer of fill ca. 10-15 cm thick. If the rotation hypothesis is correct, this fill would have to be deposited in the moments that it took for the stone to shift – essentially instantaneously since the earthquake is hypothesized to have lasted only ca. 19 seconds.232 Given the near impossibility of this, a more plausible scenario is that the upper block was originally a door jamb that fell during or shortly following the earthquake after the fill

the water contained in the cistern on its own. The earth buttressing was therefore essential to the functioning of this installation. 229 Molinari et al. 1988, 175. While it is possible that the wear was caused by prolonged usage, no other thresholds appear to show the same characteristics.

230 Molinari et al. 1988, 176. The hypothesis is that Room 19 was used for the storage of ephemeral material, most likely having to do with the maintenance of the mule in the adjacent stable space of Room 2. 231 Davis and Soren 1985, 297; Soren and James 1988, 87. 232 Soren 1988, 37. The most damaging wave lasted only 10 seconds.

Figure 4.7: Ceramic fragments used as wall chinking in Wall T, Room 14. superstructure construction, because of its constant exposure to moisture, an all stone elevation for Room 12 is highly likely.

37

Figure 4.8: Earthquake House – Plan showing additions and modifications. Adapted from Buell et al. 2009, fig. 1. was deposited. Room 6 and Room 7: The two rooms south of Room 3 were created from a single architectural space, bounded by Wall E to the north, Wall J to the south, and Wall F and Wall K on the west and east respectively. It was divided roughly in half by a short partition wall, Wall L, but left an opening whereby the adjoining spaces could communicate: Room 6 to the south and Room 7 to the north. These rooms measured 3.7 m N-S by 3.05 m E-W and 3.24 m N-S X 2.9 m E-W; areas of 11.28 m2 and 9.39 m2 respectively. A doorway in Wall K originally linked these two spaces to the remainder of the structure via Room 8. The occupation surfaces of both rooms were of packed earth. Like Room 3, they were roughly level with the Wall K threshold, which was significantly higher than the occupation surface of Room 8.

Figure 4.9: Proposed “double” threshold in Wall D linking Room 2 and Room 3. Note the thick layer of fill between the upper and lower stones.

In Room 6, a long, roughly trapezoidal stone feature built using the same material and techniques as the walls was located in the southwest corner abutting Wall F and Wall J. This feature is ca. 1 m wide on its southern end (Wall 38

J), tapering to 58 cm at its northern end where two blocks appear to form steps. The precise function of this feature is unknown at present. It was originally hypothesized to be a staircase foundation providing access to a door in Wall J.233 This interpretation is problematic, however, since no large threshold and/or jamb blocks characteristic of doorway construction were found in Wall J or Wall F (Fig. 4.10).

from the structure is the most logical and probable arrangement. Addition B: The second addition is a large open air space, Room 11, built on the southeast side of the structure. The space measures 7.7 m N-S by 13.4 m E-W; an area of 103.18 m2. It is bounded on its southern side by an extension of Wall J236 and on the east and west by Wall S and Wall K respectively. A section of Wall E may define the northern limit of Room 11, although its line is lost ca. 3 m east of Wall V. This raises questions as to whether the space was completely enclosed or remained open on its northeastern side. At the northern end of Wall S, a small wall fragment (Wall Y) may have formed part of the northern boundary of Room 11. However, the possibility that it represents an internal partition of this space cannot be discounted at present.

A great deal of decayed mudbrick was recorded in the fill of Room 6 and Room 7, indicating that the upper portions of the walls and probably the roof incorporated a significant amount of this material in their construction. Although roof tiles were recovered from both rooms, their numbers were insufficient to provide full coverage over this area. Instead, they were likely utilized to perform ad hoc repairs on damaged areas of the roof.234 The tile roof covering the adjacent Room 8 would have required the Room 6 and Room 7 roof to be sloped at least slightly east to west to insure the significant amount of runoff generated was evacuated efficiently. This architectural arrangement is distinctly different from the published reconstructions, which depict a tiled roof over this part of the structure as well.235 Such a design is counterintuitive architecturally since the west slope of Room 8 and the east slope of Rooms 6 and 7 would intersect atop Wall K, directing their runoff to a single point along the length of the wall. If put into place, the published roof system would have caused serious issues within these three rooms as well as significant architectural degradation to the mudbrick portion of Wall K. Based on this, the hypothesis of a single sloped roof covering Room 6 and Room 7 to channel water away

Given the uncertain northern boundary of Room 11, the degree of access from the exterior cannot be determined with any degree of confidence.237 Interior access to Room 11 was possible from two adjacent spaces: the short corridor (Room 18) via the doorway in Wall E; and Room 8 via the doorway in Wall N. Based on the elevation data collected during the excavation, the thresholds of both doorways were raised above the original surface of the courtyard, necessitating the placement of steps to facilitate access.238 4.2.3: Modifications/Divisions With these two major expansions the Earthquake House reached its greatest architectural extent, with further modifications taking place through the remodeling and division of pre-existing space. Such breakup of structures into smaller units is a common phenomenon during Late Antiquity. While this has been predominantly observed in public buildings, examples in private structures have been documented as well.239 In the case of the Earthquake House, the alterations significantly transformed the overall use of space. It is probable that most (if not all) of these changes took place during the final occupation, but as with the architectural additions a definitive chronology cannot be determined. Division of Room 1: One the most notable alterations to the Earthquake House plan is the division of Room 1 into two spaces: Room 1 to the east and Room 2 to the west. These measure 3.25 m N-S X 2.4 m E-W and 3.2 m N-S X 3.0 m E-W; areas of 7.8 m2 and 9.6 m2 respectively. This division was

Figure 4.10: Intersection of Wall J and Wall F showing the lack of evidence for a doorway to Room 6.

236 This is a separate construction from Wall J in Room 8. Evidence for this is a definitive abutting relationship along with the fact that Wall J in Room 11 is noticeably narrower than that in Room 8. 237 The presence of a doorway in Wall V near the intersection with Wall E that appears to provide access to the exterior of the structure may indicate that Room 11 was in fact closed off. 238 The locus sheets indicate the initial packed earth surface in Room 11 was at an elevation of ca. 67.47 ASL. The threshold elevations for the doorways in Wall E and Wall K are 68.29 and 67.84 ASL respectively. 239 See Ellis (1998: 567-9) for a general discussion of the archaeological evidence for this phenomenon.

233 GFR 1985, 23; Soren and James 1988, 107. It is possible that the level of the doorway was above the preserved height of the wall. 234 Room 6: Schiffer has pointed out that it would take 4 tiles to cover 1 m2. Since only ten tiles were recovered in Room 6, an area of 2 – 2.5 m2 could have been covered by tiles. Room 7: Based on the number of left and right bottom corners Schiffer provides estimates of 14 pan tiles, which could cover an area no larger than 3.0 m2 (SFN 1985, 27-9). 235 Soren 1988, 32-3. See also the first set of color plates in Soren and James 1988. This problem was brought to the author’s attention by Rachel Zebrowski, a professional architect.

39

accomplished through a crudely built north-south wall, Wall C. An examination revealed that this wall does not bond to either Wall B or Wall E, indicating that it was not part of the initial construction. The northern end of Wall C makes a 90° turn to the west, forming a short section of Wall B that appears to follow its original line. In the adjoining Room 1, however, Wall B seems to have been shifted ca. 10-20 cm to the north, possibly the result of reconstruction following a collapse.240 Given the fact that this is a late subdivision, it is likely that Wall C did not rise to the full height of the room, a phenomenon observed at other sites throughout the Mediterranean.241 At the time the structure was destroyed, Room 2 was being utilized as a stable. The space was dominated by a large rectangular limestone trough, measuring 2.35 X 0.65 X 0.43 m with an interior depth of ca. 26 cm (Fig. 4.11).242 This fixture was oriented on the east-west axis of the room, essentially bisecting it into northern and southern halves with a 90 cm gap on its west side. The sides of the trough are pierced with a number of tethering holes, some of which had been worn through indicating that it had been in use for a significant amount of time. While this object possibly formed part of the structure’s “original equipment,” given its size and apparent long use-life it more likely represents salvaged material (possibly from a public location) installed just prior to or following the subdivision.

Figure 4.11: Stone trough in Room 2. stones with less chinking. The stones were set to form a smooth surface on the west face of the wall, with the eastern face left rough and irregular (Fig. 4.13). This attention to detail on the Room 6 and Room 7 face seems to imply that the blocking was initiated from this side and may indicate that these spaces were conceived of as being separate at the time of construction and not intended to interact with Room 8. Since this doorway was blocked, another entrance into Room 6 and Room 7 must have existed, possibly in Wall J. As discussed previously, however, the architectural evidence as it stands does not support this hypothesis.

Blocked Doorways: The spatial remodeling of the Earthquake House was also achieved by blocking two doorways; one in Wall E joining Room 1 and Room 8; the other in Wall K joining Room 6/Room 7 and Room 8. The blocking of Wall E appears very regular and even. It was built using the same materials and techniques as the walls, so that the only indications of a doorway are the ashlar jambs (Fig. 4.12). Although this modification may be contemporaneous with the division of Room 1, the possibility it occurred earlier or subsequently cannot be excluded. The closure of this doorway essentially divided the structure into two wings separated by Wall E that were accessible only via the doorway at the southern end of the portico corridor (Room 18) and possibly Room 11.

Room 11: Over the course of its use, the ground level of Room 11 was raised so that its final occupation surface was ca. 68 cm above where it had originally been. Limited excavations below this level revealed a series of fill deposits containing significant amounts of material culture that covered the original occupation surface. It is unknown whether or not this buildup represents gradual deposition of occupant-generated waste over time, a rapid accumulation such as that resulting from cleanup efforts

In the case of Wall K, blocking the doorway isolated Room 6 and Room 7 from the remainder of the structure creating a completely separate and autonomous architectural unit. In contrast to Wall E, the Wall K blocking appears distinctly different, utilizing larger 240 In the section of Wall B in Room 1, two construction phases were discerned during excavation (DavisFN: Locus 007; AZLS Room 1, Locus 007). The original section of the wall was well constructed using the traditional irregular stone and chinking technique found in the remainder of the structure. Above this, the later wall was constructed, creating a small setback between the two phases. This later phase is less carefully built and contains a significant amount of reused material including dressed ashlar blocks and a large column drum. 241 An especially good example of this is a low wall dividing the triclinium in the Freudenhaus in Ephesus, Turkey (Ellis 2000, 111). 242 DavisFR 1984, 3. This feature was estimated to have weighed in excess of 350 kg. It was supported on two rough ashlar blocks, one of which cracked during the earthquake.

Figure 4.12: Blocked doorway in Wall E originally linking Room 1 and Room 8.

40

following a disaster, or some combination of both.243 Elevation of the Room 11 ground level above the Wall N doorway threshold necessitated installing a series of vertical, flat stones along the outer edge of the block that acted as a barrier to help prevent fill from spilling into Room 8.244 In contrast, the threshold of the doorway in Wall E that provided access to the portico corridor (Room 18) was not affected in the same way, although the need for the additional step would have been negated.

relationship between these two features and the order in which they were constructed cannot be determined. Unfortunately, the ability to answer questions regarding these features may be permanently lost since the excavation notes record only cursory descriptions of both in and the remains themselves do not appear to have survived. Rooms 15 and Room 18: In the paved courtyard, a number of significant changes occurred during the final occupation phase. The most notable is the collapse and partial rebuilding of the Doric colonnade supporting the portico roof.246 A number of column drums as well as one capital were stacked in the northeast corner of the courtyard. These may have been reused as building material/foundation elements for a shallow (ca. 20 cm deep) basin constructed on the paved surface (Fig. 4.14).247 Small patches of hydraulic cement preserved on Wall T may indicate that this feature was some type of water catchment/containment installation. If this is the case, it may have been built as a replacement (possibly temporary) for the cistern of Room 12A, which had ceased functioning sometime during this period. Other architectural elements from the colonnade were placed against Wall T, possibly as stabilizers to prevent its collapse.

Following the infilling of this space, two features were constructed in the southwest quadrant of Room 11. The first is a set of two perpendicular non-bonding walls, Wall AI and Wall AJ, which formed some type of crudely built space identified as a “lean-to” in the excavation notes. The function of this feature cannot be determined at present. Construction of the north-south Wall AI created a small rectangular space in the southwest corner of Room 11. Within this area, a hemispherical mudbrick oven of the same general type found in Room 14 was built, occupying an area of ca. 3.2 m2.245 Currently, the

In the portico itself two mudbrick ovens, one small (0.5 m2) and one medium-sized (1.2 m2) were constructed adjacent to Wall V. Given the two functional ovens in Rooms 14 and Room 11, the construction of these additional cooking installations is curious. Based on their size, it is likely they were used for smaller scale activities that required less fuel and/or prep-time than the other two. It is possible that they may also reflect seasonal cooking activities (e.g. summer vs. winter), temporary replacements for damaged ovens, differential cooking based on food types, or a combination of all these factors. Room 20 and Room 25: The final modifications to the Earthquake House occurred on its northern side, where the small alley that provided access to the paved courtyard was walled off. This created an enclosed rectangular space, Room 20, which measures 3.1 m N-S X 6.85 m E-W; an area of 21.23 m2. This room is bounded on the north by Wall AG, the south by Wall T, and on the east and west by Wall V and Wall W respectively. The only means of access appears to have been through the doorway in Wall T, which previously served as the structure’s main entrance. If this was in fact the case, with the construction of Wall W the main entrance must have shifted to another location, possibly the doorway in Wall V that opened onto Room 18. The final occupation surface in Room 20 appears to have been only slightly lower (ca. 5-10 cm) than the

Figure 4.13: Blocked doorway in Wall K joining Room 6/7 and Room 8. Top: Looking east toward Room 8. Note the “finished” appearance of the blocking and its flushness with the wall face. Bottom: Looking west toward Room 6/7 showing the unevenness of the blocking on this side. 243 Unfortunately, only a small section of these earlier deposits has been excavated, so a full picture of the processes that created them cannot be fully developed at present. 244 GFR 1985, 42. 245 No published dimensions exist for this feature. Based on the plans generated during excavation, the dimensions of the oven are ca. 2.2 m north-south, and 1.8 m east-west.

246

Soren and James 1988, 129. AZLS Room 15, Final Report: 15-6. Unfortunately, Room 15 was one of the most poorly documented during the excavation with no detailed descriptions or drawings of this “installation” in the northeast corner. Although the feature does appear on the computer generated plans, these do not provide a good overall visual record. No photographs of the feature could be located in the excavation archive. 247

41

the keyhole nature of the excavation. At present, the Earthquake House and Building B stand as isolated structures on the eastern end of the acropolis. Virtually nothing is known about the surrounding urban fabric so that their relationship with neighboring buildings and their role in the ancient city cannot be determined.250 Problems with contextualizing the Earthquake House architecture extend beyond Kourion itself to include not only late Roman Cyprus but the eastern Mediterranean in general. This lack of context begins at the most basic level, since published excavations of comparable nonelite structures are generally lacking; the focus instead being predominantly upon public and elite domestic architecture. Excavations in Cyprus at the sites of Kalavassos-Kopetra and Agios Kononas have revealed the remains of non-elite domestic structures, which are the closest architectural parallels to the Earthquake House on the island.251 Regrettably, these are of limited comparative value since they were only partially excavated and are significantly later in date (6th-7th centuries AD). Nevertheless, a number of observable similarities indicate an architectural continuity in these types of structures on Cyprus between the Roman and Byzantine periods.

Figure 4.14: Room 15 and Room 18 plan indicating the locations of late features and loci.

Perhaps the best comparative evidence for the Earthquake House comes from the so-called “Patrician House” at Meiron, Israel, believed to have been constructed in the 3rd century AD and abandoned sometime around AD 360.252 In terms of its architecture, the Patrician House possesses many similarities to the Earthquake House including a paved courtyard, L-shaped staircase, and a series of upper rooms. Although it is roughly half the size of the Earthquake House, the Patrician House is described by its excavator as a “villa.” The use of this term is misleading, since it is laden with implications of elite architecture (e.g. mosaic floors, colonnades, wall paintings, etc.) and society. Because the Patrician House and the Earthquake House lack these traditionally “elite” characteristics, they represent examples of non-elite “vernacular” architecture253 that rely heavily on the local traditions and cultures that produced them.

entrance threshold, with the upper sections of the walls and the roof being composed largely of mud and/or mudbrick. At some point, Room 20 was subdivided by a crudely built north-south wall, Wall AE, immediately east of the doorway in Wall T. This created Room 25, a small, square space measuring 3.1 m N-S X 2.8 m E-W; an area of 8.68 m2. Given its low quality construction,248 it is almost certain that Wall AE did not extend to the full height of the room. Instead, like Wall C, it served as a partition that created two separate spaces while still permitting communication between them. 4.3: Discussion In examining the architectural evidence, the Earthquake House appears to be a fusion of Greco-Roman and traditional Cypriot elements in its overall plan and construction. Unfortunately, at this time several obstacles prohibit a more definitive understanding of the various additions and changes made to the structure throughout its history. Perhaps the most critical is the lack of extensive excavation below the final occupation level. The results from such work would undoubtedly help determine at least a rough chronological framework for its initial construction and clarify the progression of expansions and internal divisions.249 A second obstacle is

Despite these problematic aspects, the architectural analysis of the Earthquake House presented here does permit some general conclusions about the structure to be drawn. As it stands, the evidence does not support the excavator’s interpretation of an upper-scale residence that was in a state of decline at the time it was destroyed. Rather, its simple, functional architecture points to the Earthquake House as being a non-elite “vernacular” structure, an interpretation also supported by its relative overall size. At its greatest extent, the Earthquake House occupied an area of ca. 397 m2, which is substantially

248 At the time of its excavation, much debate is recorded as to whether or not this feature was actually a wall, or “created” by the excavators from the fallen debris. Based on the position of a large ashlar block on the southern end, which appears to be deliberately set, it was finally agreed that Wall AG was intentionally constructed. See AZLS Room 20 “Addendum to Summary of 1986.” 249 Very limited excavation occurred below late 4th century AD levels in Rooms 1, 2, 6, 11, and 20.

250 This is one of the research questions that will hopefully be answered by the Kourion Urban Space Project (KUSP) excavations. The author is currently attached to this project as Director of Material Culture. 251 Rautman 2003; Fejfer and Mathiesen 1995. 252 Walmsley 2007. 253 Ellis 2000, 17-8.

42

smaller than the other definite domestic structure excavated on the Kourion Acropolis, the House of the Gladiators. This was an elite construction, which despite not being preserved or excavated to its full extent occupies an area of ca. 1,110 m2; roughly 2.79 times larger than the Earthquake House (Fig. 4.15).254 In addition to its size, the House of the Gladiators incorporated elaborate architectural elements and mosaic pavements signifying it belongs to a substantially superior class of construction than that represented by the Earthquake House. Conversely, the artifact assemblage from the Earthquake House, which will be discussed in Chapter 6, indicates that while its inhabitants likely occupied the lower end of the socio-economic scale, they were not completely without means. It is precisely because the Earthquake House represents urban non-elite architecture, whose study has traditionally been neglected in favor of more elaborate public and private buildings that makes its study and the insights gleaned from it particularly significant for studying this aspect of late Roman society.

Figure 4.15: Comparative plans of the House of the Gladiators and the Earthquake House illustrating their difference in size. After Loulloupis 1971, fig. 1; Buell et al. 2009, fig 1.

254 Although other structures excavated on the Kourion acropolis are identified using the designation “house” (i.e. The House of Achilles;The House of the Apsed Triclinium; and The House of Eustolios), their identification as having a domestic function is not certain. It is possible that they were used in a public or semi-public capacity.

43

Chapter 5: Room Stratigraphy/Formation Processes The Earthquake House represents a rare instance of controlled excavation on a structure catastrophically destroyed with little subsequent disturbance. Despite these circumstances, surprisingly, substantially less uniformity was recorded in the stratigraphic sequence than would likely be expected. This posed challenges for the present study in identifying, reconstructing, and analyzing the in use and associated artifact assemblages present when the structure was destroyed.

uniformly throughout the Earthquake House. It is possible that their compositional similarity with the underlying earthquake debris deposit in many cases caused them to be overlooked. Although some objects from these layers may have fallen from shelves or an upper story, the bulk represents objects deliberately discarded or residual material deposited through natural formation processes (e.g. wind, slope wash, etc.). 4) Post-Earthquake Deposition: Created following the earthquake and not associated with the continued collapse of the structure. These layers can be divided into those formed by natural processes such as wind action and slope wash and those formed by cultural processes such as dumping. The material culture from these layers represents discard and residual material, some of which may have infiltrated into lower strata presenting the possibility for some degree of contamination. The value of these layers lies in their ability to elucidate formation processes that affecting the Earthquake House following its destruction and provide insight into how this part of Kourion was used following its reoccupation.

An understanding of the stratigraphy is crucial for examining two factors that can impact the interpretation of the artifacts in the assemblage: 1) assessing the context from which a particular object was recovered; and 2) appreciating the natural and cultural formation processes that affected the structure pre- and post-earthquake. As outlined in Chapter 3, each archaeological stratum or feature excavated at the Earthquake House was identified as a “locus” and given a numerical designation differentiating it from the others.255 With the exception of Room 2 and Room 3, which were excavated based on 5 X 5 m subdivisions of the 10 X 10 m site grid; the architectural spaces were used as the main units for excavation and recording. The loci identified in the Earthquake House can be divided into four types. Each deposit represents a distinct phase in the formation of the archaeological record: 1) Final Occupation: Created during the period prior to the earthquake. It consists primarily of occupation surfaces but may also include other features such as refuse deposits and pits. These loci provide the archaeological context for the bulk of the assemblage from the Earthquake House, containing in use and discard material as well as objects that may have been dislodged from shelves or fallen from upper stories. 2) Earthquake Debris: Material deposited as a result of the initial collapse of the structure during the earthquake. Laid down directly above the final occupation loci and composed predominantly of building materials (stone and mudbrick) from the walls and roofs of the structure. In use objects from these deposits were likely dislodged from shelves or fell from upper stories. Gaps in this layer permitted some deposition of post-earthquake material either incidentally or through deliberate discard; the latter being less probable. Because of the possibility for contamination, these loci were carefully scrutinized to determine if the objects recovered from them were in use or represent post-earthquake contamination. 3) Post-Earthquake Collapse: Formed by the continued degradation and collapse of the structure following the earthquake. Loci of this type were not encountered and/or recorded 255

5.1: Stratigraphic Anomalies While the above deposit sequence likely existed to a greater or lesser extent in every room of the Earthquake House, in many cases one or more was not identified by the excavators. As a result, factors contributing to the variance between the probable strata present in each room and what was recorded (or not recorded) by the excavators must be examined and explained. No previous excavation that could have disrupted the archaeological stratigraphy had taken place prior to that of J.F. Daniel and Trench III. Likewise, no ancient disturbance by individuals attempting to free trapped survivors, recovering the remains of victims, or simply looting the ruins for valuables appears to have taken place. The only pre-modern disturbance of any kind may have been robbing walls for building material, but even this was not extensive. One of the primary reasons why some stratigraphic layers were not recognized has to do with the construction materials used in the Earthquake House. The majority of the structural components, including the upper portions of the walls and roofing elements were made of ephemeral and perishable materials (e.g. mudbrick, wood, reeds, etc.). These elements would have formed the earthquake debris layer and been a major component of the postearthquake collapse layer. As these materials were exposed to formation processes, they would have begun to disaggregate and decay to the point where these two layers became difficult to differentiate from one another.

Supra p. 25.

44

Figure 5.1: Room 1 – E-W section looking south. Reconstructed by the author from elevations recorded in AZLS from AZLS: TRIII/Room 1. The only exception is Room 8, whose tile roof acted as a horizon marker between the earthquake debris and postearthquake collapse deposits. Without such an indicator, the differences between these two layers at the time of excavation were likely so subtle that they were virtually indistinguishable. The introduction of wind and water deposited material into the gaps of these layers would have contributed significantly to their homogenous appearance.

distinct strata. The lower almost certainly represents the earthquake debris layer. The upper probably being postearthquake collapse, although its upper zone likely contained post-earthquake material as well.256 While the 1934 excavation cleared Room 1 to the wall tops in order to record its plan, it was not completely cleared of earthquake debris until Trench III was reopened in 1984.257 During re-excavation of Room 1, the University of Arizona team identified a number of loci, the majority of which consisted of slump and backfill from Daniel’s activities. Although small patches of undisturbed earthquake debris were discovered, it was only along Wall E that a significant section of undisturbed earthquake debris and post-earthquake collapse/deposition remained. This fill was designated by the Arizona Mission as Locus 012 (Fig. 5.1). Below this, the occupation surface, which represented the lowest extent of Trench III in this space, was designated as Locus 015 and Locus 017.

Another reason for the lack of locus differentiation in the rooms of the Earthquake House may be the inexperience of some field personnel with the locus system and excavation methodology in general. The assignment of loci was done very sparingly, being used primarily for large stratigraphic blocks and occasional architectural features. This resulted in a minimum number of loci being designated in many of the rooms where several more could and should have been identified. 5.2: Room Stratigraphy: All the rooms of the Earthquake House were overlain by disturbed topsoil. Below this was a thick layer of naturally deposited fill that accumulated following the site’s final abandonment. In some rooms these two layers were excavated and given locus numbers, with the topsoil normally designated as Locus 001. In others, both were stripped mechanically to what was believed to be the upper limit of the earthquake debris deposit. Locus numbers were not assigned to the stripped material, with the supposed earthquake debris layer being identified as Locus 001.

5.2.2: Room 2 258 The final occupation level in this space was not clearly 256 DFN: 1935, 39. The main description deals with the stratigraphic layers north of wall P-Σ (later Wall B); identified by the Arizona team as the colonnade of Room 15. Here, Daniel identified a transition from “darkish topsoil (rocky) to yellowish, somewhat sandy soil, in which are many rocks, and fragments fallen from the building (squared blocks, capitals, columns, etc.). This strosis (sic – the term “strosis” appears to have been used in place of “layer” or “stratum”) seems to end at 1.45 (at wall P-Σ), and is followed by a harder yellow strosis with many small pieces of plaster.” In the following paragraph, Daniel states that “This second strosis (0.50-1.45) seems to correspond to the second strosis in Σ1 (within the area of the later Room 1) though the latter is relatively clear of stones.” 257 DFN: 1935, 27. Limit of House 1, Room 1 excavations recorded on sketch plan. 258 Room 2 and Room 3 were excavated during the 1984 season using 5X5 m units based on the 10X10 m site grid. Loci from these two rooms were numbered consecutively, rather than each having their own set of locus numbers.

5.2.1: Room 1 The majority of this space was excavated as part of Trench III during the 1934 field season. Stratigraphy was only documented in a very cursory manner, so that little can be discerned about the deposits. Above the final occupation locus, Daniel appears to have discerned two 45

Figure 5.2: Room 2 and Room 3 – E-W sections looking south. Reconstructed by the author from elevations recorded in AZLS: Room 2. Hatching in Room 3 indicates portion excavated in 1985. Dashed line dividing Locus 007 indicates horizon between earthquake and post-earthquake deposits. defined, likely due to its uneven nature. This resulted in no locus number being assigned to it. A total of four depositional layers were documented during the excavation of this room (Fig. 5.2).259 The lowest, Locus 012, is the earthquake debris layer. It is recorded as having significant gaps among the stones and mudbrick that made up the bulk of its composition. This description indicates a certain degree of permeability that would permit infiltration of objects deposited following the earthquake. This situation has implications for interpreting the artifact assemblage from the room, which will be discussed in the following chapter. Locus 005, which overlay Locus 012, represents post-earthquake collapse that would have at least been partially responsible for infilling the gaps in the layer below. Like Locus 012, Locus 005 is recorded as having significant gaps throughout, which presents similar concerns for contamination by later material as well. One postearthquake deposition layer, Locus 003 and the disturbed topsoil that overlay it, Locus 001, were also identified.

do not indicate how thoroughly the earthquake debris layer was sealed in this area. This raises questions as to what degree (if any) post-earthquake contamination may have affected this room. The same post-earthquake deposition layer, Locus 003, and topsoil layer, Locus 001 that were identified in Room 2 also extended over Room 3. 5.2.4: Room 6 260 In Room 6, the post-earthquake deposition layer(s) was (were) removed mechanically down to what was believed to be the top of the earthquake debris layer. During excavation a total of three stratigraphic loci were identified (Fig. 5.3). Two of these represent final occupation deposits, with Locus 004 being the final occupation surface itself. The excavation notes record that Ruben Bullard, an environmental geologist, identified Locus 004 as being composed, at least partially, of waste from mudbrick that had been disseminated across the surface.261 This likely represents mudbrick detritus from the walls and ceiling that had spalled off and been compacted over time. The nature of the surface, however, also led Bullard to conclude that it was not subject to heavy traffic, which seems to indicate that Room 6 was not intensively utilized or was used for only a short period prior to the earthquake.

5.2.3: Room 3 In Room 3 only two relevant stratigraphic loci were discerned (Fig. 5.2). The lowest, Locus 013, represents the final occupation surface, with Locus 007 designating the earthquake debris layer. Unlike the adjacent Room 2, where earthquake debris and post-earthquake collapse loci were discerned, no such distinction was made in Room 3. However, a largely intact roof tile recovered high in Locus 007 likely marks the uppermost limit of the earthquake debris layer. The matrix above represents post-earthquake collapse, post-earthquake deposition, or some combination of both. Thus, the lower portion of Room 3, Locus 007 is contemporaneous with Room 2, Locus 012; with the upper portion corresponding to Room 2, Locus 005. Despite the presence of roof tiles, some of which were largely intact, the excavation notes

Immediately above Locus 004 was the other final occupation layer, Locus 002, which was used to designate the material culture assemblage present in the room at the time of destruction.262 Above this, Locus 001 represents the earthquake debris layer, which consisted of ashlar 260 Although Room 6 and Room 7 are in fact a single space as discussed in the previous chapter, they were excavated independently. Despite having similar stratigraphy, for clarity they will be presented separately. 261 AZLS: Room 6, Locus 004. 262 The majority of this material was recovered resting directly on Locus 004. The matrix within which the objects were found was that of Locus 001.

259 The general level for the final occupation was the lowest elevations recorded for the bottom of the earthquake debris layer Locus 012.

46

Figure 5.3: Room 6 and Room 7 – N-S sections looking west. Reconstructed by the author from elevations recorded in AZLS: Room 6 and AZLS: Room 7. blocks, stone cobbles, mud-brick remains, tile fragments (some with plaster adhering), plaster chunks, and slope wash.263 The identification of Locus 001 as being composed solely of earthquake debris is incorrect based on the stratigraphic evidence from other rooms in the structure. Most likely, what the excavators designated as Locus 001 was composed of several different layers, with the earthquake debris forming the lowest part of the deposit. This was likely overlain by a post-earthquake collapse layer and post-earthquake deposition layers created through a combination of natural (e.g. wind, slope wash) and cultural (e.g. dumping) formation processes. A comparison between the residual (i.e. broken) ceramic material from the upper part of Locus 001 in Room 6, and re-deposited material from the same stratigraphic layer in Room 7 (see below), appears to confirm that the upper part of Locus 001 was largely post-earthquake in nature.264

record that the upper portion of Locus 001 consisted of stone rubble with tile mixed in, with another layer immediately below composed of smaller stone rubble with some tile. These observations, when compared with the stratigraphy from other rooms, indicate that the upper portion of Locus 001 likely represents post-earthquake deposition. Small finds, including a coin of Probus (AD 276-282)265 were present throughout Locus 001, but it is noted that the bulk of these objects were found predominantly in the upper elevations, a phenomenon nearly identical to that observed in the adjacent Room 6. Immediately below the material-rich post-earthquake deposition was an earth matrix that contained fewer artifacts and likely represents the post-earthquake collapse/deposition layer. Below this was a layer of mudbrick detritus, which is the earthquake debris layer created by the collapse of the mudbrick walls along with the packed mud and wood of the roof. The occupation surface (Locus 002) possessed the same general characteristics as that observed in Room 6, although it was noted as being quite uneven and less compacted than the adjoining room. This observation is noteworthy and, if accurate, likely indicates that Room 7 was used even less frequently than Room 6.266

5.2.5: Room 7 Like Room 6, the post-earthquake deposition layer(s) was (were) removed mechanically to the supposed top of the earthquake debris layer. Only two loci were formally identified by the Arizona team in Room 7; the so-called “debris” layer of Locus 001 and the occupational surface designated Locus 002 (Fig. 5.3). The excavation notes

5.2.6: Room 8 As in Room 6 and Room 7, the post-earthquake deposition layers were removed mechanically to the supposed earthquake debris layer. Below this, four relevant loci were identified during the Room 8 excavation (Fig. 5.4). Two are final occupation loci. Locus 004 represents the final occupation surface of the room. Upon this a small, roughly circular deposit of fine sand was discovered in the southern part of the room, which was discovered as Locus 005. Above these, two “debris” layers were distinguished, the lower, Locus 002, and the upper Locus 001. Locus 002 represents the earthquake debris layer. Its upper limit was defined by a layer of broken tiles created by the roof collapse. Although this provides a distinct horizon marker between the earthquake debris and post-earthquake deposits, the size of the room prevented Locus 002 from being completely sealed. The excavation notes record numerous gaps in the tile layer, particularly around the perimeter of the room where no tiles were recovered. Such voids and gaps create potential conduits for material deposited post-

263

AZLS: Room 7, Locus 001. The ceramic from Room 6, Locus 001 consists of vessels that display an overall low level of completeness, and a high level of brokenness. This material provides an average EVREP of 103.5. Of these, a total average EVREP of 59 (46 definitive; 13 probable) appear to represent post-earthquake deposition, with 21 (16.5 definitive; 4.5 probable) representing material belonging to the final occupation or earthquake debris. Unfortunately, due to the lack of diligent basket elevation recording, an additional average EVREP of 23.5 are of indeterminate status, generally occupying a level where the transition between the true earthquake debris layer and post-earthquake material is believed to have been. The level of this transition was based on the uppermost limit of material definitively attributable to the in use assemblage (68.50 ASL). Given their general elevation, it is highly probable that the majority of these ambiguous vessels were deposited during the post-earthquake period. Using only the base numbers (59 and 21), the average EVREP in the probable zone of post-earthquake deposition is 2.8 times that of the final occupation and earthquake debris layers. If the unknown vessels are divided equally between the upper and lower zones, the concentration is reduced to 2.08 times greater. This increased concentration in the upper zone appears to indicate that following the earthquake, discard of ceramic material occurred in this area at a rate greater than would be expected through natural formation processes. This cultural deposition does not seem to have been as intense as in other rooms of the structure and likely represents a gradual buildup over the course of time rather than organized discard/dumping activity. The increased levels of material culture appear similar to conditions observed in Room 7, Locus 001. 264

265 266

47

AZLS: Room 7, Locus 001. Identified by E. Lane. AZLS: Room 7. Observations of Reuben Bullard.

Figure 5.4: Room 8 – N-S section looking east. Reconstructed by the author from elevations recorded in from AZLS: Room 8. layers of some sort. The formation of these loci ultimately caused the ground level in Room 11 to be raised substantially. This necessitated placing a vertical stone barrier across the width of the Wall N doorway to keep this material from spilling into Room 8. A compact layer, identified as Locus 011 seems to have formed some type of occupation surface within Locus 013; both likely belong to the final occupation phase of the structure. Over Locus 011 was another fill deposit, Locus 009, which may have been created as a result of discard processes that occurred during the final occupation. While the deposits of Room 11 overall remain relatively unknown, the small vignette provided by the western probe indicates the area was filled-in over a period of time with ad hoc surfaces within debris layers that may represent “paths” through this accumulated material. However, this hypothesis needs to be investigated further in order to determine its validity and how this area developed over its use life.

earthquake as part of Locus 001. 5.2.7: Room 11 Much of the excavation of Room 11 cannot be reconstructed since only a portion of the locus sheets could be located in the excavation archives. The postearthquake deposition layers were removed mechanically down to the top of the earthquake debris layer. The two main loci identified were Locus 001, the earthquake debris layer, and Locus 002, the final occupation surface. Earlier deposits in Room 11 remain largely unexcavated and were only exposed as the result of a deep probe on the western edge of the space. Because of the limited view obtained by this “keyhole” operation, these preearthquake layers can only be generally discussed, as their character over the remaining space is unknown. The earliest stratigraphic levels identified represent virgin soil and were designated Locus 017 (Fig. 5.5). Above this was a compacted soil layer, Locus 016 that was likely the first occupational surface of this space.267 A series of strata, (Locus 015, Locus 014, and Locus 013) overlay Locus 016, all of which appear to represent debris and/or fill

5.2.8: Room 12A A total of five stratigraphic loci were recorded within

Figure 5.5: Room 11 – E-W section looking north depicting the relationship between the doorway in Wall N and the strata recorded in the western part of Room 11. The elevation of the threshold block is the horizontal surface of Wall N. Reconstructed by the author from elevations recorded in AZLS: Room 8 and AZLS: Room 11. 267

GFR: 1985, 44.

48

Figure 5.6: Room 12A and Room 14 – N-S sections looking west. Reconstructed by the author from elevations recorded in AZLS: Room 12 and AZLS: Room 14. fragments.270 In conjunction with this surface, Locus 005 is recorded in the excavation notes as a thin layer of ash with some small “chunks” of mudbrick. This was interpreted by the excavator as being deposited as a result of periodic cleaning and maintenance activity carried out on the oven (Locus 003) and also perhaps the scattering of ash at the time this feature was destroyed during the earthquake.271 Locus 006 is recorded as being at the same general level as Locus 004, but in contrast to the hard packed surface, it was composed of loose soil and confined to small strips on the east and south sides of the oven along Wall D and Wall Q. This locus contained a significant amount of material culture, but its formation and significance, which will be explored in the following chapters, was not completely understood at the time of excavation.272

Room 12A, of which three are relevant to the present study (Figs. 5.6, 5.7). The lowest, Locus 006, was a thin layer (ca. 6 cm) of fine silt covering the bottom of the cistern. This represents suspended particles that settled out of the water column during the installation’s lifetime. This deposit almost certainly pre-dates the final occupation. Resting on Locus 006, but confined to the southwest corner was Locus 005, a thick deposit of fine ashy soil that contained a significant amount of broken ceramic and ecofacts (predominantly animal bone and charcoal fragments). The general relationship of these two layers led the excavators to conclude that Locus 005 was formed by discard of household waste (primarily hearth debris) after the cistern had gone out of use.268 Covering both of these was Locus 003, which was identified by the excavators as the earthquake debris layer. Locus 003 was composed of a significant number of wall stones and hydraulic plaster that had detached from the sides of the cistern within a matrix of loose, light-colored soil.269 The material from within this locus is consistent with other post-earthquake fill layers and based on its characteristics, Locus 003 is likely a hybrid layer, with the lower portion composed of earthquake debris and the upper post-earthquake collapse and/or deposition. The quantity of material recovered from Locus 003 seems to indicate sporadic rather than organized deposition, which is consistent with postearthquake discard activity observed in other areas of the structure. The final two layers are Locus 002, which represents definitive post-earthquake deposition and Locus 001, a layer of disturbed topsoil.

Overlaying all of these was Locus 002, a layer of decayed mudbrick and stone rubble that represents the earthquake debris deposit. Locus 002 was excavated as a homogenous layer, but based on evidence from other areas it seems probable that only its lower portion is actual earthquake debris. The upper portions were likely formed by ongoing post-earthquake collapse as well as natural and cultural post-earthquake deposition. This layer was covered by Locus 001, composed of postearthquake deposition and disturbed topsoil. 5.2.10: Room 15 and Room 18 Room 15 is one of the most poorly documented spaces in the entire structure. This is unfortunate since the discernment of loci would have been invaluable in helping to reconstruct activities that occurred here. As it

5.2.9: Room 14 Of the loci identified in Room 14, the most relevant are Locus 004, Locus 005, and Locus 006. Together these form the final occupation deposit within this space (Fig. 5.6). Locus 004 designates the occupation surface of the room. It was uneven and composed of compact earth containing ash from the oven as well as small ceramic

268 269

270

AZLS: Room 14, Locus 004. This represents incidentally deposited material incorporated into the floor matrix through use over time. 271 A general layer of ash is recorded throughout the room in the excavation notes, with some deeper pockets occurring due to the uneven nature of the floor. The largest concentration of this material appears to have been in the northwest corner. 272 AZLS: Room 14, Locus 006. The excavation notes pose a number of possibilities for the formation of this locus including “an animal burrow, a drainage system, or just an area that was not frequently walked on.”

Molinari et al. 1988. Locus 004 designates the fill in the eastern half of the room.

49

stands, the loci descriptions are very brief and offer little or no insight into their formation. Like the remainder of the structure, Room 15 and Room 18 would have been overlain by disturbed topsoil and post-earthquake deposition layers. Although many loci were recorded in Room 15, the majority are identified simply as “earthquake debris layer” with no explanation of their location and composition. As was the case with Room 11, an examination of the Room 15 and Room 18 loci provides evidence for cleanup and/or construction activities that may have occurred following an earlier seismic event (Fig. 4.18). In the southeast corner of Room 15, a deposit of hard lime cement plaster, Locus 014, was identified that contained small rounded pebbles, medium sized cobbles, and fragments of tile and coarse pottery throughout its matrix.273 The plans and descriptions indicate this material formed a rough mound in the eastern end of the portico that extended to the north, covering the courtyard paving as far as the “basin” in the northeast corner. Two superimposed compact earth occupation surfaces in the portico, Locus 007 and Locus 009, partially overlay Locus 014. A small hemispherical oven against Wall V is also recorded as cutting through Locus 014.274 The deposits of Room 15 and Room 18 are poorly understood, a situation that is unlikely to improve since both spaces were excavated completely.

than in the larger rooms. Three loci were identified in Room 19 (Fig. 5.7), all of which were interpreted as “a fairly homogenous deposit of [earth] quake debris.”275 The lowest was Locus 004, composed of loose soil, moderately-sized stones and some larger architectural blocks.276 A modest amount of material culture was recovered within this layer; primarily broken ceramics and a few plaster fragments. The overall character of this stratum as well as the material evidence indicates that the lower portion of Locus 004 represents the final occupation deposit and the upper the earthquake debris layer. Above this, Locus 003 was a layer of compact soil throughout the space although some smaller patches of loose soil were also observed. The notes make it clear that no lamination was present in this layer and that it was not an occupational surface. It contained very little material culture and based on its description and position, Locus 003 likely represents a combination of earthquake debris and post-earthquake collapse. Locus 002 was composed of large stones, including some architectural blocks within a matrix of loose soil and very little material culture. Rather than being contemporary with the earthquake as asserted in the excavation notes, Locus 002 clearly represents a layer formed by continued postearthquake collapse and deposition. The increased stone content compared to the strata below may reflect an accelerated degradation of the architectural remains, possibly coinciding with the site’s reoccupation. The uppermost layer, Locus 001, was composed of postearthquake deposition and disturbed topsoil.

5.2.11: Room 19 Because of the small and confined nature of this space, the identification of strata may have been more difficult

Figure 5.7: Room 12 and Room 19 – E-W sections looking south. Reconstructed by the author from elevations recorded in AZLS: Room 12 and AZLS: Room 19. 273 AZLS: Room 15, Locus 014. No size estimates are provided for the material incorporated in the matrix. 274 AZLS: Room 15, Locus 014.

275 276

50

This observation is recorded in AZLS: Room 19, Locus 004. AZLS: Room 19, Locus 004.

5.2.12: Room 20 In this space, the packed earth occupational surface was designated Locus 005 (Fig. 5.8). On the northern edge of the room, a concentration of what was identified as possible debris clean-up similar to Room 15, Locus 014 was assigned as Locus 006. This locus was overlain by Wall AE, which separated Room 20 from Room 25 indicating that it predates the space’s division. Locus 004, which overlay Locus 005, contained a significant amount of decomposed mudbrick and stone. It was formed by the collapse of the upper walls and roof during the earthquake and thus represents the earthquake debris layer. The most compelling evidence for this is the fact that the remains of three victims, two young adults and one infant were discovered within this stratigraphic unit.277 Above this, a layer composed of loose, yellow soil was designated as Locus 003. Unlike the stratum below, Locus 003 was noticeably devoid of material culture and almost certainly represents a post-earthquake deposition layer created principally (if not wholly) through natural formation processes when the site was abandoned. Overlying this was Locus 002, which contained a significant amount of material culture including large carved architectural blocks, fresco fragments, worked stone, and a large quantity of ceramic. Based on the concentration and variety of the artifacts recovered, it is probable that this locus represents a post-earthquake deposition layer formed predominantly by deliberate dumping, possibly from clean up following reoccupation of the site. Thus, rather than being roughly contemporaneous earthquake debris deposits as the excavators asserted, the material evidence indicates that Locus 002, Locus 003, and Locus 004 represent strata formed during three distinct phases of the site: the earthquake; the post-earthquake abandonment; and the reoccupation.

occupation surface was defined as Locus 003 and the room fill, Locus 002, was excavated as a homogenous earthquake destruction layer (Fig. 5.8). This lack of stratigraphic differentiation and recording was a problematic aspect in analyzing the assemblage from this room. Another complicating factor was that during its excavation, elevations appear to have been taken only sporadically or inaccurately so that the precise horizontal and vertical contexts for many probable in use artifacts cannot be reconstructed with any degree of confidence. Material recovered from the upper part of Locus 002 confirms that the dumping observed in Room 20 extended at least partially into Room 25 and Room 27 in the adjacent Building B. Evidence for this is based on crossjoining pottery fragments from several different vessels, one of which, a Late Roman C, Form 1 dish, has significant portions originating from each of these three spaces. Although it is possible to discern the presence of this dump deposit in Room 25, the extent to which this space was filled with intentionally discarded material is difficult to determine. As a result, the only objects that can be assigned to the in use and associated assemblages are the completely or largely reconstructible vessels recovered from the occupation surface itself. Locus 002 was overlain by Locus 001, a layer of post-earthquake deposition and disturbed topsoil. The stratigraphic units and their relationships outlined in the previous sections constitute the archaeological context from which the Earthquake House artifact assemblage was recovered. Its full comprehension is a critical element in this study for accurately interpreting the probable use status of the objects. Only now, with the overall archaeological context of the assemblage appreciated (architectural and stratigraphic), can the inquiry turn to a detailed examination of the artifacts recovered from each room.

5.2.13: Room 25 Unlike Room 20, in Room 25 no stratigraphic distinctions were made or recognized in the fill of the room. The

Figure 5.8: Room 20 and Room 25 – E-W sections looking north. Reconstructed by the author from elevations recorded in AZLS: Room 20 and AZLS: Room 25. 277

These human remains were named the “Christian family” (Soren 1988: 131-4). Although originally the victims would have been lying on the floor (Locus 005), the remains were encapsulated by the matrix of Locus 004.

51

Chapter 6: Room Assemblages In Chapter 3, a methodological framework for evaluating the Earthquake House artifact assemblage, particularly the ceramic and glass assemblages was outlined. This process assigned a particular object into one of two groups: those that had a high probability of being in use at the time the structure was destroyed (the life assemblage); and those that were not (the death assemblage). Each object’s assignment was based solely on its condition and did not take into account their archaeological context, a key component in their accurate interpretation. Having examined the contextual elements of architecture and stratigraphy in Chapter 4 and Chapter 5, it is now possible to fully evaluate the artifacts and place them into a typological system that will form the basis of their analysis.

up of ceramic for use as a component of wall construction, etc.). 280 II.   

6.1: Terminology Because this study utilizes specialized vocabulary in discussing behaviors and practices impacting the assemblage, particularly the ceramics, it is necessary to provide brief definitions for the terms used. These can be divided into two groups: types of use(s) an object could have prior to discard; and categories of discard into which an object could be placed:278 I. 









Terms Related to Categories of Discard:281 Primary Discard: Deposition of waste material at the site of its creation and/or use. Secondary Discard: Deposition of waste material away from the location where it was created and/or used. Provisional Discard: The collection and temporary caching of waste material, normally away from its site of generation. This can be done with the possibility that the material may be reused and/or recycled at a later date. Definitive Discard: Material that has been placed in its location of permanent deposition (e.g. landfill).

6.2: Artifact Typology: An examination of both the condition (e.g. completeness, brokenness, evidence for use, evidence for maintenance, etc.) and context (e.g. elevation, location in room, proximity to fixtures and other objects, etc.) of the Earthquake House artifacts permits them to be ascribed to one of four classes:282  Class 1: An object that was either in primary or secondary use at the time the structure was destroyed (life assemblage).  Class 2: An object located inside the house that was no longer in use and/or possibly awaiting reuse at the time the structure was destroyed (death assemblage).  Class 3: An object incorporated/recycled into the architecture of the structure (e.g. sherds used as chinking, sherds incorporated into mudbrick or mud roofing, etc.).  Class 4: An object deposited post-destruction that is unrelated to the architecture or final phase of the structure’s occupation.

Terms Related to the Use of the Object Prior to Discard: Primary Use: Use of an object for the purpose it is traditionally associated with (e.g. amphora used as a transport container, normally for liquid commodities; cookpot used in the process of food preparation, etc.). Secondary Use/Reuse: Use of an object for a purpose other than it is traditionally associated with (e.g. amphora used as a fixed storage container for both liquid and non-liquid commodities; amphora necks used as funnels;279 cookpots used as chamber pots, etc.). This may or may not involve modification of the object. Maintenance: A term indicating manipulation of an object so that it can continue to function either in a primary use or secondary use/reuse capacity. This includes such diverse activities as cleaning objects following use, repairing damage, and removal of fragments and/or portions of the object. Recycling: The break-up of objects or parts of objects so they can be used as raw materials in a manufacturing process (e.g. melting down of glass or metal; breaking

Class 4 objects provide evidence for post-destruction behaviors and formation processes, predominantly in the form of dumping that occurred at the site. Class 3 objects encountered during the study were larger sherds employed as chinking between wall stones or small sherds recycled into building materials (e.g. mudbrick) that were virtually impossible to differentiate from Class 4 material. Examples of Class 3 artifacts are low-density sherd concentrations or small deliberately280 Darnay and Franklin 1972, 2; Schiffer 1996, 29-30; Lister and Lister 1981. 281 Primary and Secondary Discard cf. Schiffer 1996, 58-64; Provisional Discard cf. Deal 1985, 253-9; Kamp 1991, 25; Schiffer 1996, 66. 282 In cases where the condition and/or context of the artifact are problematic, attribution to more than one class is possible.

278 The general definitions of terms applied to the Earthquake House assemblage have been adapted from those formulated by Peña (2007, 10) in his discussion of the ceramic life cycle. 279 Slane 2004; Soren and James 1988, 141. Berry 1997, 192; Deal and Hagstrom 1995.

52

squared/worked amphora sherds. Class 3 and Class 4 artifacts were not of direct interest to this study and were excluded from the overall analysis.

personal and luxury items (e.g. a marble tabletop, carved bone hairpins, a bronze lampstand, etc.) and coins, whose findspots were inconsistent with the utilitarian architectural context within which they were found.

Class 1 and Class 2 objects form the basis of this study. Both have the potential to answer questions regarding aspects of daily life including artifact use, repair, and reuse, identification of room function, the locations and nature of activities within rooms, aspects of household refuse disposal (the waste stream), and the identification of potential residential groups.

The Earthquake House artifacts assigned to Class 2 are predominantly ceramics with a moderate to low level of completeness combined with a moderate to high level of brokenness. While in some cases portions of Class 2 vessels may be reconstructible, not enough was present to permit them to be in either primary or secondary use. Because of their “broken” state, Class 2 objects were interpreted as waste material that was either in discard (secondary or provisional) or in storage awaiting some type of reuse and/or recycling at the time of the earthquake. Instances where joining fragments of a vessel were recovered from multiple areas in a single room or more than one room in the structure also resulted in Class 2 identification. These artifacts were recovered predominantly from contexts on or near the final occupation surface. In many cases, they formed deposits close to the walls as well as small and/or isolated spaces within the structure.

Class 1 objects can be further divided into four subclasses:  Class 1a: Objects on/worn by the victims at the time of death recovered in close association with the human remains.  Class 1b: Objects in use on the structure’s ground floor on or near the occupational surface that were recovered more or less in situ.  Class 1c: Objects in use on the structure’s ground floor stored on shelving or suspended from the roof that fell during the earthquake or shortly afterward.  Class 1d: Objects in use on the structure’s second story that fell at the time of the earthquake or subsequently as the remains degraded over time.

6.3: Possible Factors Affecting Vessel Completeness Since completeness is one of the main criteria for determining to what class a particular vessel should be assigned, it was necessary to examine factors that could contribute to its incompleteness other than identification as waste material. This insured that each vessel was evaluated and placed within the proper typological category to provide the most accurate picture of the assemblage possible.

Class 1a consists predominantly of personal objects that can be inferred as being worn by the individuals at the time of the earthquake (e.g. rings, hairpins). Class 1b represents objects in use at the time the structure was destroyed such as ceramic vessels with a high degree of completeness (e.g. pithoi, amphorae, etc.). Such vessels could have been in primary or secondary use, but it was not always possible to determine to which category a particular vessel should be assigned based on the archaeological evidence.

One factor that could affect overall completeness was deliberately removing a portion of the vessel, normally as a maintenance process to permit continued use or so that certain elements could be reused. In the former cases, specific parts were missing, but the vessel was highly complete overall indicating it could have functioned as part of the life assemblage. The most common maintenance process observed was removal of the rim, which was normally broken off quite evenly around the circumference of the vessel. In these cases, the rim was usually the only element missing, with the completeness and context indicating that the remainder continued in use. This typically resulted in the vessel being assigned to Class 1b.

Class 1c objects were recovered on or near the final occupation surface, although they could also occur higher up within the earthquake debris layer. The primary criteria in differentiating these objects from those of Class 1b are their size and material; with Class 1C tending to be smaller and more fragile (e.g. glass, metal, fineware ceramics). Based on these criteria, it is likely that items such as these were stored above the level of the final occupation surface (e.g. on a shelf) to protect them from accidental damage/breakage.

In some cases, instances of reuse saw the removal of significant portions of the vessel beyond what would be expected for maintenance; for example the complete removal of the neck, handles, and shoulders of amphorae. It is possible that the lower portion of the vessel could have been used as a storage container of some type, thus leading it to be identified as Class 1b. Likewise, the upper portion could have been reused (e.g. funnel), which resulted in it being given a Class 1b identification as well. Depending on their completeness and find context, however, these same objects could also be designated as

Class 1d artifacts are those believed to have originally been located on the structure’s upper story. They were deposited either by collapse during the earthquake or as the structure continued to degrade over time. Objects in this category were generally difficult to identify since they could occupy a number of contexts depending on their size (e.g. coins), ranging from near the final occupation surface to levels significantly above it. On the whole, artifacts assigned to this class are predominantly 53

Class 2. In instances where both identifications are equally plausible, both are presented.

where an object has been previously published283 or could not be photographed, illustration is omitted.

Other factors potentially impacting vessel completeness were its fabric characteristics and the likelihood of differential recovery and/or discard of fragments during excavation. Friable and/or brittle fabrics are more likely to be affected by formation processes, which may cause increased degradation over time. This often resulted in pronounced voids in the vessel walls, raising questions as to whether the vessel should be identified as Class 1 or Class 2.

6.4.1: Room 1 The 1934 Trench III excavation mainly focused on the two victims recovered in the center of the room (Fig. 6.1). Despite their intact nature, the number of Class 1a artifacts associated with the remains appears minimal. From one skeleton (identified as R in the excavation notes) two rings, one gold (S41-1934)284 and one copper alloy were recovered, both of which are recorded as being on the same finger.285 The second set of remains (identified only as L in the excavation notes) produced a bone hairpin (S44-1934) and a hexagonal bead of green paste (S43-1934). Both were recovered under the skull.286

In Trench III, much of the ceramic was ultimately returned to the field and discarded. Because of this, vessels recovered by the Arizona Mission near the Trench III boundaries may exhibit levels of completeness lower than what they actually possessed. In conjunction with this is the possibility of fragments being broken and/or discarded inadvertently by excavators in the field. A number of factors could have contributed to this including size, friability (causing them to break apart during recovery), and the rapid pace at which the Arizona excavation progressed. The latter factor, especially from the second season onward, adversely affected identifying all joining vessel fragments during conservation and reconstruction efforts. Whenever possible, these missed fragments were identified by the present study and considered in evaluating the vessel’s overall completeness. They were also examined as possible indicators that the vessel was broken and scattered prior to the earthquake.

Two additional metal objects, possibly of a personal nature, were found in the southern portion of the room during the 1984 excavation. The first is a copper alloy ring (M5), which is either a true “key” ring or a ring reused as a key. It is Class 1b or Class 1c.287 The second, a small irregularly-shaped copper alloy object (M26) (Fig. 6.2) that was possibly part of a buckle or clasp is either Class 1b or Class 1c. For both, the latter classification is the most likely. Three vessels from Room 1 can be identified as Class 1b: Vessel P1, Vessel P134-1934, and a “cooking vessel complete with lid (UR-1934).”288 Of these, only P1 (Fig. 6.3), a small two-handled pithos in the southwest corner of the room could be located and examined. Vessel P134-1934 is recorded in the Daniel field notebook and a Trench III field photograph as being found near the center of the room close to victim R. Daniel’s description of the vessel refers to it as a “large jug with [a] base ring and one handle to the shoulder.”289 In the same entry, he notes that both the neck and handle of the vessel were missing at the time of excavation. Given the level of completeness observable in the field photograph, the missing elements suggest that the vessel was likely broken prior to the earthquake but remained complete enough to permit continued use. Such use of highly complete “broken” vessels appears to have been a common practice, at least in the Earthquake House where numerous examples of this behavior were documented.

6.4: Room Assemblages The previous discussion demonstrated that determining an artifact’s status at the time the Earthquake House was destroyed is a complex and nuanced process that must be undertaken on a case by case basis. However, the use typology and evaluation protocols developed for this analysis provide a means of organizing and understanding the individual room assemblages to effectively investigate aspects of artifact use, reuse, discard, patterns of use, and room function. The following discussion is accompanied by two types of figures: AutoCAD plans of each room showing the findspots of relevant objects; and images of the artifacts themselves. On the plans, objects whose precise findspots are unknown, but can be generally reconstructed are enclosed within a square/rectangular frame. The artifact images consist predominantly of digital photographs taken by the author during analysis. While all objects from the in use and associated assemblages are discussed and presented on the plan, for simplicity only images of the most relevant objects from each room are included. When artifacts could not be photographed directly due to inaccessibility (e.g. on display in museum, unable to be located, etc.), photographs and/or scale drawings found in previously published material or the excavation archives were used to record the relevant object data. In cases

283

References for previously published images included in footnotes. Original registry number recorded as S41. The additional designation of “1934” was added by the present study to differentiate finds recovered during the University Museum Mission from those recovered by the Arizona Mission. 285 DFN: 6/6/1934, 28. The notes state that the rings were removed from what was believed to be the left hand of the victim, with the copper alloy ring on the outside, and the gold ring on the inside. 286 DFN: 6/6/1934, 28. Recorded as found by “B.H.H.” This is likely Bert Hodge Hill director of the University Museum Mission to Kourion. The hairpin was located in the storerooms of the Episkopi Museum by David Soren and Christofis Polycarpou in 1980 and identified by Soren as probably belonging to a woman (Soren and Davis 1985, 295). 287 Previously published in Soren and James 1988, 84. 288 DFN: Winter 1935, 29. This vessel could not be located for examination in the Kourion Museum storerooms. No drawings of the vessel were made in the notebook. 289 DFN: 6/6/1934, 28. 284

54

Figure 6.1: Room 1 plan. Reconstructed by the author from DFN: 6/6/1934, 27 and AZLS: TRIII/Room 1. 55

Class 1b or Class 1c.294 According to the Room 1 sketch plan in the Daniel notebook, this lamp was recovered directly in front of the doorway, close to Vessel P1341934.

Three other Room 1 vessels are identified as Class 1b or Class 2: Vessel P22/28/153, Vessel P221, and Vessel P135-1934. The remains of Vessel P22/28/153 were found adjacent to Vessel P1 and consist of the joining rim and upper shoulder fragments of a medium-sized pithos. The excavation notes and field photographs clearly indicate that the rim was imbedded in the earthen occupation surface, raising questions about its status at the time of the earthquake. In addition, Vessel P22/28/153 was found close to the Trench III boundary,290 presenting the possibility that fragments may have been recovered and discarded during the earlier excavation. If this object is 1b, it was almost certainly in some type of reuse context, possibly as a vessel stand.

The ceramic assemblage is enhanced by a copper alloy jug (UR-1934) found in front of the doorway near victim R. Based on the 1934 field photograph, it possesses a single handle and biconically shaped body similar to others found elsewhere in the structure.295 While it is possible that this object could be Class 1b, as a metal vessel it would have had a higher probability of being stored on a shelf making Class 1c identification more likely. While its findspot does not reflect its use or storage location, the lack of any drinking vessels recovered in Room 1 indicates that UR-1934 was probably stored rather than used in this space.

The two other vessels are amphorae from the east side of the room. Vessel P221 is a Kourion Class IV/Late Roman 4 (KT IV/LR4) “Gaza” amphora located in the southeast corner (Fig. 6.4).291 The base was intact and the entire rim is accounted for. Although a number of joining and non-joining fragments of this vessel were recovered, they are insufficient for a full reconstruction. Like Vessel P22/28/153, the proximity of Vessel P221 to Trench III may have led to the recovery and discard of fragments during its excavation. If this vessel was Class 1b, it was used for the storage of a solid rather than a liquid commodity. This conclusion is based on the presence of a small hole (ca. 0.05 cm) drilled into the vessel’s bottom, likely to facilitate emptying the original contents.292 The hole does not preserve evidence of repair during antiquity, which would have rendered it incapable of holding liquid.

Only two stone objects were recovered from Room 1, both within the earthquake debris layer (Locus 012). Based on this, they are identified as Class 1c, Class 3, or Class 4 objects. The first, S5, is possibly a small column drum although a suspensura is also a plausible identification (Fig. 6.5). While it likely represents some type of architectural element, S5 was not being used in this role when the structure was destroyed. Sooting on one surface indicates at one time the object was associated with small-scale burning suitable for a variety of purposes (e.g. warmth, cooking, etc.).296 It is small and light enough to have been stored on a shelf, so Class 1c identification is possible but not definitive. The second stone object, S6, is a fragment of dark, finegrained stone likely originating from the Troodos massif. It is broken on two sides, indicating it was once part of a larger object, most likely a quern. While its present size is not conducive to this function, it could have been used as a hand grinder if it does, in fact, belong to the life assemblage. If this is the case, it should be identified as Class 1c. If not, it is Class 3 or Class 4.

The second amphora, Vessel P135-1934, was located along Wall A immediately north of the doorway. It is recorded in Daniel’s notes as “the bottom of a large pointed amphora” with a brown fabric and yellow slip.293 The description seems to indicate this portion of the vessel was recovered intact, presenting the possibility that P135-1934 was Class 1b with additional fragments overlooked and/or discarded. Since this cannot be confirmed, Class 2 identification must also be considered.

The final component of the Room 1 assemblage is a significant number of copper alloy coins recovered from the occupation surface (Locus 015 & Locus 017) and earthquake destruction debris layer (Locus 012). The known findspots reveal no concentrations suggesting a single origination point, which would be expected if they were contained in a vessel or sack. A total of 57 coins were collected by the two excavations; the majority by the University of Arizona during the 1984 season. It is unknown how many coins were discarded during the Trench III excavation, as the notebook only records those found on the occupation surface or close to

During the course of both excavations, several lamps were found in Room 1. The majority of these are identified as Class 2 or Class 4 based on their lowmoderate levels of completeness. Only one, L47-1934, is 290

During the first season especially, the Arizona team placed a great deal of emphasis on the reconstruction of vessels. Because of this, if any sherds belonging to these vessels were recovered, it would have been noted and they would have been attached. 291 The Kourion amphora typology used is that published by Williams (1987), which includes petrographic descriptions. Drawings of each type with estimates of capacity are located in Appendix B. See Keay 1984 for the Late Roman amphora typology. For a study of Late Roman 4 amphorae specifically, see Zemer 1978. 292 Bonifay (2004, 467-8) lists four different methods of opening amphorae that involved either holing or removal of the lowest part of the vessel. The first, which involves “the creation of one or two small holes ca. 1-2 cm in diameter,” appears to be the technique utilized on Vessel P221. 293 DFN: 6/6/1934, 28. This vessel was unable to be located in the museum storeroom. No determination of typology or assessment of probable status at the time of the earthquake was able to be carried out.

294 This lamp was unable to be located in the Episkopi museum storerooms. A rough description is present in the Room 1 notes of the Kourion archives obtained from David Soren. It is described as “nearly complete,” with “traces of use.” 295 Specific examples will be presented in subsequent room assemblage discussions. 296 Based on other examples found at Kourion, this object does not appear to be an altar although this function cannot be excluded.

56

Figure 6.2: M26 – Unidentified CA object, possibly part of a buckle or clasp.

Figure 6.4: P221 – KT IV/LR4 amphora. Note small hole drilled in base to facilitate emptying.

Figure 6.3: P1 – Two handled pithos.

Figure 6.5: S5 – Possible column drum or suspensura. Note traces of burning on surface of object denoting likely reuse. 57

the human remains. 297 In addition, a number of coins from the Arizona excavation were recovered during screening so their exact findspots cannot be determined. Of these 57 coins, 24 (42.1%) were from the occupational surface, with the remaining 33 (57.89%) found throughout the earthquake debris layer. This distribution is noteworthy, indicating the bulk of the coins should be identified as Class 1c, Class 1d or Class 4. It is likely that the majority of those on or near the occupation surface and human remains should be identified as Class 1c or Class 1d that infiltrated into the lower stratum through gaps in the debris layer. The possibility must be acknowledged, however, that at least a small percentage of the coins may represent one or more of the following: Class 1a, Class 1b, or Class 2 (through incidental loss).

likely formed elements of a rope or leather bridle to which the chain was attached. 299 A group of four metal objects was located on the west side of the room against Wall D. At least two of these, a fragmentary iron ring (M112) and an iron curb bit (M117)300 (Fig. 6.10) were part of the mule tack assemblage. They were likely hung, possibly on a nail, on Wall D and are Class 1c. The other two objects, a heavy teardrop-shaped iron ring (M113) and a rectangular object with both iron and (apparently) lead components (M116) (Fig. 6.11) may belong to the tack assemblage, but what role they fulfilled is unknown. If they are tack components, they are Class 1c; if not, they are likely Class 1b or Class 2 objects.301 Because the space was not utilized for human habitation during its final phase, the recovery of two luxury objects southwest of the trough is noteworthy. The first is a three footed copper alloy lampstand in the shape of a spirally fluted (spiraliform) Corinthian column (M69). Three dolphin heads spring from the capital to form a tripod supporting a separately attached dish lamp.302 The second is an ornately carved marble tabletop (Fig. 6.21).303 Its completeness level indicates it was intact at the time of the earthquake. The absence of any supporting elements (i.e. legs) suggests that it rested on some type of wooden support. The presence of these two objects in a utilitarian space is puzzling, since their location essentially blocked the narrow opening between Wall D and the trough. It is highly unlikely that these objects were placed intentionally as a barrier to prevent the mule from entering the south side of the room, since they would have been largely ineffective against an animal of such size and strength. The most probable explanation for these objects being in this space and their findspots is that they represent material from an upper story that fell when the floor collapsed during the earthquake. Given the contextual evidence, both the lampstand and the table can

6.4.2: Room 2 Although one human victim, a female aged ca. 13 years,298 was recovered from this space, the only human Class 1a object is a bone hair pin (BI3 and BI4) (Fig. 6.6). A second carved bone pin fragment (BI2) (Fig. 6.7) was likely either Class 1d, Class 2 or Class 4 that infiltrated into the earthquake debris layer. Other fragments of an intricately carved bone pin, BI5 (Fig. 6.8), and what may be bone inlay carved with a bead and reel motif (BI6 and BI7) (Fig. 6.9) were recovered in the post-earthquake collapse layer (Locus 005). Both represent either Class 1d or Class 4 objects. The Room 2 ceramic assemblage is entirely Class 2, containing no reconstructible or partially reconstructible vessels. The majority of the glass from this room was unable to be located, but based on descriptions in the notes and registries it also appears to be predominantly Class 2. The only exception is the joining and non-joining fragments of a large glass jar (G104) from the postearthquake collapse layer (Locus 005). The fragments that were located account for less than 50% of the total object, but based on the quantities in the registry, its overall level of completeness is significantly higher. The find context of the vessel indicates a probable Class 1d identification, with Class 4 being a possible but less likely option.

299

Soren and James 1988, 88; Davis and Soren 1985, 298-99. M117 appears to be a textbook example of a Manning Type 1 Curb Bit. See Manning 1985, 68 fig. 17. 301 Initially, M113 and M117 were believed by the excavator to be the handle and lock for a wooden door separating Room 2 and Room 3. However, the identification of M117 as a lock is highly questionable. It does not resemble any other lock recovered at the site in terms of its size or shape. No other evidence for the presence of a door in the form of identifiable fittings or cuttings in the threshold and jambs between the two rooms were identified. 302 Soren and James 1988, 96-8; Dias et al. 1988, 182-4 Plate LIII #6. This object has been identified by Soren as an “heirloom” dating to the late 2nd or early 1st century BC based on the form of the lamp and the spiral fluting of the column. He states that “while spiral shafts are quite common among the Etruscans, the Romans seem not to have fancied them.” However, columns with spiral fluting appear frequently at Roman sites throughout the eastern Mediterranean including Kourion, where several examples have been recovered. It is more likely that the object is based on contemporary artistic influences from cities in the Roman Near East, particularly the Syro-Palestinian coast. A second copper alloy lamp was found atop of Wall C between Room 1 and Room 2. It was recovered from Locus 003, a post-earthquake deposition layer above the post-earthquake collapse layer of Locus 005. Based on this context, this lamp appears to be an object deposited significantly after the earthquake. 303 The following registry numbers are recorded for this object: S20, S21, S33, S24, S41, S42, S44, S47, S51, S52, and S59. 300

Unlike other rooms of the structure, where metal objects are absent or a minor assemblage component, in Room 2 they form the largest category of material. A significant number are Class 1a objects associated with the mule remains, predominantly iron chain links or link fragments that tethered the animal to the trough. Copper alloy (M97) and iron (M118) rings found by the skull of the animal

297 The majority of these coins were not identified in either set of excavation documentation. The coins from the Arizona excavation were cleaned by the conservators and then to the late Eugene Lane for identification (Soren and James 1988, 86). No report or copy of Lane’s notes was present in the Kourion archive. Efforts to locate notes regarding the Kourion material in Lane’s archives were unsuccessful. 298 For an osteological report of the remains recovered in Room 2, see appendix “Report on the Human Skeletal Remains from the 1984 Excavations, Kourion City, Grid L 9” by A. Galloway 302-6 in Davis and Soren 1985.

58

Figure 6.6: Room 2 Plan. Reconstructed by author from AZLS: Room 2. be securely identified as Class 1d.

were recovered in situ so their general pattern of dispersal can be visualized.304 Horizontally, the main scatter concentration measures ca. 1.5 X 0.5 m. It is oriented roughly east-west directly in front of the Wall B doorway. Vertically, coins begin appearing at an elevation ca. 85 cm. above the bottom level of Locus 012 and continue down to the lowest level excavated. The bulk (136/80.00%) occur in a 61 cm wide band ca. 74-11 cm

Further evidence supporting the hypothesis of a second story is the substantial number of copper alloy coins concentrated on the north side of Room 2. One hundred and seventy coins were recovered from this area, of which 101 (59.41%) were from the post-earthquake collapse layer (Locus 005), with the remaining 69 (40.59%) from the earthquake debris layer (Locus 012). The majority of the coins (100/58.82%) were recovered during screening, but a significant number (57/33.53%)

304 The remaining thirteen coins have no provenience data recorded. It is unknown whether they were recovered in situ or during screening.

59

Figure 6.7: BI2 – Bone pin, top and side views.

Figure 6.8: 6.8: BI5 BI5 –– Bone Bone pin. pin. Figure

Figure 6.11: M116 - Unidentified iron and lead(?) object. Originally identified as a lock, but fits no known typology. From top to bottom: top view, back view, side view.

Figure 6.9: BI6 and BI7 – Carved bone object, probably inlay or veneer. Top and side views.

Figure 6.10: M117 – Iron curb bit. Part of the mule tack assemblage. Figure 6.12: S20, S21, S33, S34, S41, S42, S44, S47, S51, S52, S59 – Round marble tabletop with carved border. 60

Figure 6.13: Room 3 plan. Reconstructed by the author from AZLS: L9/Room 3.

61

above the bottom levels for Locus 012.305 Based on their context, the majority of the coins from Room 2 should be identified as Class 1d. However, the possibility exists that at least a portion, particularly those outside the main concentration, may be Class 1c, Class 2 (deposited as a result of incidental loss) or Class 4. Given the dispersal pattern, it is highly probable that the glass jar (G104) found on the eastern edge of the concentration was the vessel in which the coins were stored. 6.4.3: Room 3 In contrast to Room 2, Room 3 preserved a substantial ceramic assemblage (Fig. 6.13). Only three vessels, however, have a completeness level that permits Class 1b identification. Two of these, a jug (187) similar to a Vasa Type 1H (Fig. 6.14)306 and a small single handled cup/mug (235) were found on the occupational surface along Wall B. In the northwest corner, a single unbroken and complete Vessberg 18 lamp (362) of the “fisherman” variety was recovered within the earthquake debris layer (Locus 007) 14 cm above the occupation surface.307 Based on its findspot, vessel 363 is likely Class 1c, although Class 1d is also possible. The last high completeness vessel is a Kourion Type III (KT III) amphora (161/170/209)308 from the south side of the room against Wall E (Fig. 6.15). At the time of its reconstruction, several large body fragments were unaccounted for. A search of the Room 3 sherd material produced a number of additional fragments that account for the majority of the gaps. It is noteworthy, however, that these were recovered from the area around the doorway; the opposite side of the room from the remainder of the vessel.309 The findspot of these missing fragments is telling, indicating that despite its level of completeness 161/170/209 was a Class 2 object broken prior to the earthquake.

Figure 6.14: 187 - Jug. Similar to Vasa Form 1H. objects is confirmed by the presence of at least two vessels with cross-joins from other areas of the structure. The first is a partially reconstructible plainware bowl (182) from the northern part of the room along Wall B. The fragments display moderate brokenness with some edge rounding, indicating minor-moderate exposure to During post-breakage formation processes.310 examination, three joining base fragments were identified: two from the area immediately around the doorway; and one from the Room 2earthquake debris layer (Locus 012).

The four examples discussed above are anomalies within the Room 3 assemblage. While many of the remaining vessels in general have elevated degrees of completeness, all are far below the in use percentage threshold, which warrants Class 2 identification. A search of the sherd material did identify some joining fragments but not enough to significantly alter this status assessment. The interpretation of the Room 3 assemblage as being at least partially, if not predominantly, composed of Class 2

The second example is more definitive due to the distance between findspots. Immediately inside the doorway of Room 3, fragments of a moderately-sized pithos rim were recovered.311 Although many fragments are sizable, the overall vessel completeness is low (ca. < 20%), indicating Class 2 status at the time of the earthquake. Two nonjoining pithos rim fragments of similar profile and decoration to the Room 3 fragments were observed in material from the northwest part of Room 11. An examination of the rims from both rooms revealed that the larger from Room 11 joined with those from Room 3, creating a definitive link between them (Fig. 6.16). While in the case of Vessel 182 an argument could be made that the joining sherd from the adjacent Room 2 is

305 The published elevation for the floor of Room 2 is 67.94 ASL (Davis and Soren 1985, 299). However, the floor was uneven and several finds have a recorded elevation below this level. Because of this, 67.83 ASL, the bottom level recorded for Locus 012 is used here instead. 306 For the Vasa jug typology, see du Plat Taylor 1958, 28. The Kourion specimens have a more rectangular and flared rim profile than those recovered at Vasa, with the upper handle attachment point below rather than on the rim. 307 For typology see Vessberg 1953; Vessberg and Westholm 1956. Published in Dias et al. 1988, 179 fig. 1. 308 Williams 1987, 236. 309 The transitional area between Room 2 and Room 3 was given its own locus designation, Locus 008, in the excavation notebooks. The missing fragments were found in Basket 19, which is recorded as originating from Locus 008.

310

See Bradley and Fulford 1980. The numerous fragments of this vessel were registered with the following numbers: 174, 196, 197, 208, 213, 214, 217 and 390. The fragments were also designated in the excavation notes as “Pot CC,” “Pot A4,” and “Sherd Pile B5.” The rim profile is similar to KalavasosKopetra 215 (Rautman 2003, 206) and Kourion Survey 63n (Leonard 1987, 109). 311

62

The glass from Room 3 appears to be Class 2, with the only clustered fragments belonging to a vessel of indeterminate form (G133/134) along Wall B. Due to their fragile nature, the fragments were consolidated in situ, but with the exception of the handle, which may indicate a small flask or jug, no other identifying characteristics could be identified.313 The remaining Room 3 material can be divided into two categories: non-personal objects that are Class 2, and personal objects that are likely Class 1d. The former category consists of two metal objects. The first is a flat, irregularly shaped piece of lead (M48) with a knobshaped protrusion on one side (Fig. 6.17). Given the amount of residual ceramic material in this space, it is possible that M48 may be a lead clamp or plug used to repair a ceramic vessel. The second object is a small, roughly oval copper alloy ring (M55) whose ends were attached using a lap and rivet joint. The shape of the object and rough nature of the join points toward some type of utilitarian function. Personal objects from Room 3 include a small complete copper alloy tool (M44), which based on its size and shape is probably either a kohl stick or a cosmetic spatula.314 A bone pin fragment (BI1) that preserves the head and upper shank (Fig. 6.18) was also recovered. The head and a portion of the shank is covered with a thin metal sheath that is identified as silver in the excavation notes. If this is correct, the pin represents a luxury object that even if broken is unlikely to have been discarded without first salvaging its precious metal cover. In addition to these two personal objects, 124 copper alloy coins were also recovered from Room 3; four from the occupation surface (Locus 013) and the rest from the

Figure 6.15: 161/170/209 – KT III amphora. incidental,312 no such claim is possible for the pithos since the straight line distance between the fragment findspots is ca. 10 meters. Based on the structure’s architectural layout during its final phase, an actual “walking” distance of ca. 17 meters separated them. Although the presence of cross-joining fragments in multiple rooms may indicate post-occupation dumping, this does not appear to have been the case in Room 3. The fragments of both vessels were recovered from the occupation surface with no evidence of post-deposition disturbance observed during excavation. While the exact process by which the pithos fragments were deposited on opposite sides of the structure is difficult to determine, their separation represents a deliberate behavior related to discard processes during the final occupation.

Figure 6.16: 174/196/197/208/213/214/217/390 – Partially reconstructible pithos rim. The two lighter colored fragments on the left were recovered from Room 11; the remainder from the eastern side of Room 3. 313 As in Room 2, glass from the 1984 season was unable to be located. An examination of the field notes and plans from 1984 do not indicate any substantial glass finds in the eastern half of Room 3. 314 Shape similar to specimens published in Meyers et al 1981, 218 (nos. 17 and 18).

312 In this case, fragments of vessel 182 were recovered from the Wall D threshold area, which could have been displaced into Room 2 by seismic waves from the earthquake.

63

amphora (V478) is noteworthy, favoring a Class 2 identification despite its high completeness level. Evidence supporting this interpretation will be discussed below. The largest storage vessel in Room 6 is a single moderately-sized pithos (V443)317 identified as Class 1b. It was recovered from the southeast corner of the room (Fig. 6.20) with its base on the occupation surface. The vessel was not equipped with any provision for drainage, indicating its contents were removed from the top. The absence of a lid may point toward V443 being either covered by one in wood or that its contents were left open to the air. A single VTC amphora (V476) in the northern part of the room (Fig. 6.21) possesses completeness that also permits Class 1b identification. Unlike the larger vessels in Room 6, the small capacity of V476 made it quite portable, providing a means to transport commodities within and outside this space. The remaining ceramic Class 1b object is a fully reconstructible funnel (520) from the south side of the room, close to the pithos (Fig. 6.22).

Figure 6.17: M48 – Unidentified lead object. earthquake debris layer. Based on their context, the majority of the coins should be identified as Class 1d, although a small percentage could possibly be Class 4. While the majority of the coins (95/76.61%) were recovered during screening, the remaining in situ specimens (29/23.39%) display a rough southeastnorthwest diagonal scatter pattern across the center of the room. The Class 2 status assigned to the majority of the ceramic vessels as well as their findspots makes it unlikely that one of them held the coins. Their wide distribution and context appears characteristic of having fallen from an upper floor.

Room 6 contains the remains of five vessels (508, 474-1, 505, 523 and V532) whose completeness and/or find context indicates they should be identified as Class 2. Of these, only the single LR3 amphora (508) was not largely reconstructible. In contrast, KT I amphora 474-1 (Fig. 6.23) preserves completeness and brokenness levels consistent with being intact at the time of the earthquake. However, the presence of a ca. 4 cm circular hole deliberately cut into its wide base rendered it ineffective as a storage container except for larger, non-liquid commodities, for which no evidence was recovered in Room 6.318

6.4.4: Room 6 The bulk of the Room 6 assemblage (Fig. 6.19) consists of fourteen amphorae of five types. Most abundant is the KT III represented by seven examples, followed by the larger Kourion Type I (KT I) with four examples. One each of the Kourion Type II (KT II), KT IV/LR4, and Late Roman 3 (LR3) complete the amphora assemblage.315 Of these fourteen vessels, nine316 (64.28%) preserve a completeness level permitting Class 1b identification. However, the context of one KT I

The single KT IV/LR4 amphora (505) found against Wall K was largely reconstructible. However, it exhibited a higher degree of brokenness than other vessels and was missing a large section of its lower body, characteristics consistent with breakage prior to the earthquake. For KT III amphora V532 and KT I amphora 523, evidence for pre-earthquake breakage is definitive. Although their primary findspots are on opposite sides of the room, the excavation and registry notes record fragments from both were part of a sherd concentration recovered near the passage connecting Room 6 and Room 7.319 In the case of 317 Profile similar to Kalavasos-Kopetra 215 (Rautman 2003, 206) and Kourion Survey 63n (Leonard 1987, 109). 318 This seems to be a smaller version of the second method of opening amphorae discussed by Bonifay (2004, 467-8). This involved the removal of a circular sherd ca. 10 cm in diameter, normally from the shoulder of the vessel. It is possible that one of the KT III amphorae (V403) has a deliberate hole cut in the vessel wall towards the bottom, but this cannot be conclusively determined due to modern plastering by conservators. 319 The excavation notes are confusing and reflect the jumbled nature of the material when excavated along Locus 003. On a sheet labeled “Finds from Room 6 (21 Pots)” the note for the amphora identified as 523 indicates a position “over 474,” with the note for amphora 474-1 recording “many pieces from no. 523, found 474…” Similarly the entry for amphora 479 recovered in front of the passage reads “partly into [room] 7. Also mends with nos. 523, 532.”

Figure 6.18: BI1 – Bone pin with a metal covered head. Metal identified by the excavator as silver. 315

The KT III accounts for 50.0% of the assemblage from Room 6, with the KT I accounting for 28.57%. The other three types make up the remaining 21.3%. 316 These include two KT I, the single KT II, and six KT III amphorae.

64

Figure 6.19: Room 6 Plan. Reconstructed by the author from AZLS: Room 6.

65

Figure 6.22: 520 – Plainware funnel.

Figure 6.20: V443 – Pithos.

Figure 6.23: 474-1 – KT I amphora side and bottom view showing deliberately cut hole.

Figure 6.21: V476 – Vasa Type C amphora.

66

vessel 523, multiple fragment findspots combined with significant missing portions confirms Class 2 identification. For V532, information regarding the fragment findspots is crucial since the vessel’s completeness could lead to it being identified as Class 1b when it was not. The examples of Vessel V532 and Vessel 523 demonstrate some of the pitfalls of permanently reconstructing vessels. While informative regarding vessel shape and size, removing fragments from their find contexts can greatly impact assessing a vessel’s status. Due to the reconstruction of these vessels, the extent to which their fragments were dispersed and what percentage of each vessel was found in each location cannot be determined since only minimal documentation for this exists in the excavation archives. Class 1c ceramic objects consist of a partially reconstructible (ca. 50%) two-handled cookpot (404/444)320 and a nearly complete (≥ 95%) small singlehandled cookware pitcher (445). Both were found among the pithos fragments near the intersection of Wall J and Wall K. Although they were found somewhat high in the stratigraphic sequence, raising the possibly they are Class 4, their contextual relationship with the pithos indicates this is unlikely. Rather, their findspot likely reflects either shelf storage on Wall J or Wall K above the pithos, or within the pithos itself. Since no facilities for cooking or evidence of burning were observed during excavation, it is probable that 404/444 and 445 were employed in a secondary use role, possibly as dippers or scoops to measure out commodities. In addition to the ceramics, a significant glass assemblage was also recovered from Room 6. The majority came from the same stratigraphic context as the two cookware vessels 404/444 and 445. Because of their fragile nature and find context, they are identified as Class 1c. The main concentration has an Average EVREP of 8, with a small bottle (G138), being almost completely reconstructible (Fig. 6.24).321 All of the identified glass vessels are small, with the better preserved examples ranging in size from 9.5-11.5 cm in height and 3-4 cm in width.

Figure 6.24: G138 – Glass bottle. 6.4.5: Room 7 The Room 7 assemblage consists predominantly of amphorae; nineteen vessels of three types (Fig. 6.25). Numerically, the KT I is represented by nine examples, followed by the Kourion Type V/Late Roman 1 (KT V/LR1) and KT III with six and four examples respectively. Of the nineteen vessels, only four (21.05%) have a completeness level permitting Class 1b identification. Of the nine KT I amphorae, only one, V551, is fully reconstructible. Three, V477, P510, and V535, were partially reconstructible324 and are either Class 1b functioning at diminished capacity or Class 2. The five remaining KT Is are represented almost exclusively by diagnostic elements (necks, handles, bases/spikes) and are Class 2.

The final glass vessel from Room 6 is a partially reconstructible juglet (G137), 322 recovered near the center of the room close to KTIII amphora V542. Although parts of G137 are reconstructible, its overall low completeness level combined with its findspot indicates it is Class 2.323

The general characteristics of the KT I amphorae are also observable among the KT IIIs. Of the four vessels of this type, two, V519 and V536, can be designated Class 1b. The remaining two were recovered in the northwest corner of the room and are Class 2, represented primarily by diagnostic elements with no associated body fragments.

320

Profile similar to Vasa Type 3B (du Plat Taylor 1958, 37) and Dhorios Form P111. For Dhorios cookware typology see Catling 1972. 321 Form similar to # 200 in Hayes 1975. 322 Vessel identified in the excavation notes as an amphoriskos, but a more likely identification is a juglet. 323 The alkaline soil conditions at the site are not ideal for glass preservation (Soren and James 1988, 114), and no mention is made in the excavation notes as to completely degraded material recovered around G137. If degraded and/or fragile fragments had been encountered, they would have been consolidated in situ as in other cases.

324 V477 – No neck; P510 – Large shoulder fragment missing; V535 – Large shoulder fragments missing.

67

Figure 6.25: Room 7 plan. Reconstructed by the author from AZLS: Room 7. Figure 6.25: Room 7 plan. Reconstructed by the author from AZLS: Room 7.

68

The last type is the KT V/LR1, which were clustered in the southwest corner and represent the only specimens of this amphora type recovered from the Earthquake House. All possess red dipinti,325 indicating that they had been utilized for shipment at least once. Only one example (V552) was completely reconstructible (Fig. 6.26), permitting a Class 1b identification. The remaining KT V/LR1s generally possess higher completeness levels than the other types represented in the room. They preserved the top portion including the neck, both handles, and upper shoulder along with a significant portion of the body. However, the bases/bottoms of all except V552 are missing, necessitating Class 2 identification.

and Locus 002, as well as objects from Locus 001 that have joining fragments recovered elsewhere in the structure.326 The presence of such post-earthquake material is problematic and has the potential to seriously impact interpretations of the artifactual evidence. Because of the heightened susceptibility of Room 8 to this phenomenon, extra effort was placed on identifying and mitigating Class 4 contamination so that conclusions regarding object status and assemblage composition at the time of the earthquake were as accurate as possible. The majority of the reconstructible Room 8 vessels were recovered in its northern half; larger ones near the walls with smaller vessels located more centrally (Fig. 6.27). In the northwest quadrant, a total of three vessels; V560, an unregistered KT III amphora (BSK 119), and V651 preserve a completeness level that permits Class 1b identification. Vessel V560 is a large two-handled plainware basin327 (Fig. 6.28) located in the northwest corner near the KT III amphora (UR: BSK 119). The fragment findspots and elevations for the latter indicate that it was probably upright against Wall K and fell to the south during the earthquake.328 Vessel V651 was located south of the KT III amphora at a slight distance from Wall K. Its type has been classified by Hayes as an “eggshaped bottle,”329 a form that appears relatively frequently throughout Egypt in 4th-5th century AD contexts. Based on its fabric, V651 may have origins in the Aswan region.

6.4.6: Room 8 Analysis of the Room 8 assemblage poses several challenges. Perhaps the most significant is the room’s size, which in several areas did not permit the earthquake debris layer to be completely sealed. The archaeological evidence indicates the tile roof over Room 8 collapsed towards the center, leaving voids where Class 4 material could infiltrate. Evidence for post-earthquake deposition takes the form of ceramic cross-joins between Locus 001

In the northeast quadrant, only one vessel, a small Cypriot Red Slip Ware (CRSW) or Late Roman C (LRC) cup (V550) recovered towards the room’s center can be identified as Class 1b. The remaining vessels include a KT III amphora along Wall E, a CRSW Form 1 plate (V555), and a Dhorios Type P132 cookpot (V567). Although all possess relatively high completeness, significant missing fragments as well as their findspots raise questions as to whether they should be identified as Class 1b, Class 2, or Class 4. Of these three vessels,330 V567 has the greatest probability of being Class 1b despite its high brokenness; a condition no doubt exacerbated by the thin and brittle nature of its fabric. The southeast quadrant of Room 8 produced three vessels. Two can be assigned to the pre-earthquake 326 The most notable example is a large stone table fragment (S69) recovered from Room 8, Locus 001 (outside the structure) that joins with a fragment (S73) recovered from Room 11, Locus 001. 327 Similar to Dhiorios Type P307 (Catling 1972, 67). 328 The findspot of the intact base as well as the area of scatter where this vessel was found are recorded in the field notebooks and plans for this room. The base of the vessel was north of the main fragment concentration. Either a horizontal position or a fall to the south during the earthquake is the most probable explanation for this pattern. 329 Soren and James 1988, 113. V651 is the vessel on the right of the photo. Nearly identical vessels have been recovered in 4th-5th century AD fill contexts in the Athenian Agora (Hayes 2008, 93; 253-4). 330 Unlike the other vessels, V550 was recovered below the tile collapse layer. The rest of the vessels were recovered toward the northern part of the room, which was not sealed by the tile collapse and more prone to infiltration. As a result, the possibility that these vessels represent material discarded post-earthquake cannot be dismissed, particularly in light of their overall levels of completeness.

Figure 6.26: V552 – KT V/LR1 amphora. 325 Dipinti observed on the KT V/LR1 amphorae were sketched and photographed. The majority of the vessels preserve only fugitive traces of paint, which cannot be reconstructed with any degree of certainty. These markings most likely represent notations of capacity, contents, or ownership.

69

Figure 6.27: Room 8 Plan. Reconstructed by the author from AZLS: Room 8.

70

Figure 6.28: V560 – Plainware basin. (Note: Missing rim fragments found in loose sherd material postphotograph.) assemblage although their characteristics indicate at least expected. Given these problematic aspects, no definitive one, and probably both are Class 2 objects. These classification of V578 can be made. If Vessel V578 was vessels, a moderately sized pithos (V568) (Fig. 6.29) and part of the pre-earthquake assemblage, its findspot likely a VTC amphora (V594), were located on the eastern edge indicates storage either on a high shelf or in a loft of some of this quadrant but not directly against the wall. Of the type formed by the roof framework. In this case, it would two, V568 is more complete and is similar in size and be designated as Class 1c. shape to the Room 3 and Room 11 pithos rims, and the Room 6 reconstructible vessel (V443). Despite V568’s The final ceramic vessel is a small CRSW Form 1 bowl high completeness level, several large fragments remain (579/601) with low brokenness and high completeness unaccounted for, with a search of the Room 8 sherd (Fig. 6.30). Although a small section of its floor is material producing no definitive or probable matches. missing, the fabric thinness caused its breakage into very Based on their size, number, and location, it is unlikely small fragments that likely went unnoticed during that this vessel could have functioned as a container in excavation and screening. The large fragment findspots any meaningful capacity and thus has to be identified as are only generally known, but their recovery elevations Class 2. indicate the vessel is Class 1b.332 Like the ceramic assemblage, the glass from Room 8 is problematic. A significant amount of Class 4 glass appears to have been deposited post-earthquake, infiltrating the earthquake debris layer through gaps in the rubble. Of the glass recovered, no identified vessel has a completeness level permitting Class 1 identification. The only possible exception is a fragment concentration from the northwest quadrant (UR: BSK 191), but whether it represents a complete or partial vessel cannot be determined. The findspot of this material from an area outside the tile fall presents the possibility that it Class 4, but their proximity to the occupation surface may indicate an object in the room when the structure was destroyed.

Unlike V568, less ambiguity exists with V594, whose high brokenness and moderate completeness are indicative of a Class 2 object. At some point, the vessel underwent maintenance that resulted in the entire rim being removed. No additional fragments attributable to this vessel were recovered, indicating that the maintenance processes either occurred elsewhere and/or were not recent. The third vessel from this quadrant is a large, heavy mortarium (V578).331 Despite its high completeness level, a section of the vessel floor was not recovered. No additional definitive or probable vessel fragments were located during a search of the Room 8 sherd material. Questions regarding the status of V578 are complicated by its findspot high in the earthquake debris layer among the roof collapse. This, combined with the missing section raises the possibility that V578 is a Class 4 object. However, based on other vessels of this class identified throughout the structure, if V578 is Class 4, an overall lower completeness level combined with a more widespread dispersal of the fragments should be

331

Perhaps the most noteworthy aspect of the Room 8 assemblage is its significant number of metal objects. These include M161, a large hanging lamp complete with chains and hook (Class 1c), M164 and M165, two singlehandled pitchers with incised decoration and separately

332 Based on the excavation notes, the fragments of V579/601 were all recovered in general pottery baskets located in the area immediately around the central column. Although one fragment is recorded as being recovered from Locus 001, the lower portion of this basket overlaps elevations generally assigned to Locus 002.

Rim profile similar to Kourion Basilica L14 (Megaw 2007, 468).

71

stored. They were significantly corroded, which complicated identification in some cases, but all were interpreted as being Class 1b. Two (M174, M191) are likely tools, although their exact function(s) could not be determined with any degree of confidence. The first (M174) is a large cylindrical object that tapers towards one end. It was found in the northwest quadrant close to the KTIII amphora (UR: BSK 119) along Wall K. Given its size and shape, a number of possible identifications are possible, including a spike, chisel, or punch. The second, M191, is a slightly concave rectangular object with rounded corners that was found close to the stone table.335 Although M191was identified as a knife by the excavators, its characteristics are not consistent with others that have been recovered at Kourion and elsewhere. It more likely represents some type of multifunctional tool such as a pry-bar or lever. In contrast to M174 and M191, objects M177/M180 and M175 can be identified with certainty although the exact function of the latter remains unclear. Three joining iron fragments (M177 [2 Pieces] and M180) found north of the stone table form a complete iron spearhead (Fig. 6.32).336 Despite being heavily corroded, the general characteristics of the object can still be discerned; an elongated elliptical blade that transitions to a cylindrical socket for the shaft. Given the context in which the object was found as well as the lack of any significant military presence on Cyprus during the Roman period, it is probable that M177/M180 would have been used primarily, if not exclusively, for hunting.

Figure 6.29: V568 – Pithos. added feet (Class 1b or Class 1c),333 and M183, a set of proportional dividers (Class 1b) found atop the circular sand deposit (Locus 005) on the south side of the room.334 Five smaller copper alloy objects that are likely Class 1b or Class 1c were also recovered. One is a straight cylindrical shaft (UR: BSK 166), which could be part of a needle, straight pin, or nail. The other four (M178, M179, M185, M192) are U-shaped, one of which (M178) is recorded as having preserved fibers attached to its surface (Fig. 6.31). Although the type of fiber was not identified, its presence may indicate that these objects were used as some type of pin, possibly for clothing or other textiles. The spatial distribution of these U-shaped “pins” appears random and does not presently permit any insights as to their function.

Food procurement may also have been the function of M175 (Fig. 6.33), a four-tined “fork” recovered south of the central column and table. The object was created from a single U-shaped piece of iron used to form the external tines, with the two interior tines added separately. The tips of each tine are pointed, but not barbed, and no evidence for a hafting socket is present. If it was used with a handle, this would need to have been attached separately. Based on its size and shape, it is

The Room 8 iron objects were recovered predominantly towards the room’s center where they were most likely

Figure 6.30: 579/601 – CRSW Form 1 bowl.

Figure 6.31: (L – R) M178, M179, M185, M192 – CA objects, likely pins.

333

Similar in overall shape to the pitcher recovered from Room 14. M161: See Soren and James 1988, 115; Dias et al. 1988, 182-3, fig. 3. Three small chains attach to a central one with a suspension hook. All four chains show evidence of ancient repair. M164 and M165: Soren and James 1988, 115; Soren et al. 1986, 203, Pl. XLIV: 2. These vessels appear similar to the specimen recovered by Daniel in Room 1. M183: Soren and James 1988, 116. 334

335 A note found on the M191 provenience tag records that it was found in association with charcoal. No description of this relationship appears to have been made elsewhere in the excavation notes. 336 The total length of M177/M180 is 25 cm, with the blade having a maximum length of 16.5 cm and a maximum width of 4.0 cm. Similar to Manning Type V105 (Manning 1985, Pl. 79).

72

Figure 6.32: M177 and M180 – Iron spearhead. that the object represents a drinking vessel.338 Given its findspot, M176 may have been on the table at the time of the earthquake and fell onto the occupation surface during the event.339

tempting to identify M175, it as a “trident” for fishing. Its characteristics, however, are not consistent with known examples, which normally have three large barbed tines and a hafting socket.337 More probably M175 represents some type of agricultural implement, possibly a “rake” for harvesting olives or a multi-tined hoe. At present, no parallels for this object have been found in published material.

Five copper alloy coins were recovered from Room 8. Of these, only two (C389, C390) are from the occupation surface: directly south of the central column; and under the stone table respectively. This permits an identification of Class 1b or Class 2. The remaining three coins were collected during screening so their context is only generally known. It is possible that some or all of them may represent Class 4 objects.

One Class 1b or Class 1c lead object, a small cup (M176) was found immediately south of the stone table (Fig 6.34). Based on its size and shape, it is highly unlikely

The Room 8 stone assemblage is composed of two objects: a completely reconstructible large limestone mortarium (S75/S86/S86) from the northwest quadrant (Fig. 6.35); and a roughly-carved square stone basin (S84) found on the south side of the room. The mortarium is a standard type with numerous examples being found elsewhere on the site. It is roughly circular with a hemispherical depression and two rectangular “lug” handles on either side. The rim shows signs of significant wear in some areas, which indicates significant use prior to deposition.340 Based on its completeness and findspot it is identified as Class 1b. Small square stone basins such as S84 were ubiquitous at Kourion with intact and broken examples reused for wall construction in many structures throughout the site. Although S84 is largely complete, ca. one-third of the basin wall has been broken off, in some areas down to the inner floor. No joining fragments appear to have been recovered, indicating that the object was likely broken prior to the earthquake. The missing section would have prohibited the object from functioning as a container, 338 A note on the field tag records that the contents of this object were kept. No comments or analytical results of this material, which could potentially help elucidate its function, were found in the excavation documentation. 339 A small fragment of lead sheet (UR: BSK 97) was also recovered in the southwest quadrant of the room close to Wall J. Although its function cannot be determined with any degree of confidence, based on its size and character, two interpretations seem the most probable: raw material for some activity carried out in the room, or scrap material possibly in some type of discard context. 340 The softness of the stone from which this object was carved may have enhanced its worn appearance, giving the impression that it was in use for a longer period of time than it actually was.

Figure 6.33: M175 – Four tine iron fork. Likely an agricultural implement or possibly used for fishing. 337

This was tentatively identified as a “pitchfork” at the time of excavation (Soren and James 1988, 116). However, its small size would seem to preclude it from the type of agricultural activity this nomenclature implies.

73

Figure 6.34: M176 – Lead “cup.” Top and side views. making Class 2 identification possible. However, an alternate interpretation is that S84 was in some type of secondary or reuse context, in which case it would be identified as Class 1b. The justification for the latter is the fact that it was found upside-down, which would have permitted its bottom surface to be used as some type of work platform or as a seat. This hypothesis gains credibility based on its proximity to the circular sand deposit and copper alloy proportional dividers (M183) on the occupational surface of the room (Locus 005). When taken together these three elements seem to form some type of distinct activity area, although for what purpose cannot be determined at present.

Figure 6.35: S75, S86, S87 – Stone mortarium. was utilized as a receptacle for household refuse during its final phase. The present study has determined that at least part of this waste material originated from a provisional discard area in the adjacent Room 14. This link between these two spaces has significant implications for reconstructing behaviors pertaining to the domestic waste stream of the Earthquake House during its final occupation. A detailed model of these processes derived from the Room 12A, Room 14, and Room 19 material will be presented and discussed in the following chapter.

6.4.7: Room 12A This space was not a room but a dedicated water storage installation (cistern). The various stratigraphic units and artifacts from Room 12A are highly informative regarding its use, particularly during the terminal phase of the Earthquake House.

6.4.8: Room 14 Room 14 produced the most diverse assemblage of the Earthquake House (Fig. 6.36). A total of six storage vessels are Class 1b. The first is a large, intact pithos (V617) located in the northwest corner of the room (Fig. 6.37).343 In contrast to the pithoi from Room 6 and Room 8, which had a wider, flat base, the Room 14 example has a base diameter that is considerably smaller than the vessel’s body and rim. This likely would have caused V617 to be relatively top-heavy when full, necessitating the placement of three stones around its base to provide stability. The occupation surface surrounding this vessel is recorded as being considerably harder than in the rest of the room. Although this may indicate the pithos held an essential commodity that required frequent access, it may also represent a specially prepared surface intended to provide a stable platform for the heavy vessel. Areas of intact plaster patching demonstrate that V617 was repaired at least twice during its use life, a common phenomenon on large vessels due to their high cost and possibly limited access to replacements.344

All objects from Room 12A can be identified as Class 2. Those from Locus 006 represent material accidentally deposited when the installation was functioning as a cistern. These consist of a small quantity of broken ceramic, a coral, jet, and amber bead necklace, and a silver-plated copper alloy bracelet.341 Since these objects were not or could not be retrieved, they remained in the accumulated sediment layer on the tank’s bottom. In contrast, the material from the ash deposit in the southwest corner (Locus 005) and the lower part of Locus 003 represent Class 2 objects deposited when Room 12A ceased functioning as a cistern and was being used in a secondary capacity. These loci produced a moderate amount of broken ceramic (Average EVREP = 32) as well as a significant amount of animal bone (ca. 200 fragments). No reconstructible or partially reconstructible vessels were identified. The characteristics of the Locus 003 and Locus 005 material are consistent with domestic refuse deposits documented at other Mediterranean sites.342 A detailed examination of the Room 12A material has confirmed the excavator’s hypothesis that it

In conjunction with the pithos, additional storage was 343 Profile similar to Kalavasos-Kopetra 206. The Earthquake House example has two bands of impressed diamond shaped decoration below the rim. 344 See especially Peña for discussions concerning the economic value (2007, 27-31) and the maintenance/repair operations carried out on large ceramic vessels (2007, 210; 213-227).

341

Molinari et al. 1988, 173; Soren and James 1988, 124-5. See especially De Caro 1994, Small et al. 1994, Stampolidis 2004, Maiuri 2002, Chiaramonte Treré 1986, and Ault 1999. 342

74

Figure 6.36: Room 14 plan. Reconstructed by the author from DFN: 6/4/1934, 21 and AZLS: Room 14. provided by three amphorae, two KT IV/LR4s (V621 and P129-1934), along Wall T on the north side of the room and a KT II (V615) in the southwest corner.345 The two KT IV/LR4s appear to be of the standard type, although

V621 only possesses a single handle. In the case of P1291934 (Fig. 6.38: A), its overall capacity was diminished since a significant fragment of the shoulder appears to be missing. It is unknown if this was removed intentionally

345 DFN: 6/4/1934, 22. The Trench III notes also record what Daniel describes as “the top of a large amphora,” which he designated P132, among the material in the northeast corner of the room. This object does not appear in the field photographs. The fact that Daniel mentions specifically the top of the vessel seems to indicate that the remainder

was not recovered. This likely indicates reuse or residual status. The vessel is described as having “handles neck to shoulder” with “orange brown clay.” Although the specific amphora type cannot be determined from this description, it fits the general characteristics of those possessed by a Kourion Type III.

75

half is present. The upper portion appears to have been deliberately removed, since the preserved edge is very even, much more so than would be expected if caused by accidental breakage. It seems that P665 had undergone maintenance to remove its (presumably damaged) top, with the bottom half retained for continued use. Despite its largely reconstructible state, sections of the lower walls as well as nearly the entire base of P665 are missing. This would have made it inadequate for use as a storage container except for large, non-liquid commodities such as fruits, vegetables, etc. If this was the case, P665 should be identified as Class 1b. If not, it probably represents a Class 2 object, whose findspot is significant given the substantial quantity of Class 2 material found clustered around the oven. Two ceramic vessels, a flat bottomed basin with thumb impressed handles (V620) found adjacent to V615, and a two handled cookpot (P131-1934) in the northeast corner (Fig. 6.38: C) can also be identified as Class 1b. The basin is of a form common during the late 4th century AD based on the number of Class 2 fragments with similar handles observed throughout the assemblage.348 The cookpot appears to be of the standard shape and size as other examples recovered at Kourion and other sites throughout Cyprus. It is the only cooking vessel within Room 14 whose overall completeness level permits Class 1 identification.

Figure 6.37: V617 – Pithos. Note ancient plaster repairs on side. since the vessel could not be located for physical examination, but the regularity of the edges without any apparent surrounding damage is noteworthy and may indicate deliberate removal. The KT II amphora (V615) is nearly identical to the example from Room 6 (Fig. 6.39). It preserves several prominent manufacturing defects, including a warped and dented wall, an oval rather than a round neck, and a general asymmetry in shape. Both V621 and V615 possess multiple dipinti,346 indicating that they had been utilized at least once as transport vessels. Both were likely being reused as general storage (probably for foodstuffs) during the final occupation. The final Class 1b storage vessels are two VTC amphorae: one near the northeast corner of the room (P130-1934) (Fig. 6.38: B); and the other along Wall F (V619) (Fig. 6.40). The lower half of a third VTC amphora (P665) was recovered adjacent to V619, but its condition makes classification problematic. At the time the structure was destroyed, V619 was missing the rim, one handle, and a significant portion of its neck. The fact that none of these fragments were recovered during excavation supports its breakage prior to the earthquake with the vessel continuing in use.347 The status of the second example, P665, is less certain since only the lower

Figure 6.38: Field photograph recording the group of objects recovered in the northeast corner of Room 14 during the Trench III excavation in1934. A: P1291934; B: P130-1934; C: P131-1934; D: P133-1934; E: ST13-1934. Photo courtesy of the Cyprus Department of Antiquities archives.

346 The second KT IV/LR 4 specimen P129-1934 was recorded in excavation notes as being “painted,” which may indicate that dipinti were present. No dipinti are observable in the excavation photos. 347 Molinari et al. 1988, 175. The incomplete nature of this vessel was noted by the excavators, who interpreted the continued use of a “broken” vessel as “a possible sign of the depressed state of the local economy.” However, this interpretation does not consider the possibility that broken vessels often continued to be used in a primary or secondary role.

348 The vessel is similar to one recovered at Salamis identified as used for making cheese or bread (Soren et al. 1988, 175 c.f. Diedrichs 1980). While the Earthquake House vessel would have been suitable for this purpose, in all probability its function was much less specific, being utilized for a multitude of daily activities that may or may not be associated with food preparation.

76

A total of three ceramic vessels from Room 14 would have been utilized for serving food; one fineware bowl (P133-1934), a CRSW dish (V622), and a single plainware jug (V618). All are Class 1b. The vessel recovered during the Trench III excavation (P133-1934) (Fig. 6.38: D), could not be located for examination. It is described by Daniel as a “red bowl,” but is identified in a subsequent notebook entry as a Late Roman B shallow bowl with a spreading rim.349 The second fineware vessel (V622) is a CRSW Hayes Form 1B dish (Fig. 6.41)350 found in the central part of the room. The fact the vessel was unbroken, right-side up, and sitting upon the occupation surface when discovered led the excavators to conclude that it was in this location at the time of the earthquake. While this is possible V622 may have been stored on a shelf that collapsed during the earthquake; its condition and position a matter of chance rather than purpose. The last ceramic serving vessel is a Vasa Type 3G jug (V618) (Fig. 6.42) found near the center of the room close to V622.351 Although largely reconstructible, the rim and a portion of its neck are absent. The similarity between the missing portions of V618 and V619 is significant and may indicate both vessels were damaged but continued in use as part of the functioning room assemblage.352 While in the case of V619 this scenario appears certain, the situation is less clear for V618. This is because its findspot near the earlier Trench III may have resulted in the missing fragments being found and discarded. One of the most notable characteristics of the Room 14 ceramic assemblage is the large number of lamp fragments representing an Average EVREP 12.5 (Min EVREP: 10; Max EVREP: 15). Of greater significance than the actual number of lamps is their condition, with no examples able to be identified as Class 1. All are Class 2, which permits three inferences to be drawn: 1) Room 14 was used during hours of darkness with lamps providing at least some of the light required; 2) Lamps were not stored in this room, but brought in when needed and then removed; 3) When lamps were broken during use, they remained in the room as provisional discard.

Figure 6.39: V615 – KT II amphora. This Class 2 material was deliberately collected in the small spaces between the sides of the oven and the walls as provisional discard where it awaited ultimate disposal. Within this collection area, a conspicuous difference among the Class 2 ceramics can be observed. Those near the north end of the east side are more intact, exhibiting moderate completeness and low brokenness with many larger fragments present. Based on their location, it is probable that this material was placed intentionally so as to be readily accessible “clutter refuse” with potential for future re-use. In contrast, the material from the south side of the oven possessed moderate to low completeness and high brokenness. These characteristics likely indicate they had been in discard for an extended period of time, causing increased fragmentation and dispersal.353 These

As previously mentioned, the Class 2 material from Room 14 provides valuable evidence concerning the disposal of household waste; a little studied and poorly understood aspect of life in the ancient world. A significant quantity of ceramic was recovered around the east and south sides of the oven. Unlike the other Room 14 vessels that are Class 1, those from this area exhibited low to moderate completeness and moderate to high brokenness. Based on these characteristics, they could not have been in use at the time of the earthquake and instead represent objects broken while being used in the room. 349

DFN: 6/4/1934, 22. Molinari et al. 1988, 175; Meyza 2007, 46-7, pl. 2. 351 For Vasa jug typology see du Plat Taylor 1958, 28-35. 352 Molinari et al. 1988, 175. 350

353 Fragments of V658 were recovered in the west and northwest parts of Room 14.

77

Figure 6.40: V619 – Vasa Type C amphora.

Figure 6.42: V618 – Vasa Type 3G jug. groundstone is a heavy, dense, fine grained stone (probably Gabbro) with a roughly oval shape. The primary work surface has been worn flat with several secondary surfaces as well as pecking/pounding areas able to be identified. The findspot of this object, within the provisional discard area south of the oven is curious. However, an in situ iron nail in the Wall Q masonry may indicate the presence of a shelf in this general area that could be where this object was stored. The collapse of this feature during the earthquake would have caused UR: BSK 81 to fall and come to rest in the location where it was found. Based on this evidence it is identified as Class 1c. The small stone mortarium (Fig. 6.38: E) recovered in the northeast corner by Daniel is likely Class 1b, although because of its size, Class 1c cannot be definitively excluded. In his field notebook, Daniel records that the object was “rough-hewn of soft sandstone,” with a “flat base.”354 Based on the field photograph,355 the vessel is shallow with two rectangular “lug” handles on either side.

Figure 6.41: V622 – CRSW Form 1B dish. smaller fragments would have been unsuitable for reuse except in very specific circumstances. Analysis of this material combined with that from the adjacent Room 12A and Room 19 provided definitive evidence to reconstruct a portion of the domestic waste stream during the final phase of the structure. This will be explored in detail in the following chapter.

354

DFN; June 4, 1934; p.22. The identification of sandstone is curious and probably incorrect, as all other objects of this type appear to be made of local limestone. The assessment of Daniel cannot be discounted, however, since the object could not be located for examination. 355 Like the majority of the ceramic objects recovered from this area, ST13-1934 could not be located for physical examination.

Two stone objects were recovered from Room 14: a small groundstone tool, likely a hand grinder (UR: BSK 81), from the southeast corner of the room; and a small stone mortarium (ST13-1934) from the northeast corner. The 78

It appears to be a similar, but smaller example of the mortarium (S75/S86/S86) from Room 8.

its condition and value that this vessel was Class 2 slated for eventual discard. As with the glass vessels, M193 almost certainly represents a Class 1c object stored on a shelf that was dislodged during the earthquake.

The ceramic serving assemblage was augmented by two glass bowls (G147 and G148) and a copper alloy pitcher (M193). Both glass vessels appear to have been flaredrim bowls, although they are too badly broken for a secure identification. They were found next to one another and were probably Class 1c, stored on a shelf on Wall T. The collapse of this fixture caused both vessels to fall and be crushed on the occupation surface.356 The rim of G147 (Fig. 6.43) was preserved and consolidated in situ, but a substantial amount of G148 is missing, suggesting that it may be Class 2 material. However, given its proximity to G147 and the lack of any other significant glass finds in the room, it is likely that both vessels shared Class 1 status. The lack of completeness exhibited by G148 may result from a combination of two factors: 1) recovery and discard of fragments during the 1934 excavation, which is probable given its proximity to the edge of Trench III; 2) differential recovery exacerbated by extreme friability due to the alkaline nature of the soil.. Both could have caused fragments of the vessel to be unaccounted for, giving it the appearance of a lower level of completeness than it actually possessed.

A number of smaller finds were also recovered in Room 14, among which are four coins from the occupation surface that are Class 2. Three were found south of the doorway along Wall D and one in the northern part of the room near G147. The coins range in size from 1.1-1.6 cm in diameter and were badly corroded, which prevented their identification.358 They almost certainly represent objects accidentally dropped/lost in the room with the three along Wall D probably collected and deposited with other Class 2 material as a result of periodic sweeping/cleaning of the space. A small copper alloy fishhook (M198) was found in the northwest quadrant in the area affected by Trench III.359 This raises the possibility that the object was not in situ, but contamination introduced from an unknown context. Because of this, the object cannot be classified with any degree of certainty. Given its size, M198 would not have been stored on the occupation surface making it Class 1c if it was, in fact, part of the assemblage. The last small find is a bone object (B58) (Fig. 6.44) fashioned from the long bone of a small animal with a joint forming the proximal end. Based on its characteristics, it is probably an awl, punch, or some other utilitarian tool used for a similar purpose. It was recovered east of the oven near the provisional discard area, possibly indicating that like the majority of the material found here B58 should be identified as Class 2. However, a storage situation similar to that of M193 offers an alternative designation of Class 1c.

The copper alloy pitcher (M193) is similar to those found in Room 1, Room 8, and elsewhere on the Kourion acropolis in later 4th-5th century AD contexts.357 Although the vessel is a Class 1 object, it was found along Wall D among the Class 2 material in the provisional discard deposit. Despite its findspot, it is highly improbable given

6.4.9: Room 19 The bulk of the Room 19 assemblage was recovered from the lower levels of Locus 004, all of which is Class 2. This locus produced a moderate amount of broken ceramic (Average EVREP = 34.5), animal bone (ca. 110 fragments) and an unidentifiable coin (C406). No reconstructible or partially reconstructible vessels were identified. The characteristics of this material are consistent with domestic refuse deposits such as those observed in Room 12A. A detailed examination of the

Figure 6.43: G147 – Vessel rim and fragments consolidated in situ during excavation.

Figure 6.44: B58 – Worked bone object. Possibly an awl or punch.

356

Molinari et al. 1988, 175. Molinari et al. 1988, 175. A copper alloy pitcher similar to those of the Earthquake House was recovered by Christou during his excavations in the forum. Published as a color plate in Soren and James 1988. 357

358 359

79

Molinari et al. 1988, 175. Published in Soren and James1988, 128.

Room 19 material has shown that it was utilized as a location for caching household waste. Like Room 12A, at least some of the Room 19 material originated from and can be tied to the adjacent Room 14. Ceramic cross-joins between the Room 14 and Room 19 vessels further elucidate behaviors concerning refuse disposal during the final occupation of the structure. Data from Room 19 illustrating and confirming these discard processes will be analyzed in conjunction with that from Room 12A and Room 14 in the following chapter.

other parts of the structure and largely overshadowed by the three victims discovered in its northwest corner (Fig. 6.45). A small number of Class 1a objects were associated with the remains: two finger rings (one iron and the other copper alloy); and a bone hairpin. The rings are of the same basic shape, with a round, flat bezel. The thin band on the iron ring (UR: BSK 566) is largely missing (Fig. 6.46), likely resulting from corrosion during deposition rather than wear. The copper alloy ring (UR) is intact. It was found next to the left hand of the male individual and was almost certainly being worn at the time of death. Unlike its iron counterpart, the bezel of this ring is inscribed with a chi-rho symbol flanked by the

6.4.10: Room 20 In Room 20, the artifact assemblage is slight compared to

Figure 6.45: Room 20 Plan. Reconstructed by the author from AZLS: Room 20. 80

Figure 6.47: UR/BSK 1412 – Iron and CA tintinnabulum. Object is uncleaned. striking that no intact or reconstructible Class 1b/1c lamps were recovered.

Figure 6.46: UR/BSK 566 – Iron finger ring. Side and top views.

The only Class 1 vessel from Room 20 is a small glass juglet (UR: BSK 1407) located near the center of the space. Its findspot is notable and based on the tight clustering of the fragments it appears to have been sitting on the floor when the structure was destroyed. It is Class 1b.

Greek letter alpha on one side.360 The bone hair pin was located ca. 2 cm north of the female cranium, indicating it was in the individual’s hair at the time of death. Apart from these three items, no other personal objects were recovered. However, because the remains were blocklifted as a single unit, the occupational surface where they rested could not be investigated, permitting the possibility that other personal objects remain beneath them.

In addition to the rings associated with the victims, two other metal objects, a copper alloy lion paw (UR: BSK 1409),361 possibly from a lampstand/candelabra or table and a copper alloy and iron tintinnabulum (UR: BSK 1412) (Fig. 6.47) were found near the interface between Locus 004 and Locus 006 in the northeast corner. Both are from the same general context as lamp (UR: BSK 841) and mortarium (UR: BSK 1404) fragments that are identified as Class 2. While their context suggests Class 2, it is highly probable that the tintinnabulum was a Class 1c object employed as an apotropaic device at the time of the earthquake. This interpretation is particularly compelling given the presence of an infant among the victims.362

The Room 20 ceramic assemblage is limited, consisting predominantly of Class 2 material confined to the margins of the space. One possible exception is a large amphora of indeterminate type (P687) found west of the Wall T doorway. Its completeness indicates this vessel could be identified as Class 1b, but curiously its bottom (which is intact) was recovered at a higher level than the neck. This suggests that P687 was either Class1b and upside-down, or Class 2 located on the margin of the room. No evidence of post-earthquake disturbance is recorded and based on the character of the other ceramic from this space the latter scenario is the most probable. Since Room 20 appears to have been a living space, it is

361

This object could not be located and was not examined. It is possible that this object was suspended from a roof beam. However, no elevation data was found in the excavation notes. The function of this object will be discussed in the following chapter.

360

362

The omega on the other side is not visible although it almost certainly existed. The iron specimen is too corroded to determine if it was inscribed. Published in Soren 1988, 30.

81

Figure 6.48: Room 25 Plan. Reconstructed by the author from AZLS: Room 25. 82

Only one possible coin was found, which is striking given their ubiquity in other parts of the structure, especially those containing human victims (Room 1 & Room 2). This absence of coins may indicate that the victims in Room 20 were of a lower socio-economic status than the individuals recovered elsewhere in the structure. 6.4.11: Room 25 Although a significant amount of material was present in Room 25 at the time of its destruction, the findspots of only three objects can be reconstructed with any degree of confidence (Fig. 6.48). The reason for this is that no plans of this space could be located and no findspot descriptions or provenience measurements were recorded in the notebooks. Findspots that can be reconstructed are based on field sketches made on the backs of provenience data tags. Since one-half of the room was excavated during the 1986 field season with the other completed the following year, the provenience tag dates provide the means to determine which half of the room specific objects were located.363 The three objects with known findspots are ceramic vessels in the northeast corner along Wall AG. At the intersection of Wall V and Wall AG, a complete and unbroken KT III amphora (V679) was recovered, which is Class 1b. Next to this was a largely reconstructible Dhorios Type P132 cookpot (V678), almost identical to the example recovered in the northwest quadrant of Room 8. This vessel appears to have been unused, possessing no sooting or other evidence of thermal exposure. A significant portion of the vessel is missing, however, and a search of the Room 25 sherd material failed to produce any additional fragments. This lack of completeness necessitated designating V678 as Class 2. The third is a fineware vessel (V677) identified by Meyza as a CRSW

Figure 6.50: P684 – Two-handled pithos/amphora. Hayes’ Form 12A decorated jar (Fig. 6.49).364 Originally the vessel possessed a high neck. At some point this was deliberately removed, probably mainly for aesthetic reasons although inhibited functionality may have also been a factor. The edge is broken around the circumference, much too evenly to have been damage caused by the earthquake. In addition to the missing rim, the vessel also has a large hole in its side. It appears that the rim removal and side damage occurred prior to the earthquake since no additional fragments belonging to this vessel were recovered during excavation. Although highly complete, the hole in its side would not permit the vessel to function in any meaningful capacity. Based on this, V677 is identified as Class 2. The remainder of the Room 25 ceramic consists of a variety of storage vessels. On the east side of the room, the vessels include a fully reconstructible KT III amphora (P683), a partially reconstructible KT III amphora (UR: BSK 561), and a large two handled pithos/amphora (P684) of indeterminate type (Fig. 6.50). Two of these (P683 and P684) can be identified as Class 1b. The status of the third vessel is less certain, as a portion of the neck and upper shoulder was not recovered. If the vessel was intact at the time of the earthquake, it could have been Class 1b with diminished capacity, although Class 2 is also possible.

Figure 6.49: V677 – CRSW Hayes Form 12A jar. Note missing rim removed by maintenance processes. (Undamaged side shown).

On the west side of the room, one KT III amphora (UR:

363

Based on the find locations recorded on the tags, it appears that the eastern portion of the room was excavated during 1986, with the western half completed in 1987.

364

83

Meyza 2007, 76.

Figure 6.52: S110 – Pestle. either stored with it or on a nearby shelf (Class 1b/1c). This support subsequently decayed, leaving the mortar at the approximate upper limit of the earthquake debris layer. While it is possible that S109 and S110 were Class 4 objects, this seems unlikely since their creation represents a substantial labor investment and both were still capable of functioning. The similarity of these two objects, as well as their proximity to one another, strongly suggests that they form part of the Class 1 assemblage of the room. The fact that both were located in the eastern half of the room along with Class 1 ceramic provides additional support for this interpretation.

Figure 6.51: S109 – Mortar. BSK 255) and a VTC amphora (UR: BSK 257) were recovered, both of which were only partially reconstructible. Fragments of the latter were spread among a number of baskets collected from this half of the room, indicating breakage and dispersal prior to the earthquake. Its overall completeness permits the KT III as possibly being identified as Class 1b with diminished capacity or as Class 2. This is not the case with the VTC amphora, which based on its completeness can only be Class 2. A large, heavy pithos (UR: BSK 236; 252; 564), with fragments found in both halves of the room, presents a similar situation to the VTC amphora (UR: BSK 257). Although a significant amount of the vessel was present and the fragments were partially reconstructible, its overall completeness indicates Class 2 status. Apart from the ceramic material, the only other significant objects recovered as part of the Room 25 assemblage are a large mortar (S109) (Fig. 6.51) and a pestle (S110) (Fig. 6.52). Both were made from finegrained black and white granite,365 which suggests that they formed a “matched set.” The mortar is large with a hemispherical work surface. The pestle is roughly cylindrical, although a noticeable thickening occurs toward one end. While both ends of the pestle are rounded, the larger appears to be the primary work surface. A secondary work surface is also present on this end, where a wide, shallow depression gives the object an asymmetric appearance. The notes record that both objects were found “high” in the room fill. This is not problematic since the mortar could have rested on an object of perishable material (Class 1b) with the pestle 365

A second pestle (S111) was also recovered from Room 25 but its probable elevation indicates it was deposited post-earthquake. In addition, S111 is of a completely different material than the other two objects with much of its work surface broken away.

84

Chapter 7: Activities and Aspects of Household Organization The previous three chapters have presented a detailed examination of the architecture, stratigraphy, and artifact assemblage from the Earthquake House. These data sets can now be brought together to create a comprehensive picture of the structure during its final phase. The purpose of this chapter is to synthesize the archaeological evidence preserved in the Earthquake House and exploit it to investigate the range and types of activities that occurred in each of its spaces. This analysis will first determine probable activities and behaviors that occurred in individual rooms followed by an examination of their distribution throughout the entire structure. The latter analysis lead to the discernment of behavioral patterns that reveal how the domestic environment of the Earthquake House was organized at the time it was destroyed, including the probable number of residential groups occupying it.

Type C storage has been observed in several locations throughout the structure. It consists of complete and partially complete vessels; the latter most likely being retained as clutter refuse. In situations where an object could be assigned to more than one of these three categories based on the evidence it was designated as “Type Indeterminate.” In order to establish the quantity of Type A storage in the Earthquake House, it was necessary to calculate individual volumes for Class 1 pithoi and amphorae. This was done using the stacked/summed cylinder methodology, which visualizes the interior of each vessel as a stack of thin cylinders.366 Individual cylinder volumes were calculated and then added to provide an Following this estimated total vessel capacity.367 determination, individual Class 1 vessel volumes were combined to approximate the total Type A storage capacity for specific rooms and the structure as a whole. A summary of individual vessel capacities can be found in Appendix C, Table 2.

7.1: Methodologies and Calculations Used in Analysis Specific components of the Earthquake House assemblage possess the ability to illuminate two key behaviors that the inhabitants engaged in during the structure’s terminal phase: storage and household waste disposal. Since both required specialized methodologies and calculations in their investigation, before proceeding it is necessary to discuss the procedures utilized in analyzing the data.

7.1.2: Discard In the Greco-Roman world, discard behaviors have been examined generally at the settlement level, but largely overlooked at the household level until relatively recently.368 Archaeologically, two primary lines of evidence for investigating the domestic waste stream exist, the architectural and the artifactual.

7.1.1: Storage Storage is one of the most prolific domestic activities that can be archaeologically documented, normally as one of three types: 1) storage of consumable items (e.g. grain, pulses, oil, wine, etc.) often kept in specific vessels (e.g. pithoi, amphorae, etc.); 2) storage of in use (life assemblage) objects (e.g. ceramic, metal, glass, etc.); and 3) storage of items for the purpose of reuse, recycling, and/or discard (death assemblage). These three categories have been designated by this study as Type A, Type B, and Type C respectively. Type A storage has been documented throughout the Earthquake House and is of primary interest since it forms a fundamental line of evidence for investigating the domestic economy, particularly questions surrounding socio-economic status and overall levels of household self-sufficiency.

Architectural Evidence: A number of ethnographic studies have examined questions concerning the relationship between architecture and the discard of household waste.369 This research has shown that refuse generated in domestic settings is normally cached as provisional discard within the structure prior to being removed and transformed into definitive discard.370 In the Greco-Roman world, this behavior often involved placing household waste into a built feature within the structure, a process classified by Peña as a form of provisional discard.371 Such installations would be emptied periodically with the waste normally transported to a location outside the domestic space, although deposition in an out of the way

Evidence for Type B storage is difficult to identify since some objects may have been recovered in their use rather than storage location. In the Earthquake House, Type B storage would have consisted predominantly of objects on shelves or, more simply, nails from which they were hung. No clusters of objects indicating the possible presence of cupboards, cabinets, or chests were identified. Larger ceramic vessels and/or heavy objects recovered on the ground/occupational surface may also represent objects in Type B storage, although in many cases this can be virtually indistinguishable from their place of use.

366

Nelson 1985, 312. The calculations employed are based on measurements collected during object assessment. 367 A headspace of 2-3 cm from the vessel rim was allotted for each specimen. 368 For the Greco-Roman world see especially Dupré Raventós and Remolà 2000; Libeschuetz 2000; Guidobaldi et al. 1998; Robinson 1993; Scobie 1986; Owens 1983. For general discussion see Needham and Spence 1997; Halstead et al. 1978; Hoffman 1974; Green 1961a, 1961b. 369 See Hayden and Cannon, 1983; Kamp 1991; Deal 1985. The models for these behaviors were developed based on ethnographic data collected from contemporary groups, predominantly in the New World. 370 E.g., Hayden & Cannon 1983; Kamp 1991. 371 Provisional discard installations have a variety of names in the Roman world including latrina, lacus, and stercilinum (Peña 2007, 308). In the Greek world, the term used for these types of installations is kopron (See Owens 1983 and Ault 1999 for discussion).

85

place within the structure has also been documented.372 Based on these ethnographic observations, the plan of the Earthquake House was examined for architectural features or spaces that could potentially be used as areas for this activity.

because of the nature of the assemblage from the Earthquake House, overall levels of completeness and brokenness were assessed based on comparison to complete/fully reconstructible Class 1 objects. The following procedure was used to identify probable discard deposits:

One characteristic observed in discard deposits is an elevated concentration of artifacts, often of multiple classes in a damaged or fragmentary condition within a relatively confined area. In order to compare concentrations of artifacts within locations of probable discard to the surrounding space, it was necessary to obtain three calculations for rooms where such deposits were believed to exist:

I. Findspots of all Class 2 ceramic material within a room were examined as thoroughly as possible to identify areas of apparent concentration. II. All material from these concentrations was examined to determine that it was definitively Class 2, and could not be assigned to any Class 1 vessel. III. Based on the findspots, the boundaries of the possible discard space were determined and its area calculated. IV. The EVREP within the possible discard area and the remainder of the room were determined. These were then divided by the TAA and TDA respectively to calculate the Broken Vessel Density (BVD) for each. This figure provides the number of Class 2 vessels per square meter. V. If a significant difference was discovered between the BVD calculations, it was determined that the area of concentration had a high probability of being used for the disposal and/or caching of household waste. To provide confirmation, the density of other materials, specifically animal bone, was calculated to see if a pattern similar to the ceramic evidence was revealed. In cases where similar concentration values emerged, a discard area was definitively identified. VI. The final step was to actively seek out definite and probable cross-joins between discard loci to identify and trace the household waste stream.

I. Total Usable Area (TUA) – Calculated by determining the total area of the room and subtracting the amount of space occupied by any built features. II. Total Discard Area (TDA) – The amount of space that was utilized for discard. III. Total Available Area (TAA) – The amount of open and unoccupied space where daily activities could be carried out. It was calculated by subtracting the space occupied by Type 1b objects as well as any area determined to have been utilized for the disposal of household waste (TDA) from the TUA.373 Based on the results of these calculations, the TDA could then be quantified as a percentage of the TUA and TAA, indicating how much space was utilized for discard activity. These figures also provided the means for determining the density of material recovered within each space, an essential component in confirming the presence of deliberate discard deposits. Artifactual Evidence: A characteristic of ceramics from well preserved domestic waste deposits is their relatively intact nature. One of the best published examples is a pit excavated at Gravina di Puglia in Southern Italy. The pottery was notable for possessing generally high completeness compared to other material combined with low brokenness.374 Unfortunately, the determination of these two characteristics is subjective and to the knowledge of the author, no studies have attempted to establish definitive values for the objective evaluation of completeness and brokenness.375 As stated previously,376

7.2: Individual Room Activities: Using the above methodology and assemblage data, individual activities carried out within each room of the Earthquake House were determined and examined. Recent assemblage studies have confirmed the multifunctional nature of ancient domestic space. This spatial flexibility is characterized by a “mixing” of the artifactual material, commonly reflected by the presence of artifacts used for two (or more) different activities in the same architectural space and evidence for the same activity occurring in multiple locations within a structure.377 Both of these situations can be observed in the Earthquake House. In a few cases it is possible to assign specific functions to rooms based on the architectural and artifactual evidence, such as Room 2 being a stable and Room 14 being a kitchen. These two spaces, however, are the exception rather than the rule. For the most part, it is only possible to identify activities and behaviors that took

372 See Peña 2007, 274-6 for a discussion of general domestic discard characteristics derived from ethnographic data. One of the primary studies of household waste installations of this type in the Greco-Roman world was conducted by Ault at Halieis in Greece. In several houses he identified a number of stone lined earthen floor pits filled with a variety of residual material consistent with household waste (Ault 1999, 550-7). 373 Available area utilizes the maximum diameter for each vessel of the in use assemblage recorded during the assemblage evaluation. 374 Small et al. 1994, 206-26. 375 Although Orton (1989, 97) provides formulae for the calculation of brokenness and completeness: (Brokenness = Sherds/Estimated Vessel Equivalency (EVE); Completeness = EVE/Estimated Vessels Represented (EVREP)), he offers no correlation as to what the results obtained from this formula should equate to in terms of physical

description (i.e. “X Value = Partially Broken,” “Y Value = Very Complete,” etc.). Supra p. 30, # 194. 376 See Chapter 3, p. 30. 377 Nevett 2010, 98; Oswald 1987.

86

87

Figure 7.1: Artifacts reflecting storage in Room 1, Room 2, and Room 3. Reconstructed by the author from DFN: 6/6/1934, 27and AZLS: TRIII/Room 1,AZLS: Room 2, and AZLS: L9/Room 3.

Figure 7.2: Provenienced coins from Room 1, Room 2, and Room 3. Reconstructed by the author from DFN: 6/6/1934, 27and AZLS: TRIII/Room 1, AZLS: Room 2, and AZLS: L9/Room 3. place within each space, with no specific overall room function able to be determined. The evidence for these activities was used to discern patterns within the assemblage, permitting the formulation of hypotheses and interpretations about how each space was utilized by the inhabitants. Finally, the individual room observations were brought together to derive spatial use patterns within the structure as a whole.

discard processes (e.g. deliberate collection in a single location). The majority of the Room 1 coins (Fig. 7.2) were recovered above the occupation surface in the earthquake debris layer, indicating deposition peri- and/or postdestruction. The coins appear randomly scattered with no central point of origin, suggesting that they were not contained in a vessel located in the room, but probably fell from an upper floor.

7.2.1: Room 1, Room 2, and Room 3 (Fig. 7.1) The assemblages from these three spaces present many challenges in their interpretation. In assessing the Room 1 material, it is necessary to keep in mind that the picture is incomplete since only cursory notes were made and not all material was retained from the Trench III excavation. The only activities that can be positively identified in this space are Type A and Type B storage. The first is attested by Pithos P1 and Vessel P134-1934, which combined provided an overall capacity of ca. 44-50 lt.378 Evidence for Type B storage is largely indirect, with shelved objects likely represented by the copper alloy jug UR1934 and possibly lamp L47-1934. Type B storage on the occupation surface may have been represented by the cooking vessel complete with lid (UR-1934) discovered during the final stages of the Trench III excavation.

One space in the Earthquake House that did have a specific function when it was destroyed was Room 2, which was utilized as a stable. The only activity discerned is Type B storage along Wall D, where a group of iron tack components (M112, M116 and M117) were probably suspended from a nail or hook. As in Room 1, the Class 2 material does not indicate any type of organized discard behavior. In spite of its certain utilitarian function, the Room 2 assemblage includes two luxury objects, a marble tabletop and a copper alloy lampstand, as well as the largest quantity of coins found anywhere in the structure (Fig. 7.2). The majority of these coins were recovered in the earthquake and post-earthquake debris layers, a situation analogous to that observed in Room 1.

Other activities such as food processing and preparation in Room 1 could be suggested, but ultimately the evidence is inconclusive. The Class 2 material from the room appears to be largely incidental and does not indicate it was subjected to any type of organized

Room 3 was also utilitarian in nature and contained a significant amount of Class 2 ceramic. Both the quantity and size of this material would have been conducive to reuse and/or recycling, making it likely that it was cached here deliberately in Type C storage. Based on the recovery of joining fragments from other parts of the structure,379 however, at least a portion of this material

378

Other vessels recovered from this room that are typically associated with Type A storage (P22/28/153, P221, P135-1934) have been excluded from the calculations due to their indeterminate status (Class 1 or Class 2) at the time of the earthquake.

379

88

Most notably the pithos rim from Room 11 (Supra p. 63-4).

led to a second story.381 An analysis of the Pella material indicates that the upper floor served as the main habitation area, with ground floor spaces used for stabling animals, storage, and light workshop activities.382 This same arrangement has also been documented ethnographically in Cyprus where living spaces in traditional houses are located close to, and normally above, those sheltering animals so that their radiant heat can provide warmth for the inhabitants.383

represents provisional discard awaiting transfer elsewhere. In all probability, the Room 3, Class 2 material is some combination of provisional and definitive discard, which in this case cannot be differentiated based on the archaeological evidence. The Room 3 Class 1 artifacts may represent objects in Type B or Type C storage, although in the case of Vessel 187, Type A storage of 6-7 lt. cannot be excluded. However, based on the find context of this vessel combined with its small capacity, its use for Type A storage is unlikely.

Based on the traditional Cypriot techniques used in its construction,384 the Earthquake House upper floor would have been constructed predominantly of perishable materials (e.g. wood and mudbrick) whose collapse debris would be largely indistinguishable from that of the ground floor. Because of this, evidence of its presence would not have been readily distinguishable during the excavation. If the roof beams had been weakened through insect action, something known to have occurred in the adjacent Room 8,385 the upper floor would have likely collapsed toward the center, spilling its contents into the spaces below. As the structure continued to degrade over time, objects from the upper floor would continue to fall causing them to be recovered higher in the postearthquake fill layers. The same pattern of evidence, particularly regarding the coins, was observed at the “Patrician House” at Meiron, Israel. Here, a significant number mid-4th century AD coins were discovered in the collapse loci. This led the excavators to conclude that they “had been stored away (in a bag perhaps?) and forgotten in the second floor living quarters and then were scattered in the debris when the second story collapsed.”386 The presence of an upper floor also explains why the human victim from Room 2 was found over the remains of the mule; having fallen from the upper floor when it collapsed beneath her.387

Despite the fragmentary and scattered nature of the Room 3 ceramic, an anomalous “blank spot” devoid of artifacts except for very small ceramic fragments is present in the southwest corner. No evidence of post-earthquake disturbance that could account for this was recorded during excavation, which seems to indicate a differential use of space in this part of the room. One possible explanation is that this area was used for Type A or Type B storage of ephemeral material (e.g. grain in sacks or bags, animal fodder, blankets, etc.) that left no trace in the archaeological record. Like Room 2, Room 3 preserved a significant number of coins (Fig. 7.2) as well as a copper alloy cosmetic tool and silver-sheathed bone hairpin, the latter two representing personal luxury items. Since these objects and a significant number of the coins were recovered within the earthquake debris layer, it is probable that they represent material originally on an upper story that fell when the structure collapsed. Much of the material from Room 1, Room 2 and Room 3 stands in sharp contrast to that found in the remainder of the structure. Architecturally, these three spaces form a distinct unit and preserve evidence for significant monetary and non-monetary wealth, including the majority of “luxury” objects recovered. The utilitarian nature of these spaces only serves to highlight the dissonance between objects such as the bronze lampstand and marble tabletop and their findspots. Based on the available evidence, the elite objects and coins from Room 1, Room 2, and Room 3 almost certainly belonged to an assemblage located on an upper floor/second story over these three spaces. Two lines of evidence, archaeological and ethnographic, can be used to support this hypothesis.

Although the evidence as it stands does not absolutely prove the existence of an upper floor, collectively it provides a compelling circumstantial case for one above Room 1, Room 2, and Room 3. 381

Fejfer and Mathiesen 1995, 85; Meyers et al. 1981, 55-8; Walmsley 2007, 251. 382 Walmsley 2007, 253, 264. House A and House B have rooms equipped with feed bins indicating they were used for stabling animals. House G preserved the remains of a number of animals killed during the earthquake. In the “Patrician House” at Meiron, Israel, although no provision for stabling animals was found, based on the artifactual evidence the ground floors appear to have been used primarily as domestic workspaces with the primary living quarters being above (Meyers et al. 1981, 59). 383 Davis and Soren 1985, 297; Soren and James 1988, 117. 384 As discussed previously in Chapter 4 (p. 31-2). 385 Soren et al. 1986, 202; Soren and James 1988, 115. 386 Meyers et al. 1981, 54. 387 Soren and James 1988, 90. Soren states that the girl was “most likely still alive and crawling through the rubble” following the first wave of the earthquake. However, the remains were found over the hind quarters of the mule, which would have been more likely to struggle through the debris. As the first wave is believed to have only lasted ca. four seconds before the second (Soren 1988, 37) it is hard to accept the timing of this scenario. The position of the human remains strongly suggests that the girl may have been on an upper floor that collapsed. Unfortunately, the condition of the remains precludes any type of forensic evaluation that could substantiate this hypothesis.

Archaeologically the evidence can be subdivided into two categories: artifactual, which has been discussed above; and architectural, embodied in the L-shaped staircase in the northwest corner of Room 15. The excavator believed that the primary function of this stairway was to provide access to Room 14 and the cistern portion of Room 12; the latter accessed from above since it was not equipped with a drain.380 While the staircase was no doubt used for these purposes, in late antique/early Byzantine houses excavated at Agios Konanos in western Cyprus, Meiron in Israel, and Pella in Jordan, stairways virtually identical to that of the Earthquake House almost without exception 380

Molinari et al. 1988 173, 177.

89

Figure 7.3: Artifacts reflecting storage in Room 6 and Room 7. Reconstructed by the author from AZLS: Room 6 and AZLS: Room 7. 90

7.2.2: Room 6 and Room 7 (Fig. 7.3) During their final phase, Room 6 and Room 7 formed a separate architectural unit. Initially, these spaces appear to have been part of the Earthquake House, but became separate by sealing the Wall K doorway linking them to Room 8.388 In examining the Room 6 and Room 7 artifacts, it is clear that the two spaces were materially linked with each having a distinct role.

storage. The majority of the remaining Room 7 assemblage seems to represent material in Type C storage that had been subjected to systematic behaviors of reuse and recycling. Distribution of the amphorae within Room 7 is notable, with like types being grouped together. Of these nineteen vessels, seven (37%) (five KT Is and two KT IIIs) are represented only by diagnostic elements. In these cases the lack of associated “body” material is striking, suggesting a deliberate and organized set of behaviors that broke down and recycled/reused the vessel walls, likely for chinking in wall construction.391 Following their cannibalization, diagnostic elements from these vessels appear to have been retained in Type C storage either as clutter refuse392 or as waste byproducts held in provisional discard. In contrast, the KT V/LR1 amphorae possessed a high degree of completeness, with only the very bottom portion of the vessel missing in the majority of cases. This difference reveals that KT V/LR1s were not subjected to the same systematic recycling/reuse behavior practiced on the other types. The reason for this may be due to their physical characteristics; possessing thinner walls and more brittle fabric that made it less desirable for recycling.393

Room 6 served as a locus for multiple activities, most notably the trans-vasing of liquid commodities. This is demonstrated by the presence of a funnel (Vessel 520), as well as at least one amphora that had been deliberately holed (474-1) to facilitate emptying. Based on the number of Class 1 vessels, Room 6 was also used for Type A storage of liquid and possibly non-liquid commodities. The latter, however, cannot be confirmed since no botanical evidence was recovered during excavation. The ten Class 1 vessels389 provided the room with an estimated Type A storage capacity of 362-377 lt. Glass vessels (G138 and G139) from the southeast corner may also represent Type A storage, being employed for small amounts of possibly more expensive commodities such as herbs and spices. Type B storage is likely represented by the cookware pitcher (Vessel 445) and partially complete cookpot (Vessel 404/444), which likely occupied a shelf with the glass vessels and the funnel.

The overall character of the Room 7 ceramic indicates that vessels were intentionally retained and cached in this space awaiting reuse and/or recycling. Although some of the remaining fragments may not have been intended for reuse (handles, spikes, etc.), forming provisional discard, the lack of material from other artifact classes (e.g. animal bone, glass, metal, etc.) points toward Room 7 being employed specifically for the storage of ceramic material and not for the general caching of household waste.

Three examples of Type C storage were identified in Room 6: the cluster of three vessels along Wall K (505, 508, V532) that appear to represent deliberate segregation of Class 2 material in this area; two KT I amphorae (4741, 523) on the west side of the room; and a sherd concentration in the northeast corner of the room near the entrance to Room 7. The final instance produced crossjoining fragments from Vessel V532 and Vessel 523, which confirms the identification of these vessels as Type C. A Class 1 KT I amphora (V478) with the same general context and findspot as the sherd concentration makes it highly probable that this vessel is also Type C.

When examined in conjunction with the Room 7 evidence, the findspot of the Type C sherd concentration and KT I amphora (V478) at the northern edge of Room 6 supports the movement of Class 1 and Class 2 material from Room 6 into Room 7. Here, it would have been retained in Type C storage until subjected to either reuse or recycling. Given the evidence for this transferal, it is logical and probable that the Room 7 vessels originated predominantly, if not wholly, from activities in Room 6. Recycling would have involved cutting up vessels for wall chinking, which, based on the Room 14 evidence,

As described in the previous chapter, the Room 7 material is predominantly amphorae of three distinct types, KT I, KT III, and KT V/LR1. Four Class 1 vessels390 could have been utilized for Type A storage, providing an additional capacity of 98-102 lt. This would have increased the total Type A storage capacity of Room 6 and Room 7 to 460-479 lt. Based on the high proportion of Class 2 material in this room, however, it is unlikely that these vessels were used for this purpose. Rather, they almost certainly represent surplus material in Type C

391 See Peña (2007, 257-8) for evidence regarding amphorae fragments used as construction elements in walls. The fabric of these amphorae is consistent with much of the chinking utilized in the walls of Room 14. The large, relatively flat body surfaces created by the shape of these vessels would have been well suited for this purpose. 392 Slane 2004; Sullivan 1989. Necks could have been utilized as funnels; bases as vessel stands, bowls, or dishes. 393 A field interpretation put forth during the excavation was that the bottoms of these KT V/LR1 amphorae had been intentionally removed so they could be used as a conduit to repair a water line (Tom Davis – Personal communication). Although this is theoretically possible, the diameter of the KT V/LR1 neck is quite small and would not provide a diameter wide enough to ensure an efficient flow of water. A number of preserved terracotta pipes recovered in the area of the structure would seem to negate reusing these vessels for this purpose.

388

For discussion of this division see Chapter 4 (p. 40-1). These include the pithos (V443), the VTC amphora (V476), one KT I amphora (V502), the single KT II amphora (480/500), and six KT III amphorae (474-2, V521/533, P522, V403, 534/541, V542). 390 These include a single KT I amphora (V551), two KT III amphorae (V519, V536), and the single KT V/LR1 amphora (V546). Three KT I amphorae (V477, P510, V535) could possibly have been used for Type A storage at reduced capacity (ca. 80%), which would have provided an additional 84-87 lt. However, given the presence of complete amphorae, it seems unlikely that these broken specimens were being utilized for Type A storage, but were cached in this space as Type C. 389

91

appears to be principally derived from KT I and KT III amphorae. Given this scenario, the lack of tools in Room 7 for these activities is noteworthy and possibly indicates they were either brought in when needed and removed or the amphorae were taken and broken down elsewhere.394 The lack of small amphora fragments and other detritus from such an activity in Room 7 points toward the latter scenario being more probable. Afterward, elements such as necks and bases with potential for secondary use (e.g. funnels, vessel stands, etc.) appear to have been replaced as “clutter refuse” awaiting future use or discard.

instance of Type B storage may be represented by the large mortarium (V578) found in the southern part of the room. However, questions concerning the identification of this vessel (Class 1 or Class 4) make Type B assignment problematic.397 In general, the Room 8 Class 2 ceramic does not suggest any type of organized discard activity with the possible exception of the pithos and VTC amphora (V568 and Vessel 594), which may represent Type C storage. An examination of this material did reveal a curious discrepancy in the southwest quadrant that appears to indicate this area was used differently than the remainder of the room. Overall, the southwest quadrant has a lower artifact density than the remaining three. This same pattern is also present in the BVD, which is relatively constant in the other quadrants but drops dramatically in the southwest. The calculated area of each quadrant is 10.39 m2. In the northwest quadrant the Class 2 EVREP is 41, which provides an density of 3.94 vessels/m2. Similar figures were obtained for the northeast (Class 2 EVREP: 44.5 = 4.28 vessels/m2) and southeast (Class 2 EVREP: 43 = 4.13 vessels/m2) quadrants. These BVDs are consistent with those calculated from other rooms of the structure398 but are significantly higher than that obtained from the southwest quadrant of Room 8. Here, the Class 2 EVREP is 9.5, providing an BVD of 0.91 vessels/m2, ca. 4.15 times less than the other three areas. The difference in BVDs between the southwest quadrant and the remainder of Room 8 is too remarkable to be coincidental and indicates that for some reason, deposition of Class 2 ceramics was significantly hindered in this area. Although it is possible the reason for this absence is behavioral (e.g. deliberate cleaning), it seems unlikely this would have been confined to the southwest quadrant alone. More likely, this disparity reflects some activity that caused the occupation surface in this area to be covered. One possible explanation is Type A storage in perishable containers (e.g. cloth or leather sacks, wooden bins/barrels, etc.) that would have prohibited deposition of Class 2 ceramics on the occupation surface while leaving minimal, if any, traces in the archaeological record.

By definition, the transference of material from Room 6 to Room 7 does not technically constitute a “waste stream” since the material was not intended for discard but reuse and/or recycling.395 However, it is a similar process that provides evidence for behaviors influencing the movement of objects from one space to another for purposes other than primary use. In this case, the transfer between Room 6 and Room 7 is a preliminary stage that occurs prior to material entering the waste stream proper, a “use stream,” where elements that can be utilized are, with the rest of the material discarded. 7.2.3: Room 8 (Fig. 7.4) As discussed previously, the Room 8 assemblage poses numerous challenges due to the unevenness of the occupation surface, the loosely sealed nature of the earthquake debris deposit, and probable post-earthquake contamination. Like many spaces of the Earthquake House, Room 8 was multi-functional, with several different identifiable activities. Based on the ceramic evidence, Type A storage in Room 8 was minimal with the only Class 1 vessel (KT III amphora [BSK 119]) having a capacity of ca. 25-26 lt. Two probable locations for Type B storage were identified. The first is represented by the copper alloy pitchers (M164 and M165), ceramic basin (V560), and object M174 in the northwest corner of the room. It is likely the pitchers were stored on a shelf (or multiple shelves) along Wall E and/or Wall K and that M174 was stored in this location as well. Based on its size and weight, Vessel V560 would have been stored on the occupation surface. The second Type B storage location is toward the center of the room, where a cluster of metal objects including the spearhead (M177/M180), fork (M175), tool (M191), and lead cup (M176) were recovered. Given their proximity to the stone table and central column, it is likely some of these objects may have been located on or around these features at the time of the earthquake. If the spearhead was hafted, it could have been resting against the central pillar and fallen to its findspot during the earthquake.396 A possible third

Evidence for small scale foodstuff processing and preparation is located in the northeast corner where a grinding station partially composed of reused architectural elements was set up. It is likely that the stone mortarium and ceramic basin (S75/S86/S87; V560) in the northwest quadrant of the room were also used for this activity. The presence of tablewares (V550, Vessel 579/601), a ceramic flask/bottle (V561), and the previously mentioned copper alloy pitchers indicate Room 8 was a locus for the consumption of food/drink. It is one of only two such assemblages for this purpose recovered in the Earthquake House ceramic. Tableware findspots towards

394 Although it is possible that the vessels were recycled without the use of tools, many of the edges of the remaining diagnostic fragments are quite even, much more so than would likely be expected if broken down by hand. 395 Schiffer 1996, 66-70. 396 This is likely a case of what Kent terms “informal storage.” She defines this as “the placement of objects on or in areas that are not specific in function for storage.” For discussion see Kent 1999, 80.

397

Supra Chapter 6, Room 8 (p. 72-3). See discussion of discard evidence in section 7.2.3 below for comparative BVD calculations. 398

92

Figure 7.4: Artifacts reflecting storage in Room 8. Reconstructed by the author from AZLS: Room 8.

93

the room’s center may indicate the stone table was the use and/or storage location for these objects. The surviving evidence does not support the consumption of food/drink in Room 8 on a large scale, instead reflecting meals by either a single individual or very small group. While based on the available evidence, this conclusion must be viewed cautiously since the surviving ceramic tablewares were almost certainly augmented by vessels and utensils of perishable material (i.e. wood) that have not been preserved; a fact that could alter the perceived intensity of this activity in Room 8.

identification, as with few exceptions the objects reflect activities pertaining to food preparation and consumption. The bulk of the ceramic assemblage would have been utilized for Type A storage, providing a total capacity of 267-281 lt. Although Vessel P665 could have been utilized for Type A storage, questions concerning its status at the time of the earthquake (Class 1 or Class 2) make a Type C identification as well. Type B storage is predominantly reflected by objects on shelves, namely the two glass vessels (G147 and G148) on Wall T, the copper alloy pitcher (M193) on Wall D, and the small hand grinder (BSK 81) on Wall Q. One exception is the ceramic basin (V620) in the southwest corner of the room, which represents Type B storage on the occupation surface. This may also be true for food service objects such as the fineware ceramics (V622 and P133-1934), the Vasa Type 3G jug (V618), and possibly the small stone mortarium (ST13-1934). However, the possibility also exists that the V622 and V618 findspots towards the center of the room reflect a use rather than storage location.

A probable activity area composed of the intentionally deposited sand layer (Locus 005), copper alloy proportional dividers (M183), and (possibly) the broken inverted stone basin (S84) can be identified in the southern part of the room along Wall J. The relationship between these objects, particularly the dividers and Locus 005, does not appear accidental, suggesting that they were used for a specific task. However, what this might have been cannot be determined at present. Room 8 formed the largest interior space of the Earthquake House and its artifact assemblage provides evidence for a wide variety of domestic activities that underscore its multi-functional nature. Based on this, it is possible that Room 8 was a communal space used by all the structure’s inhabitants and served as one of the primary foci for daily activities. However, its architectural isolation during the final occupation phase suggests that Room 8 could also have been autonomous, serving as a self-contained living space for one or more of the inhabitants.

Evidence for food preparation is attested by the aforementioned hand grinder (BSK 81), the small stone mortarium (ST13-1934), and the ceramic basin (V620). The size of BSK 81 and ST13-1934 indicates that they were likely used for small grinding tasks (e.g. herbs and spices) rather than large scale processing of grain, which would have occurred outside this space. Since the primary function of Room 14 was food preparation, the presence of only a single Class 1 cooking vessel is striking given the significant number of Class 2 cooking vessels in the provisional discard area. The reason for this is unknown and the general absence of Class 1 cooking vessels from the remainder of the structure does not indicate their storage elsewhere. Several possible explanations for this deficiency exist including keeping minimal cookwares on hand and procuring them only when necessary, and ready access to replacements; negating the need to stockpile spare vessels.

7.2.4: Room 11 Only two activities can be identified in the large courtyard of Room 11: food preparation and discard. The first occurred in the southwest corner where the large hemispherical oven was located. Discard appears to have been widespread over the entire area, since the final occupation assemblage consists almost entirely of Class 2 material, predominantly ceramics. While some of the ceramic fragments are large and possibly retained in Type C storage, the degree to which this activity was organized cannot be determined.

The final activity identified in Room 14 is the organized discard of Class 2 material, at least some of which probably represents “clutter refuse” in Type C storage. This was focused in the narrow spaces between the oven, Wall D, and Wall Q. Combined with the Class 2 material from the adjacent Room 12A and Room 19 these objects enabled a partial reconstruction of the Earthquake House domestic waste stream, which will be discussed below.

7.2.5: Room 12A During the structure’s final phase, Room 12 A was used as an area for Type C storage of Class 2 material. Study of the objects from this space in conjunction with those from Room 14 and Room 19 has provided evidence allowing a partial reconstruction of the domestic waste stream of the Earthquake House. A detailed discussion of the archaeological evidence used in identifying and reconstructing this household system is provided below in Section 7.3.3.

7.2.7: Room 15 and Room 18 Like Room 11, the Room 15 and Room 18 assemblage appears to consist almost exclusively of Class 2 material. This, combined with the fact that both spaces are among the most poorly documented in the Earthquake House, make identifying specific activities and behaviors problematic. The only activity that can be definitively identified in these spaces is cooking, evidenced by the

7.2.6: Room 14 (Fig. 7.5) In Room 14, the mudbrick oven indicates that this space functioned primarily for preparing food, allowing it to be designated a “kitchen.” The assemblage supports this 94

Figure 7.5: Artifacts reflecting storage in Room 14. Reconstructed by the author from AZLS: Room 14.

95

Figure 7.6: Artifacts reflecting storage in Room 25 and Room 25. Reconstructed by the author from AZLS: Room 20 and AZLS: L9/Room 25. two mudbrick ovens along Wall V on the eastern end of the portico.

low partition wall (Wall AE). Based on the presence of three victims, who may have been asleep at the time of the earthquake, Room 20 likely functioned predominantly as a habitation space.399 Distinguishing specific activities in this room is difficult, since the assemblage is much less robust than other rooms. Type A storage of 20-25 lt. may exist in the form of amphora P687, but questions regarding its status at the time of the earthquake (Class 1 or Class 2) make conclusively identifying this activity problematic. If P687 was not intact, it represents Class 2 material likely retained in Type C storage. Type B storage is equally difficult to identify, with the small glass juglet (BSK 1407) being the only possible object that may represent this activity. Given its size and fragility, however, it is unlikely that this object was stored on the floor. Its findspot, therefore, likely reflects it being in use by the victims when the structure was destroyed.

It is possible that the built feature (“basin”) in the northeast corner of the courtyard was utilized for Type A storage of some type. Unfortunately, the lack of a detailed description, drawings, and/or field photographs cannot presently substantiate this hypothesis. The presence of a few large Class 2 ceramic fragments throughout Room 15 indicates the possibility of Type C storage in this area. However, because of its poor documentation this material cannot provide insight as to the degree and organization of any discard behaviors that may have taken place. 7.2.8: Room 19 This small area formed a sottoscala utilized for Type C storage of Class 2 material at the time of the earthquake. The evidence from this space as well as Room 12 A and Room 14 has provided the means to partially reconstruct the domestic waste stream of the Earthquake House during its final phase.

No organized discard activity was distinguished in Room 20, other than Class 2 material being located along the

7.2.9: Room 20 and Room 25 (Fig. 7.6) Although excavated as separate spaces, the evidence indicates Room 20 and Room 25 formed a single architectural unit divided into two functional areas by a

399 The position of the victims may provide credence to Soren’s hypothesis that the earthquake occurred either at night or in the early morning hours (Soren and James 1988, 3). Nur and Burgess (2008,1502) provide an eerily similar comparison between the “Christian Family” of Room 20 with a family of three killed by an early-morning earthquake that struck Killari, India in 1993.

96

walls of the space. This makes sense given the primary function of the space as a habitation area.

Based on the findspots of the remains (Fig. 7.7), it appears that at least two and perhaps three distinct groups inhabited the Earthquake House when it was destroyed. The first group is composed of the three individuals found in Room 1 and Room 2 (two adults, one juvenile), who likely occupied the central part of the structure consisting of Room 1, Room 2, and Room 3 along with the upper floor over these spaces. Objects of personal adornment (particularly the gold ring), the mule, and the significant number of coins from these three rooms indicate that these individuals possessed a certain degree of wealth of a monetary and non-monetary nature.

During its final phase, Room 25 may have been used for Type A and Type C storage, which appear to have been spatially differentiated. The three Class 1 vessels in the eastern half of the room (P683, P684, V679) could have provided 85-88 lt. of Type A storage. Conversely, given that a significant number of Class 2 objects were located in this same area, it is possible that all three were in Type C storage – a situation similar to that observed in Room 7. A number of Class 2 objects were also identified in the western half of the room and in the northeast corner along Wall AG. Based on their completeness and brokenness, some may have been largely intact at the time of the earthquake making them likely candidates for retention in Type C storage.

Three individuals (two adults, one infant) comprise the second identifiable group, which occupied Room 20 and by extension, Room 25. In contrast to the victims from Room 1 and Room 2, no evidence of wealth was found associated with this group.402 The lack of evidence for any type of wealth is striking, especially when compared to the first group. This almost certainly indicates that the Room 20 occupants belonged to a lower socio-economic class than those found across the courtyard.

The presence of the mortar and pestle (S109 and S110) provides the possibility that Room 25 was utilized as a workspace for small-scale foodstuff processing, likely for the individuals in the adjoining space. Unfortunately, the absence of reliable provenience data for these objects does not allow their location to be accurately fixed. As a result, their use status remains ambiguous. Given the character of the other objects from this space, it is possible the findspot of these objects reflects their storage rather than use location. If this is the case, S109 and S110 would represent Type B storage, likely on the occupation surface, although the small size of S110 makes a shelf location possible.

The single adult individual from Room 11 could possibly belong to one of the aforementioned groups. Alternately, he could form a distinct residential unit based on the isolation of his remains and their discovery in an area largely separated architecturally from the remainder of the structure.403 No distinguishing artifactual material was found associated with these remains that conclusively points to either scenario. Based on the architectural layout and the distribution of the remains, each group appears to have inhabited a distinct area in the structure. This fact, combined with the artifacts associated with them suggests that they were discrete from one another.404 The relationship (e.g. kinship, master/slave, etc.), if any, that existed between these two or three groups of individuals cannot be determined based on the evidence presently in hand.

7.3: Aspects of Household Organization The picture of the Earthquake House that emerges from this room activity analysis is consistent with those from other domestic structures in the Greco-Roman world. Nearly every space preserves evidence for multiple activities, indicating their use was not rigidly defined but determined by the specific needs and preferences of the inhabitants.400 By examining the evidence for a number of basic domestic activities in the rooms of the Earthquake House, a more complete understanding of its overall organization can be reached.

7.3.2: Storage (Fig. 7.8) This is the most widespread activity identified in the Earthquake House, with most rooms possessing at least two of the three storage types identified by this study.

7.3.1: Residential Groups Questions concerning the residential group(s) occupying a structure are generally difficult to answer, since the inhabitants are normally not preserved along with the architecture and artifacts.401 This is not the case with the Earthquake House, where the remains of seven individuals were recovered during its excavation. Unlike Pompeii, where human remains found within domestic structures are not necessarily those of its inhabitants, at Kourion, the majority of the victims appear to have perished instantly with little or no warning. Because of this, the Earthquake House permits a more definitive identification of probable residential groups, although even these need to be considered judiciously.

Type A Storage: Excluding Room 6 and Room 7, spaces that were separate from the remainder of the structure during its final phase 402 The only exception is a single object that could possibly be a coin but cannot be positively identified as such. 403 Supra Chapter 4, p. 40. During the final period of occupation, Room 8 and Room 11 had to be accessed from Room 18 via the doorway in Wall E. The doorway linking Room 1 and Room 8 was likely blocked by this time. The degree to which Room 11 was accessible from the exterior of the structure is unknown. 404 The relationship between Room 6 and Room 7 and the residential groups identified in the Earthquake House cannot be determined, since they were architecturally separate from the remainder of the structure at the time of destruction. Although the possibility exists that these spaces were utilized by the inhabitants it is equally possible that they had been appropriated by a neighboring structure or represent an autonomous unit that had no interaction with the Earthquake House.

400 See Jameson 1990; Berry 1997; Allison 1997b, 2004; Cahill 2002; Ault 2005; Nevett 2010. 401 See George 1997a for discussion.

97

Figure 7.7: Plan of the Earthquake House showing victim locations. Plan adapted from Buell et al. 2009, fig. 1. Reconstructed by the author from AZLS: TRIII/Room 1, AZLS: Room 2, AZLS: Room 11, and AZLS: Room 20. and whose use cannot be definitively linked to its inhabitants, evidence for Type A storage was present in three of eleven rooms: Room 1, Room 8, and Room 14. Collectively, the Class 1 vessels provide a total Type A storage capacity of 336-357 lt., of which ca. 78% (267281 lt.) was located in Room 14. This space was the primary Type A storage location for the structure, which is logical since it functioned as one of the main loci for food preparation. Given this relationship, the majority of staple foodstuffs (e.g. cereals, pulses, lentils, olive oil, etc.) were likely stored in Room 14, with the remaining Type A storage in Room 1 (14%: 44-50 lt.) and Room 8 (7%: 25-26 lt.) used for other commodities (e.g. water, wine, nuts, dried fruit, etc.). If the three Class 1 vessels in Room 25 were utilized for Type A storage, the total capacity of the structure would have been increased by 25% to 421-445 lt.

other foodstuffs (e.g. dried fruits, smoked meat/fish, nuts, etc.), water, and domestic implements.405 Based on these percentages, the storage space allocated for primary foodstuffs in the Earthquake House would have been 268.8-285.6 lt. It is significant that this figure corresponds almost exactly, both from a volumetric and overall percentage standpoint, with the amount of Type A storage in Room 14 – the primary storage space in the structure. Given the lack of any functional water storage installation in the structure at this time,406 it is probable that the remaining Type A storage was given over for this purpose. It must be noted, however, that the vessels recovered only represent the durable component of the storage assemblage, with space provided by perishable containers (e.g. leather or cloth sacks, baskets, etc.) that have not survived unable to be accounted for. It is likely

Recent research on domestic storage practices that draws on ethnographic and historical records from pre-industrial Crete has shown that ca. 80% of available household storage space was employed for primary foodstuffs (cereals, pulses, olive oil, and wine), with the remaining 20% used for a variety of purposes including storage of

405 Christakis 2008, 32. These percentages are based on an analysis of records pertaining to pre-industrial (Venetian and Ottoman) households and ethnographic research on traditional Cretan practices. The 80% storage space for primary foodstuffs can be broken down as follows: Cereals – 55%; Pulses – 5%; Olive Oil – 8%; Wine – 12%. 406 The installation in the northeast corner of Room 15 may have been used for water collection, but whether it functioned for water storage as well is unknown.

98

that at least a portion of the Earthquake House Type A storage capacity was composed of these types of containers.

Type B Storage: Type B storage, both above and on the occupation surface was more extensive than Type A storage, with seven of the eleven rooms of the Earthquake House, as well as Room 6, preserving evidence for this activity. Artifacts in Type B storage encompass all classes of material (ceramic, glass, metal, stone). The objects are predominantly utilitarian in nature and likely would have been used frequently by the structure’s inhabitants. Because of this, a logical inference is that rooms with elevated concentrations of artifacts in Type B storage reflect spaces that served as focal points for day to day activities in the structure. The majority of rooms in the Earthquake House contained minimal evidence for Type B activity, normally on the order of two to three objects. However, Room 8 and Room 14 both produced significant assemblages associated with this type of storage (10+ objects), which, if the hypothesis is correct, would indicate that they played an integral role in the daily life of the Earthquake House. If the presence of large Type B storage assemblages is, in fact, an indicator of the central role of these spaces for the inhabitants, their location on opposite sides of the structure may support the hypothesis of its architectural division into two distinct residential spaces along Wall E.

Based on the number of victims (6 adults, 1 juvenile), estimates of annual dietary requirements, and consumption rates per individual,407 the inhabitants of the Earthquake House would have needed 1,636.46 liters of the four staple commodities previously mentioned per year.408 Given these figures, the Type A storage present in the Earthquake House would only hold enough basic provisions to last for 60-63 days. This could be extended to 79-83 days if the Class 1 vessels in Room 25 were pressed into service. Additional dietary supplement in the form of small scale domestic food production (such as a household garden), gathering of wild produce, and hunting/fishing were likely also used as a means to maximize and extend the staple commodities in storage. The results of this analysis indicate that residents of the Earthquake House would have needed to procure basic foodstuffs through purchase or trade frequently, probably within the city or its immediate environs. It is likely that the significant amount of small coinage found in the structure was used for such purchases, but the source(s) from which this income was derived cannot be determined from the archaeological remains. Together, these lines of evidence indicate the presence of some type of market economy at Kourion during this period, in which at least a portion of the Earthquake House inhabitants would have participated. However, the scale and frequency with which they did so can only be hypothesized at present.

Type C Storage: This is the most ubiquitous activity documented in the Earthquake House, with nearly every space, including Room 6 and Room 7 containing material that was either definitely or probably in Type C storage. In the Earthquake House proper, Room 3 and Room 25 appear to have been the primary focal points of this activity, along with Room 7 in the two room suite to the southwest. It is noteworthy that all three of these spaces are “controlled,” meaning that in order for them to be accessed, another room must first be entered and crossed.409 While in the case of Room 3 and Room 25 this control is clear, it is less so in Room 7 since the entrance location for these two spaces is unknown. Given the presence of Type A and Type B storage in Room 6, this space likely served as the center for any regular activity that occurred in these two spaces. Based on this, it is logical to postulate that the main point of access would have been in Room 6, which thus “controlled” Room 7.

Despite its separation from the remainder of the Earthquake House, Room 6 also served as a major locus for Type A storage. The Class 1 vessels from this space provide a total capacity of 327-341 lt., an amount roughly equal to that from Room 1, Room 8, and Room 14 combined. If the four Class 1 vessels in Room 7 were also utilized for Type A storage, this would have increased the total capacity of these two rooms by 30% to 425-443 lt. It is unfortunate that the relationship between these two rooms and the remainder of the structure during its terminal phase cannot be determined. By the time of the earthquake, Room 6 and Room 7 were separate from the Earthquake House, perhaps being utilized as a storeroom by a neighboring structure or as some type of small commercial establishment. It is hoped that further investigation to the south and west of the excavated area will provide answers to these questions.

In Room 8 and Room 14 where Type C storage has also been identified, the quantity of material is generally less than the rooms previously discussed. It also appears to have been relegated to peripheral areas rather than distributed throughout the space. This can be seen in Room 6 where objects in Type C storage were placed along the borders of the space and in Room 8 where this material is located opposite from what seems to be the main area of activity in the northwest quadrant. The best example, however, is provided by Room 14 where the narrow spaces between the oven and the walls were exploited for this behavior.

407 See especially Foxhall & Forbes 1982; Gallant 1991; Garnsey 1988; 1999; Christakis 2008. Figures for the annual consumption rates of foodstuffs in antiquity is highly varied and by no means agreed upon. For the purposes of this study, the amounts proposed by Christakis, whose research provides a comprehensive overview of the components of the ancient diet, are used. Christakis’ study represents the most recent attempt at quantifying ancient dietary needs and is based on data derived from an insular society (Crete), which are more readily applicable to Cyprus than figures derived from mainland populations. 408 See Appendix C, Table 3, Table 4, and Table 5 for calculations.

409

99

Grahame 2000, 33-4; Hillier and Hanson 1984, 109; Preziosi 1979.

Figure 7.8: Plan of the Earthquake House showing artifacts reflecting storages. Plan adapted from Buell et al. 2009, fig. 1. 7.3.3: Tracing the Domestic Waste Stream410 As mentioned in the previous and present chapters, Room 14, Room 12A, and Room 19 (Fig. 7.9) preserve evidence that permits a partial reconstruction of domestic discard processes, known collectively as the waste stream, in the northwest part of the Earthquake House during its final occupation. Analysis of the Class 2 material from these three spaces has provided artifactual evidence for a waste stream moving from the north and west parts of Room 14, to the southeastern part of that room, then to either the adjacent cistern (Room 12A) or nearby sottoscala (Room 19).411 One characteristic of domestic refuse at Greco-Roman sites is the diversity of material, as demonstrated by deposits from the Villa Regina and Gravina di Puglia.412 The material from the southeast side of Room 14 is consistent with this pattern and although it predominantly 410

The following discussion is the focus of a paper by the author published by the Danish Institute at Athens (Costello IV 2011). 411 The walls of these rooms are preserved to a significant height, which prevented any cross contamination of artifacts between them. 412 De Caro 1994, 96; Small et al. 1994, 201.

contains ceramics and animal bone, glass and other ecofacts do form a minor component. The Average EVREP for Room 14 is 65, with eleven vessels (17%) possessing completeness levels indicating they were Class 1 at the time of the earthquake. The remaining 54 vessels represent Class 2 material, of which 27 (50%) were recovered in the space east and south of the oven with the remaining 27 (50%) found throughout the rest of the room. Even though the number of vessels is split equally between these two areas, a comparison of the BVD reveals a significant difference. In the main part of the room, the BVD equals 5.5/m2, whereas in the area east and south of the oven the BVD is 23.27/m2. The density of Class 2 vessels in the discard area is 4.23 times greater than the remainder of the room, indicating a different set of behaviors was responsible for deposition of material in this location. The identification of this space as a discard area for waste material produced in the room is further strengthened by an examination of the animal bone. This class of material generally forms a significant component of refuse deposits and is perhaps one of the easiest ways to identify

100

potential areas where this activity occurred.413 In Room 14, a total of 347 animal bone fragments were recovered from the occupation surface and associated contexts during excavation. Of these, 159 fragments (45.8%) were from the area east and south of the oven. Like the ceramic evidence, the density of this material demonstrates a significant difference between the two areas of the room. In the main space, where 188 fragments (54.2%) were recovered, the density is 29.11/m2, whereas in the probable discard area it increases to 137.0/ m2. The animal bone density inside the discard area is 4.7 times greater than that of the surrounding space, which mirrors the types of behaviors observed in the ceramic material. Both the ceramic and animal bone from the east and south side of the oven in Room 14 represent the provisional discard of waste created during the use of the room, which eventually would have been removed to a subsequent location. The nearby Rooms 12A and 19 provide some evidence for the onward movement of material from Room 14 via the Earthquake House waste stream. In the southwest corner of the cistern (Room 12A), a deposit of fine ash mixed with bone and ceramic fragments (Locus 005) rested atop a thin layer of fine water deposited silt (Locus 006) that accumulated during the installation’s use life.414 Locating the origin(s) of this material was of central importance to the task of documenting the waste stream in this area of the Earthquake House. The ceramic assemblage from the cistern’s ash layer has a high percentage of cookware and lamps. Cookware and lamp fragments are similarly prominent in the material from the provisional discard area of Room 14. This point of similarity indicated the possibility that the cistern debris partly originated in Room 14. Supporting this possibility, a cookware fragment (Fig. 7.10: D) recovered from Room 12A at roughly the same elevation as the ash layer, but strictly speaking outside its boundary, joined cookware vessel V657 from the east side of the oven in Room 14.415 A second set of fragments from the cistern do not physically join, but almost certainly belong to cookware vessel P636 from the east side of the Room 14 oven. The two sets of fragments share comparable fabrics, a heavy accumulation of soot on their interior and exterior surfaces, the same estimated rim diameter, and an overall similarity in rim profile. The preserved wall elevations make any random, post-destruction movement 413 The cesspit investigated at Pompeii (Arthur 1993, 194-5) yielded a quantity of animal bone, although the total number of fragments has not been published. By far the best preserved feature of this type is the one from Gravina di Puglia, which yielded 1853 pieces of bone and shell (Small et al. 1994, 243). 414 Molinari et al. 1988, 173-4. Soren and colleagues hypothesized that at an earlier earthquake damaged the cistern to the extent that it could no longer function. It subsequently became a dumping area for household refuse. Although the oven in Room 14 was the most likely source of the ash, the deposit in the cistern has not been definitively linked to this feature. A similar situation of a cistern utilized as a receptacle for household discard has been documented in House 1 at Eleutherna, Crete (Stampolidis 2004, 58). 415 AZLS: Room 12, Locus 003 and Locus 005.

of fragments between these spaces virtually impossible and the stratigraphy showed no signs of prior disturbance. The cistern fragments joining to Room 14 vessels therefore attest to the deliberate transfer and deposition of Class 2 material from Room 14 into Room 12A. Despite the similar percentages of cookwares and bone and the cross-joins between rooms 14 and 12A, differences between the two assemblages indicate that some Room 12A material originated from waste produced elsewhere in the structure. However, these possibilities do not invalidate the demonstrable connection between Room 14 and Room 12A. The reuse of the non-functioning cistern is consistent with Peña’s observation based on ethnographic data that “some domestic refuse is disposed of in out of the way places around the residential compound, with features that constitute an enclosed depositional basin, such as abandoned rooms, outbuildings, fixtures, and pits, particularly favored for this purpose.”416 Room 19, like the cistern in Room 12A, was also being used for the deposition of household waste at the time of the earthquake. The excavators interpreted Room 19 as a storage area based on its marginal location, small size, and constrained space, all of which rendered it unsuitable for other household activities.417 As noted earlier, however, the threshold leading into the space from the courtyard was unusually well-worn, indicating it was frequently accessed. Since little material was recovered from the stratum contemporaneous with the house’s final activity phase,418 it was postulated that Room 19 was used to store perishable material, possibly fodder for the mule stabled in the adjacent Room 2. While this initial reading makes sense, the architectural and ceramic evidence indicate that at the time of the earthquake the space was being used for discard within the house. The lowest excavated level of Room 19 is roughly 40 cm. below the entrance threshold. This elevation difference would have made entering and exiting the space difficult, particularly if carrying goods to or from storage. As discussed in Chapter 4, many rooms in the Earthquake House had occupational surfaces well below their entrance threshold. In these rooms, some form of step was provided in order to facilitate access.419 However, no such feature was found in Room 19. Unlike a general storage area, which would have likely required entry on a regular basis, access to a room used for discard was only necessary when the space required cleaning out. The wear noted on the threshold of the room, rather than having resulted from repeated entry and exit, may be 416

Peña 2007, 276. Molinari et al. 1988, 175-6. 418 At the time of the excavation, three stratigraphic layers were identified in Room 19, all of which were interpreted as being earthquake debris. Research conducted by the author has demonstrated that the lowest layer consisted at least partially of fill deposited during the final phase of the structure. See Chapter 5, p. 50-1. 419 In Room 14, the floor is ca. 73 cm below the entrance threshold. In Room 8 the difference between the floor and threshold is approximately 43 cm.

101

417

or overflowing.420 One possible explanation is that Room 19 may have been cleaned out shortly before the earthquake. Alternatively, it may have been used primarily for disposal of organic material, possibly animal waste originating from Room 2, with more durable household waste deposited only occasionally. Whatever the situation, the positive, preserved evidence makes it clear that Room 19, at least in part, served as a component of the waste stream fed by ceramic material from Room 14. The evidence presented above reveals that Room 12A, Room 14, and Room 19 of the late 4th century AD phase of the Earthquake House were linked parts of the domestic waste stream. Pottery and animal bone debris indicates that the area east and south of the Room 14 oven was a provisional discard area for household waste at an early stage in this process. Room 12A and Room 19 provided loci for the discard of household waste, not only from Room 14 (as proven by preserved cross-joins) but also from other areas of the structure. The data from the assemblage clearly demonstrate that the domestic waste stream can be tracked successfully utilizing architectural and artifactual evidence. Ethnographic data can be used to help identify architectural features and spaces within a structure that may have been used for such discard activities. In complement to this, analysis of the general characteristics of the material (e.g. brokenness and completeness of ceramics, diversity of artifact classes, etc.) from potential discard areas, as well as a higher overall artifact density compared to the surrounding area can provide confirmation for deposits of this type. The ceramic evidence can then serve as a guide for cross referencing/mending with material from other spaces,

Figure 7.9: Detail plan of Room 12, Room 14, and Room 19. Plan adapted from Buell et al. 2009, fig. 1. Reconstructed by the author from AZLS: Room 12 and AZLS: Room 14. explained by people frequently having stood in the doorway to dispose of refuse. This interpretation is supported by the ceramic debris from Room 19. Material from its lowest stratigraphic layer is contemporary with the latest use of the structure. The assemblage here is also similar to that found in the cistern in Room 12A. As in the cistern, a greater percentage of the identified vessels are of a utilitarian nature than the discard assemblage of Room 14, indicating that it was likely derived from throughout the house. In quantifying the Room 19 ceramic, an average EVREP of 34 was obtained, of which four (11.76%) preserved cross-joins with vessels from Room 14 (Fig. 7.11) including two fragments of V657. Like Room 12, the preserved walls separating Room 19 from Room 14 would have prevented any random movement of these fragments between spaces. A striking feature of Room 19 in terms of its interpretation as a location for in-house discard is that the quantity of material is quite low compared to other rooms of the Earthquake House or refuse-collection loci documented elsewhere, which are normally found filled

102

Figure 7.10: Cookpot V657 rim fragment (A) recovered from provisional discard area of Room 14. Additional joining fragments recovered from Room 19 (B and C) and Room 12A (D). 420

E.g., Small et al. 1994; Ault 1999, 567-8.

which may provide further links in the domestic waste stream chain. 7.3.4: Domestic Production One activity that appears to be conspicuously absent from the Earthquake House is domestic production, specifically textiles, with no loomweights or other associated paraphernalia recovered during excavation. Textile production is known from both literary and archaeological evidence to have formed a major component of the ancient domestic economy421 and its apparent absence is significant, particularly since at least three of the victims were female. The lack of objects representing this activity, however, does not necessarily mean that small scale textile production did not occur in the Earthquake House, only that it did not involve the use of a weighted vertical loom. If textiles were produced in the structure, it may have employed a horizontal loom, which is believed to have developed in western Syria ca. AD 250.422 The only production definitively identified in the Earthquake House was small scale processing of agricultural commodities. This is evidenced by the groundstone assemblages recovered in Room 8, Room 14, and Room 25. Products resulting from this activity were almost certainly consumed by the Earthquake House inhabitants and not utilized as trade or retail commodities within the greater economic system of Kourion. This conclusion is based on the small number of objects associated with this type of activity found throughout the structure, as well as the total Type A storage capacity of the Class 1 vessels recovered. Both are inconsistent with large scale processing of agricultural foodstuffs for economic benefit. 7.3.5: Food Preparation (Fig. 7.12) Evidence for food preparation in the Earthquake House assemblage is sparse, being predominantly recovered in Room 8 and Room 14. The grinding implements in Room 8 indicate domestic-scale preparation of foodstuffs, with the basin in the northwest corner (V560) being suitable for a variety of tasks associated with this activity (e.g. mixing, transport, etc.). The Room 14 evidence is similar to that from in Room 8 except for the “grinding station” in its northeast corner. As discussed previously, the Room 14 groundstone is smaller than that from Room 8, being suitable for lighter tasks such as preparing herbs and spices for use. Room 14 produced the only Class 1 cooking vessel of the entire structure, with the possible exception of UR-1934 found during the Trench III, Room 1 excavation.423 The mortar and pestle from Room 25 421

Nevett 1999, Cahill 2002, 169-79, Ault 2002, 78-9 Wild 1987, 471. Although Wild focuses on the use of this loom to produce intricately patterned silk damasks, it would have been suitable for use with other fibers as well, including wool and linen. Both are known to have been used in Cyprus. A collection of bone bobbins recovered from the adjacent Building B were interpreted by the excavators as being used to produce bobbin lace. 423 Neither the vessel nor any field photographs depicting it could be located for examination to determine its status at the time of the earthquake. 422

Figure 7.11: Joining fragments of a Vessberg 18 lamp of the “galloping horse” variety. Unbroken discus (P635) and lower portion of nozzle recovered from the provisional discard area of Room 14; remainder from Room 19. could have been used for small scale preparation of foodstuffs in this space as well. Following preparation, the cooking process would have utilized one or more of the mudbrick ovens in Room 11, Room 14 and Room 18. No evidence of portable braziers or small, low-intensity fires that could be associated with this activity were discovered in any room of the structure. The presence of the largest oven in Room 11 may provide further evidence that Room 8 was an autonomous living area, precluding the need for its inhabitant(s) to use those located in other areas. Unlike Room 11, the Room 14 and Room 18 ovens do not indicate two distinct living spaces. Instead, the different locations likely reflect the seasonal nature of cooking, which would have occurred in interior spaces during the winter, and open spaces during the warmer spring and summer months.424 If this is the case, Room 14 probably represents the fall/winter kitchen with Room 18 being utilized during the spring/summer months. 7.3.6: Food Consumption (Fig. 7.12) Evidence for food consumption in the Earthquake House is slight. The majority of the objects traditionally associated with these activities (tablewares, metal pitchers, glass vessels, etc.) appear to have been recovered in their storage rather than use location. In addition, it is likely that a significant number of objects associated with this activity were in perishable materials

103

424

Nevett 2010, 106-8; Cahill 2002, 160-1.

Figure 7.12: Plan of the Earthquake House showing artifacts reflecting food preparation and food/drink consumption. Plan adapted from Buell et al. 2009, fig. 1. that do not survive. Based on the preserved evidence, of all the Earthquake House spaces, Room 8 seems to provide the best evidence for food consumption in the form of a small fineware cup (V550), fineware bowl (579/601), copper alloy pitchers (M164; M165), and the egg-shaped bottle (V561). The findspots of the first two vessels strongly suggest they were located on or around the stone table in the center of the room when the structure was destroyed. Together these objects appear to form a small service set that probably would have been used by the inhabitant(s) frequently, if not daily. 7.3.7: Ablution and Personal Hygiene No facilities that could have been used by the inhabitants for ablution or personal hygiene have been identified. It is possible that certain utilitarian objects, such as the ceramic basins from Room 8 and Room 14 could have been used for these activities in addition to any other role they would have fulfilled (e.g. food preparation, transport, etc.). The significant amount of Class 2 material in Type C storage found throughout the structure, particularly those of larger vessels such as

amphorae, could have been used for these types of activities as ad-hoc basins and/or chamber pots that would have been discarded following use. 7.3.8: Religious Activities No evidence of domestic cult activity could be discerned from the Earthquake House assemblage. The only overtly religious object was the Chi-Rho ring from Room 20 worn by the male victim at the time of death. The copper alloy and iron tintinnabulum from the northeast part of the same room likely performed a quasi-religious function. Such objects were regularly used as apotropaic devices against the “Evil Eye,” indicating the persistence of traditional (i.e. pagan) beliefs and methods of protection after Christianity became dominant.425 The recovery of this object is especially significant given the presence of a small infant among the victims in this space, as tintinnabula were often hung above cribs to provide additional protection for newborns.426 425 Use of objects to ward off the “evil eye” is still practiced in many Mediterranean countries. 426 Russell 1982, 137.

104

Chapter 8: Conclusions The data and interpretations presented in this work provide the first comprehensive account of the architecture, stratigraphy, and artifact assemblage recovered from the Earthquake House at Kourion, Cyprus. It began by presenting the theoretical background for material culture studies and reviewing previous scholarship that focused on domestic assemblage analysis from sites in the Greco-Roman world. Next, the development of Kourion as a city and an archaeological site was outlined, which provided historical context for the development of the Earthquake House as well as its subsequent excavation: first by the University Museum of the University of Pennsylvania Mission and later by the University of Arizona. Each of these projects employed distinct methodologies that had differing impacts on the assemblage as it was recovered during excavation. These impacts, the procedures used to mitigate their possible effects, and the techniques used to collect the data by the present study were delineated and reviewed prior to examining the archaeological evidence. Turning to the physical remains, the discussion first examined the architecture to investigate how the Earthquake House developed and changed throughout its history. Two additions that expanded the overall area of the structure and several internal modifications that significantly altered its spatial dynamics were identified through this process. An overview and analysis of the archaeological stratigraphy recorded during the excavation followed. Understanding this body of data is particularly critical for this study since it provides the artifacts’ context as well as insights into the formation processes that affected the structure pre- and postearthquake. Subsequently, a detailed autopsy of the in use assemblage from each room of the Earthquake House was undertaken to determine the use status of each object and identify possible behaviors/activities that occurred in each space. Finally, the results obtained from these three examinations (architectural, stratigraphic, and artifactual) were synthesized to produce conclusions regarding the structure’s overall spatial organization and the domestic activities that occurred within it during the final occupational phase. 8.1: Architecture Analysis of the architecture revealed that the Earthquake House was a modest structure of a type that was likely common in urban and rural settings throughout Cyprus during the Roman and Byzantine periods. Its construction techniques and overall plan were influenced by local building traditions and functionality rather than adherence to any specific architectural canon. Although this category of domestic architecture has recently begun to be examined in the Cypriot countryside,427 structures of this type have not been the focus of any systematic investigation. The situation is similar in urban settings, where such edifices have only been cursorily studied in 427

Rautman 2004.

favor of monumental residences with elite architectural elements. The Earthquake House, therefore, provides a rare example of a non-elite urban domestic structure. During its life cycle, the Earthquake House underwent a number of modifications. Unfortunately, the order in which these occurred and their significance are difficult to establish. They may have been carried out in response to pressure from an increase in household size, to create separate space for distinct residential/family groups, or to changes in the socioeconomic status of the occupants. Despite these alterations, the lack of any of aristocratic architectural elements (peristyle courtyard, mosaics, wall paintings, apses, etc.) indicates that the inhabitants did not aspire to elite status; or if they did, it was not expressed through this medium. In its final phase, the Earthquake House was subdivided into multiple living spaces. This, however, should not be construed as evidence for a “decline” of the structure or a low socioeconomic status for its inhabitants. It only indicates that architectural adaptation was necessary in order to serve the needs of the occupying individuals. Problems contextualizing the Earthquake House are significant, since it presently stands as an isolated structure virtually devoid of its surrounding urban landscape. While a few sites such as Kalavasos-Kopetra and Agios Kononas in Cyprus, Pella in Jordan, and Meiron in Israel provide useful comparative evidence for the Earthquake House, in general, published excavations of such non-elite structures are lacking. This study, therefore, makes a valuable contribution to our understanding of late Roman housing at the non-elite level. 8.2: Artifact Assemblage Recent scholarship has demonstrated the value of assemblage analysis as a means to identify the types and range of activities that occurred within a particular space.428 These studies have conclusively demonstrated that the majority of domestic space was multi-functional, even in cases where the architecture and its associated nomenclature have traditionally assumed functional specificity. The well preserved nature of the Earthquake House assemblage and its diligent recovery permitted placing the artifacts in one of four classes based on their condition and context: in use (life assemblage), refuse (death assemblage), recycled and/or reused, and postearthquake deposition. This system provided an effective tool for breaking down the material from each room into manageable and relevant data sets. However, not all artifacts were able to be assigned to a single category due to ambiguities surrounding their condition and context. While some of this vagueness may be due to anomalies and/or flaws in the excavation process (e.g. differential recovery, inadequate recording, etc.), in other cases it serves to emphasize the flexible nature of objects, whose 428 Nevett 2010, 2008; Ault 2005; Allison 2004; Cahill 2002; Berry 1997.

105

use can vary significantly even between identical items. Objects belonging to the life and death assemblages were examined to determine activities (e.g. storage, waste disposal, food preparation, etc.) with which they could possibly be associated. Based on the analysis of these sub-assemblages, the types and range of activities within a particular room and their spatial differentiation, if any, were determined. These findings were combined with the architectural and human remains evidence to provide a comprehensive picture of the organization of the Earthquake House. The results attest to and confirm the flexible nature of domestic space including evidence for multiple activities within a single room and the same activity occurring in multiple locations throughout the structure. The overall pattern of organization suggests that at the time it was destroyed, the Earthquake House was not a single residential unit. Based on the architectural remodeling, artifact distribution, and human remains, the existence of at least two, or possibly three, different residential groups that inhabited distinct parts of the structure can be inferred. The main part of the structure, including the upper floor, was occupied by the victims recovered in Room 1 and Room 2. Objects from these spaces, particularly the gold ring, silver-sheathed bone pin, and coinage indicates that these individuals possessed a certain amount of accumulated wealth. Rooms 20 and 25 appear to have been inhabited by a separate group composed of three victims known as the “Christian family.” The lack of personal possessions, jewelry, coins, and “luxury” items (bronze lamps, marble tables, etc.) associated with these remains indicates they held a lower socioeconomic status than those from Room 1 and Room 2. It is possible that the Room 20 individuals rented this space from those in Room 1 and Room 2, or used it in exchange for providing services. The single set of human remains from Room 11 may have used Room 8 as a multi-functional residential unit since it was largely isolated architecturally from the rest of the structure. The relationship, if any, between these individuals cannot presently be established, with possibilities ranging from a multi-generational extended family to three completely unrelated groups. The identification of distinct residential units was corroborated by the artifact and activity analysis. This revealed that two rooms served as focal points of daily activity: Room 8 and Room 14. Both produced evidence for multiple storage activities, food processing/preparation, and possible consumption of food/drink. The most telling evidence for this is the concentration of durable objects identified as being in non-perishable storage (Type B), which were likely regularly used. While Room 20 does not share the characteristics of the other two principal spaces, this does not preclude the identification of these individuals and this space as forming a separate residential unit. In this scenario, Room 14 likely served as a communal activity space for the Room 1/Room 2 and Room 20 groups.

A significant result from this study is the archaeological documentation of multiple storage types throughout the structure. Studies on domestic storage in the classical world are frustratingly limited, especially in the case of non-consumables and materials intended for reuse, recycling, and/or discard by the occupants. Storage was the most widespread activity identified in the Earthquake House and was differentiated into three types: storage of perishable/consumable commodities (Type A), storage of non-perishable objects (Type B), and storage of material for reuse, recycling, and/or discard (Type C).The available Type A storage space in the Earthquake House would have only sustained its inhabitants for ca. 2-2.5 months. The degree to which these individuals were engaged in agriculture is not known, but the quantity of Type A storage indicates that replenishment of at least some staple foodstuffs either through purchase or trade would have been necessary on a relatively frequent basis. Perhaps the most surprising and unexpected discovery from the Earthquake House storage analysis is the extent to which intact and broken ceramic vessels were retained and cached (Type C storage) awaiting reuse, recycling or discard. Although the main spaces used for Type C storage were those with “controlled” access, nearly every room preserved some object that could be associated with this activity. In spaces where this type of storage was documented, especially Room 14, the material was located toward the margins of the space to provide the greatest possible area for other activities. The identification of a Type C storage deposit of clutter refuse in Room 14 that joined to ceramic fragments recovered from the neighboring defunct cistern (Room 12A) and sottoscala (Room 19) permitted a partial reconstruction of the domestic waste stream in this part of the Earthquake House. The documentation of this process provides archaeological confirmation of waste disposal behaviors that until now have largely been based on models derived from ethnographic observations. The degree to which the Type C storage behavior documented throughout the Earthquake House represents a general cultural/societal trait or simply the practices of a single group of individuals is unknown. Undoubtedly, such activities vary significantly between households and it would be useful to compare the Earthquake House evidence to other domestic structures of equal and/or higher economic status to investigate similarities and differences in this practice. The presence of material in Type C storage is likely, at least partially, the reason behind some of the “anomalous” results obtained by assemblage analyses conducted at other Greco-Roman sites.429 This study has demonstrated that even in a true Pompeiilike scenario excavated using modern methodology a certain amount of ambiguity about the status of a particular object will still exist. This is because its condition and context will present multiple interpretations 429

106

Allison 2004, 39. See also Nevett (2010, 101-3).

for consideration. In order to account for anomalies in household artifact distribution and mitigate their impact, archaeologists must keep in mind and consider a wide variety of variables during artifact evaluation including the multi-functionality and condition of each object, whether they were in primary or secondary use, the possibility they could be in (provisional) discard, shortterm behaviors, seasonality of activities, the overall organization of space, and the possible existence of separate and/or multiple residential groups.

production and consumption of food on Cyprus during the Late Roman period. While this study of the Earthquake House has illuminated aspects of lives of its occupants to a level never before explored, it remains only one piece of a much large puzzle that awaits assembly. Nevertheless, it represents a vital first step toward exploring and understanding the complexities of daily life at Kourion during the Late Antique period.

8.3: Directions for Future Research This study of the Earthquake House has produced a nuanced model for understanding the distribution of artifacts in ancient domestic contexts. To address the question of room function it is necessary to create typological groups based on an assessment of the object context (use, storage, discard), condition, and their use in a primary or secondary role. When applied to other similarly well preserved and excavated assemblages throughout the Mediterranean, this methodology will permit the results from this analysis to be compared with structures of different periods and socioeconomic levels, providing greater insight into all levels of domestic life in the ancient world. With regards to the architecture of the Earthquake House, continued excavation is essential to obtain stratigraphic evidence for the building sequence. It is also necessary to complete excavation of the adjacent Building B and perform a similar architectural and artifactual analysis. This will contribute significantly toward contextualizing the Earthquake House and producing a more complete picture of pre-earthquake, Late Roman Kourion. To illuminate the urban context of these two structures, part of the research currently being undertaken by the Kourion Urban Space Project is to conduct a geophysical analysis of the area surrounding the Earthquake House.430 It is hoped that this along with targeted excavation will permit some estimate of housing density in this part of the city in addition to illuminating whether non-elite architecture existed alongside more monumental domestic structures or was relegated to a particular “neighborhood” on the acropolis. Architectural evidence from the main basilica and House of Eustolios indicates that following the earthquake, Kourion underwent significant changes. Architecture alone, however, cannot illuminate the nature of the change involved. By comparing the types and range of artifacts and their distribution in the Earthquake House to post-earthquake structures at Kourion, it should be possible to investigate questions about the nature and degree of any social change that might have occurred at the site following its reoccupation. By comparing the range of artifacts and their distribution in the Earthquake House to other domestic structures from rural contexts, it will be possible to further examine questions concerning the domestic economy, patterns of storage, and 430

This part of the project will be under the direction of the author.

107

108

Appendix A: Catalogue of Individual Room Assemblages



All measurements in centimeters unless otherwise indicated.

109

Room 1 Assemblage: Ceramic: Object

Ware

Reg. #

Class

Description

Storage Vessel

Plain

P134-1934

1b

Recorded in Daniel notebook as a "large jug with base ring and one handle to shoulder. Deep red clay, not ribbed. Neck and handle gone." Unable to be located - not examined. Missing neck and handle appears to indicate pre-earthquake breakage, with the vessel remaining in use.

Pithos

Plain

P1

1b

RD: 27; BD: 11.7; MaxD: 40.2; H: 52.7. Fig. 6.3. Largely complete and reconstructible two handled pithos. Findspot in the southwest corner.

Pithos

Plain

P22/28 /153

1b; 2

RD: 26.4; H: 17.8 (Max.). Reconstructible rim of a moderately sized pithos. Found upside down in the southwest corner of the room adjacent to P1. Unknown if the remainder of the vessel or additional fragments were recovered during the 1934 excavation.

Amphora Kourion Type IV/ Late Roman 4

Amph

P221

1b; 2

RD: 10.3; MaxD: 25.6. Fig. 6.4. Partially reconstructible KTIV/LR4 "Gaza" amphora. Bottom intact, rim reconstructible. Additional fragments identified, but not enough to reconstruct the vessel. Unknown if additional fragments recovered during the 1934 excavation. Bottom has a small hole (ca. 0.5 cm dia.) drilled into the bottom, likely to facilitate emptying.

Amphora Unknown

Amph

P135-1934

1b; 2

Recorded in Daniel notebook as the "bottom of a large pointed amphora. Yellow slip on brown clay. Ribbed." Form indeterminate. Unable to be located for examination.

Lamp Unknown

Lamp

L47-1934

1b;1c

Recorded as an "almost complete" lamp with traces of use. Herringbone pattern on edge of bowl. Crude floral decoration on shoulder. Description consistent with Vessberg 18.

Lamp Unknown

Lamp

L51-1934

2

Lamp Unknown

Lamp

L112-1934

2(?)

Recorded in Daniel notebook. No indication of completeness level. Decoration description consistent with Vessberg 18.

Lamp Unknown

Lamp

Unk.

2(?)

Recorded in Daniel notebook. No indication of type or completeness level. Status unable to be determined.

Lamp Unknown

Lamp

Unk.

2(?)

Recorded in Daniel notebook. Status unable to be determined.

Lamp Unknown

Lamp

P186

2

Three lamp fragments.

Cookpot with Lid

Cook

Unk.

1b

Recorded in Daniel notebook. Described as "red ribbed."

Listed as lamp fragment in Daniel notebook.

No indication of completeness level.

Glass: Glass recovered during the 1935 excavation by Daniel unable to be located. List of finds from Room 1 in Daniel notebook records a glass vessel, and a circle on the sketch plan of the room records an area in the northern part of the room labeled "brownish glass." Diameter for the vessel is recorded as 16.5 cm. The fact that the vessel diameter could be determined appears to indicate that it was relatively complete. Material recovered during the 1984 excavations is missing and unable to be located.

Metal: Object Gold Finger Ring

Metal

Reg. #

Class

Description

AU

S411934

1a

Gold ring recovered by Daniel from one of the skeletons in Room 1. Recorded as a "gold ring set with glass or a stone." Unable to be located for examination.

110

Room 1 Assemblage (Cont.): Metal: (Cont.) Object

Metal

Reg. #

Class

Description

Finger Ring

CA

Unk.

1a

Copper alloy ring from one of the skeletons recovered in Room 1. Recorded as a "bronze ring." Recovered on same finger as gold ring S41.

"Key" Ring

CA

M5

1b;1c; 1d

Copper alloy ring. Large bezel that is grooved or turreted on one side. Supposedly modified to be used as a key. Recovered in the earthquake destruction layer (Locus 012). Precise elevation not recorded in notes.

Buckle(?)/Clasp(?)

CA

M26

1b;1c; 1d

L: 2.1; W: 3.0 - 1.1. Fig. 6.2. Copper alloy object. May be part of a clasp or a buckle.

Pitcher

CA

Unk.

1b;1c

Copper alloy pitcher recovered by Daniel. Similar to other copper alloy pitchers found throughout the structure. Unable to be located for examination.

Coin

CA

CB- 1934

1c;1d;4

Copper alloy coin recovered by Daniel during the Room 1 excavation. Description/identification not found in excavation notes.

Coin

CA

CC-1934

1c;1d;4

Copper alloy coin recovered by Daniel during the Room 1excavation. Description/identification not found in excavation notes.

Coin

CA

CD-1934

1c;1d;4

Copper alloy coin recovered by Daniel during the Room 1excavation. Description/identification not found in excavation notes.

Coin

CA

CF-1934

1c;1d;4

Copper alloy coin recovered by Daniel during the Room 1excavation. Description/identification not found in excavation notes.

Coin

CA

C37231934

1c;1d;4

Recovered by Daniel 5 cm. above one of the skeletons in the earthquake debris layer (Locus 012). Identified in Room 1 notes as an early coin of the emperor Valens. Dated to late AD 365 or pre-September AD 365. SECURITAS REI PUBLICAE. Nike with wreath.

Coin

CA

C37241934

1c;1d;4

Copper alloy coin recovered by Daniel during the Room 1excavation. Description/identification not found in excavation notes.

Coin

CA

C37251934

1c;1d;4

Copper alloy coin recovered by Daniel. Identified in Room 1 notes as Constantius II. SPES REPUBLICAE. Emperor with spear and right hand on globe. AD 355-361.

Coin

CA

C5

1c;1d;4

Copper alloy Aes 1 (Locus 012). Obverse may be Jovian 363-4. Victoria Romanorum. Emperor with laberum [sic] and holding victory. Exergue: ES. Recovered in sift.

Coin

CA

C6

1c;1d;4

Copper alloy coin (Locus 012). In situ

Coin

CA

C7

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C9

1c;1d;4

Copper alloy Sestertius of Claudius or Nero (AD 41-68) (Locus 012).

Coin

CA

C10

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C11

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C12

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C13

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C14

1c;1d;4

Copper alloy coin (Locus 015).

Coin

CA

C15

1c;1d;4

Copper alloy coin (Locus 012). Constantius II (?) Mid 4th c. AD.

Coin

CA

C16

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C17

1c;1d;4

Copper alloy coin (Locus 012). Shows profile. Recovered in sift.

Coin

CA

C18

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C19

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C20

1c;1d;4

Copper alloy coin (Locus 015). Recovered in sift.

Coin

CA

C21

1c;1d;4

Copper alloy coin (Locus 015). Recovered in sift.

Coin

CA

C22

1c;1d;4

Copper alloy coin (Locus 015). Recovered in sift.

Coin

CA

C23

1c;1d;4

Copper alloy coin (Locus 015). Recovered in sift.

111

Room 1 Assemblage (Cont.): Metal(Cont.): Object

Metal

Reg. #

Class

Coin

CA

C24

1c;1d;4

Copper alloy coin (Locus 015). Recovered in sift.

Coin

CA

C25

1c;1d;4

Copper alloy coin (Locus 015). Recovered in sift.

Coin

CA

C26

1c;1d;4

Copper alloy coin (Locus 015). In situ. Damaged with pick.

Coin

CA

C27

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C28

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C29

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C30

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C31

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C32

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C33

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C34

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C35

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C36

1c;1d;4

Copper alloy coin (Locus 015). In situ. Obverse: SPF AUG. Reverse: Seems to be SECURITAS REIMBLICAE [sic]. Victory going left. Des 3 type coin.

Coin

CA

C37

1c;1d;4

Copper alloy coin (Locus 015). In situ. Mid 4th c. AD to AD 470. Reverse: SECURITAS REPUBLICAE.

Coin

CA

C38

1c;1d;4

Copper alloy coin (Locus 017).

Coin

CA

C39

1c;1d;4

Half of copper alloy coin (Locus 017). Modern break.

Coin

CA

C41

1c;1d;4

Copper alloy coin (Locus 017). In situ.

Coin

CA

C42

1c;1d;4

Copper alloy coin (Locus 017). Recovered in sift.

Coin

CA

C45

1c;1d;4

Copper alloy coin (Locus 017). Recovered in sift.

Coin

CA

C46

1c;1d;4

Copper alloy coin (Locus 017). Recovered in sift.

Coin

CA

C47

1c;1d;4

Copper alloy coin (Locus 017). In situ. Under stone.

Coin

CA

C48

1c;1d;4

Copper alloy coin (Locus 017). Recovered in sift.

Coin

CA

C50

1c;1d;4

Copper alloy coin (Locus 012). In sweep.

Coin

CA

C51

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C52

1c;1d;4

Copper alloy coin (Locus 012). Figure of emperor dragging captive & holding labarium [sic].

Coin

CA

C53

1c;1d;4

Copper alloy coin (Locus 012) (Frag.).

Coin

CA

C217

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C218

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C219

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C220

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C221

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C222

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Stone: Object

Description

Stone

Reg. #

Class

Description

Column Drum / Hearth(?)

LS

S5

1b; 3

D: 18.0 - 17.5; H: 11.7. Fig. 6.5. Limestone cylinder, probably a small column drum or suspensura. Possible evidence of burning on one surface. Possibly recycled and used as construction material in the wall.

Groundstone

Gab

S6

1c; 3(?); 4(?)

L: 11.5 - 10.5; W: 10.6 - 9.0; H: 5.0. Possibly a fragment of a quern used as a hand grinder. May be recycled material used for wall construction.

112

Room 1 Assemblage (Cont.): Bone: Object

Reg #

Class

Description

Pin

S44-1934

1a

Fragment recovered during the Trench III excavation. Shaft with teardrop shaped head. Recorded in the field notebook as found by B.H.H. "beneath the skull of L." Recorded length ca. 4 cm. Found associated with a hexagonal green paste bead.

Miscellaneous: Object

Reg #

Class

Description

Paste Bead

S43-1934

1a

Hexagonal green paste bead recovered during the Trench III excavation. Recorded as being found by B.H.H. "beneath the skull of L." Found in proximity to a bone hair pin, and it may have been used in conjunction with this object. Unable to be located.

Room 2 Assemblage: Pottery: All the ceramic from Room 2 can be identified as either Class 2 or Class 4.

Glass: Object Jar

Color

Reg. #

Class

Description

Light Blue

G104

1c(?); 1d

MaxD: ca. 24.0. Fig. Joining fragments of a glass vessel, probably a jar. Partially reconstructible. Completeness level of unable to be determined. Recovered from the post-earthquake collapse layer (Locus 005).

*Majority of the glass from this space during the 1984 excavations is missing and was unable to be located. The remainder of the glass from Room 2 can be identified as either Class 2 or Class 4.

Metal: Object

Metal

Reg. #

Class

Description

Lampstand

CA

M69

1d

Cast copper alloy lampstand. Tripod base with curved legs. Spiral shaft terminating in a stylized Corinthian capital. Tripod element above in the form of three dolphin heads supporting a small separately attached "hat" type lamp.

Ring

CA

M97

1a

Copper alloy ring. Part of the tack assemblage for the mule. Recovered on the left side of the mule skull.

Curb Bit Manning Type 1

FE

M117

1c

Fig. 6.10. Complete iron curb bit. Recovered along Wall D. Composed of two iron rods, hinged together in the middle. Heavily corroded.

Ring

FE

M118

1c

L: 9.4; W: 6.7. "Figure 8" shaped iron ring associated with the mule tack assemblage. Recovered on the right side of the mule skull.

Chain Fragments

FE

M122

1a

L: 10.3; W: 3.3 - 2.0; H: 1.6 - 1.3. Three iron chain fragments including one complete link. Fragments consolidated in the field.

Chain Fragments

FE

M120

1a

Iron chain fragments. Heavily corroded and friable.

113

Room 2 Assemblage (Cont): Metal (Cont.): Object

Metal

Reg. #

Class

Chain Fragments

FE

M121

1a

Apparent spall fragments from the main body of an iron chain link.

Description

Chain Fragments

FE

M126

1a

Two joining large iron chain fragments. Consists of two links that were overlapping at the time of deposition.

Chain Fragments

FE

M119

1a

Iron chain fragments, some of which join.

Chain Fragments

FE

M125

1a

Single chain link. Two joining fragments. Consolidated in situ.

Ring

FE

M112

1c

L: 6.8; W: 5.7. Fig. 6.16. Heavy "U" shaped iron ring. Identified as a "Tether Ring" on tag. Object appears only partially complete. Recovered along Wall D.

Chain Fragments

FE

M68

1a

Iron chain fragments from inside stone trough.

Chain Fragments

FE

M39

1a

Iron chain fragments from inside stone trough.

Chain Fragments

FE

M108

1a

Iron chain fragments from inside stone trough. Appear to be spall from larger fragments.

Chain Fragments

FE

M106

1a

Iron chain fragments from inside stone trough. Appear to be spall from larger fragments.

Chain Fragments

FE

M40

1a

Iron chain fragments recorded as being from on top of the edge of the stone trough.

Chain Fragments

FE

M124

1a

Iron chain fragments from inside stone trough. Appear to be spall from larger fragments.

Chain Fragment

FE

M123

1a

L: 7.5. Iron fragment recovered from one of the holes in the wall of the stone trough. Heavily corroded.

Iron and Lead (?) Object

FE / PB(?)

M116

1c(?)

L: 10.2; W: 5.3; H: 3.4. Fig. 6.11. Iron and probable lead object. Identified as a lock although not consistent with other locks recovered and does not fit any known lock typology. Iron elements appear to consist of two possibly interlocking plates that both terminate in a semicircular iron ring. Opposite side consists of a block (4.3 X 3.7 X 1.3) of a different metal, possibly lead. This block appears to be pierced along its long axis by two circular iron elements. The short axis appears to be pierced by five similar iron elements arranged in a rough "X" pattern. Association with other elements of mule tack along Wall D suggests that it may belong to this group of objects.

Iron Strip

FE

M111

1b(?); 1c(?)

Strip of iron. Possibly a strengthening band for a wooden object. Exact findspot unknown.

Ring

FE

M113

1c

L: 7.4; W: 5.7; H: 1.6. Fig. 6.18. Complete iron ring. Recovered along Wall D. Findspot suggests a possible function as part of the mule tack assemblage. Heaviness/thickness of the object appears inconsistent with previously examined tack elements.

Lead Object

PB

M99

2

L: 2.3; H: 1.9; W: 1.0. Small rectangular lead object, possibly a weight of some sort. Recovered from the interior of the stone trough.

Coin

CA

C58

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C81

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C85

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C96

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C133

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C134

1c;1d;4

Copper alloy coin (Locus 005).

Coin

CA

C136

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C145

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C146

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C147

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

114

Room 2 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Class

Coin

CA

C148

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Description

Coin

CA

C149

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C151

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C152

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C169

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C170

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C171

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C179

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C180

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C181

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C182

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C183

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C184

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C186

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C193

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C198

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C199

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C200

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C201

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C202

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C203

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C206

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C207

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C208

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C209

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C210

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C211

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C212

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C214

1c;1d;4

Copper alloy coin (Locus 005). In sweepings.

Coin

CA

C215

1c;1d;4

Copper alloy coin (Locus 005). In sweepings.

Coin

CA

C216

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C223

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C224

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C225

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C226

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C227

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C228

1c;1d;4

Copper alloy coin (Locus 005). In situ. Identified as Constans, ca. AD 330. Reverse: Two warriors flanking trophies.

Coin

CA

C229

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C230

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C231

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C232

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C233

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

115

Room 2 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Class

Coin

CA

C234

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C235

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C236

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C237

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C238

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C239

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C240

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C241

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C242

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C243

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C244

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C245

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C246

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C247

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C248

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C249

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C250

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C251

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C252

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C253

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C254

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C255

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C256

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C257

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C258

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C259

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C260

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C261

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C262

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C263

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C264

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C265

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C266

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C267

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C268

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C269

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C270

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C271

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C272

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C273

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C274

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C275

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C276

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Description

116

Room 2 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Class

Coin

CA

C278

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C279

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Description

Coin

CA

C280

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C281

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C282

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C283

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C284

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C285

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C286

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C287

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C288

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C289

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C290

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C291

1c;1d;4

Copper alloy coin (Locus 005). Recovered in sift.

Coin

CA

C292

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C293

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C294

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C295

1c;1d;4

Copper alloy coin (Locus 005). In situ.

Coin

CA

C296

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C297

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C298

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C299

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C300

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C301

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C302

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C303

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C304

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C305

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C306

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C307

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C308

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C309

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C310

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C311

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C312

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C313

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C314

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C315

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C316

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C317

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C318

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C319

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C320

1c;1d;4

Copper alloy coin (Locus 012).

117

Room 2 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Coin

CA

C321

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C322

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C323

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C324

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C325

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C326

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C327

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C328

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C329

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C330

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C331

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C332

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C333

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C334

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C335

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C336

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C339

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C340

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C341

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C342

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C343

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C344

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C345

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C346

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C350

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C351

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C353

1c;1d;4

Copper alloy coin (Locus 012). In situ.

Coin

CA

C354

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C357

1c;1d;4

Copper alloy coin (Locus 012).

Coin

CA

C337

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Coin

CA

C338

1c;1d;4

Copper alloy coin (Locus 012). Recovered in sift.

Stone

Reg. #

Class

Description

Marble

S20/ S21/ S33/ S34/ S41/ S42/ S44/ S47/ S51/ S52/ S59

1d

D: 69.4. Fig. 6.12. Round tabletop with an elaborate carved border. Fine-grained, brilliant white marble. Object reconstructible with the exception of ca. four fragments, which are likely in the unexcavated area under the stone trough. No evidence of attachment points for any support such as carved legs. Lack of stone pedestal/legs seems to indicate a wooden support system. Completeness of object indicates that it was likely in use at the time of the earthquake. Found in close association with the copper alloy lampstand M69.

Class

Description

Stone: Object

Tabletop

118

Room 2 Assemblage (Cont.): Bone: Object

Reg #

Class

Pin

BI2

2;4

L: 3.5. Fig. 6.7. Large bone pin fragment. Cylindrical shank with a flat discoid head.

Pin

BI3/BI4

1a

Joining fragments of a bone hair pin recovered in association with the human remains. Straight shaft with a teardrop shaped head. On display in Kourion Museum.

Pin

BI5

1d;2;4

L: 4.3. Fig. 6.8. Head and upper shank of a bone pin. Cylindrical shank demarcated from the head by incised grooves and a flared collar. Rounded ovoid head.

Inlay (?)

BI6/BI7

1d;2;4

Description

L: 5.9; W: 0.8. Fig. 6.9. Two joining fragments of a carved bone object. Possibly some type of inlay decoration. One side worked flat, with the other carved in a bead and reel pattern. Both ends appear broken. Likely part of some larger object in perishable material.

Room 3 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description

LRC Dish Form 1

Fine

181

2

RD: 17.0; BD: 9.1; H: 4.3. Partially reconstructible LRC Form 1 dish. Recovered in the northeast corner of the room.

Amphora Vasa Type C

Amph

380 (388)

2

RD: 17.9; BD: 12.3; MaxD: 33.96; H: 44.9. Partially reconstructible VTC amphora. Significant portion of the lower body is missing. Rim of the vessel also missing. Based on the evenness of the breakage it appears to have been removed intentionally as a maintenance process. Search of the sherd material may lead to the identification of additional fragments, but not enough for complete reconstruction.

Jug Vasa Type 1H

Plain

187

1b

Fig. 6.14. Reconstructible Vasa Type 1H jug. Minor fragments missing, but are likely present in the loose sherd material.

Amphora Vasa Type C

Plain

161/174/ 203

2

BD: 15.2; H: 27.2 (Max.)Partially reconstructible bottom of a VTC amphora. Fragments recovered near the entrance to the room.

Bowl

Plain

182

2

RD: 14.5; BD: 10; H: 5.2. Partially reconstructible plainware bowl with a flared rim and hemispherical bottom. Excavation notes record findspot in "east corner." Additional fragment recovered from the lower debris layer of Room 2 indicating that it was broken prior to the earthquake with the fragments being scattered between rooms.

Cup

Plain

235

1b

Largely reconstructible small single handled mug. Recovered near the northeast corner of the room. Minor fragments missing.

Lamp Vessberg 18

Lamp

362

1b;1c

Complete and unbroken Vessberg 18 lamp of the "Fisherman" variety. Recovered in the northeast corner of the room.

119

Room 3 Assemblage (Cont.): Pottery (Cont.): Object

Ware

Reg. #

Class

Description

2

RD: 39.0; BD: 17.0; H: 35.5 (Max.). Fig. 6.16. Profile similar to Kalavasos-Kopetra 215 and Kourion Survey 63n. Fragments of a moderately sized pithos. Joining and non-joining fragments recovered including a reconstructible base and a sizable rim fragment. Fragments recovered in the area of the door leading into the room on the floor. Larger rim fragment from the north half of Room 11 joins to the Room 3 rim fragment.

Pithos

Plain

174/ 196/ 197/ 208/ 213/ 214/ 217/ 390

Pithos

Plain

215/ 237

2

RD: 28.1; BD: 9.7; MaxD: 41.82; H: 55.6. Partially reconstructible small pithos. Large section of the vessel’s upper shoulder missing and do not appear to have been recovered during the excavation.

Amphora Kourion Type I

Amph

175/ 206/ 274A/ 274B/ 274C

2

Joining fragments that appear to belong to a Kourion Type I amphora with cross-joins found in the adjacent Room 2. Fragments recovered in primary destruction layers.

Amphora Kourion Type III

Amph

161/ 170/ 209

2

RD: 7.6; MaxD: 30.71; H: 76.8. Fig. 6.15. Largely reconstructible KTIII amphora. Top and bottom complete, but large body sections missing. Majority of the vessel recovered in the south part of the room adjacent to Wall E. Several fragments that account for the majority of the missing sections were recovered in a search of the sherd material from the room. Dispersal of fragments indicates that the vessel was broken prior to the earthquake.

Amphora Sphaethon

Amph

363

2

BD: 4.2; H: 29.6. Joining fragments that form the complete spike of a Sphaethon amphora.

Amphora Kourion Type III

Amph

197/ 198

2

MaxD: 32.6; H: 35.8 (Max.). Partially reconstructible amphora bottom, most probably of the Kourion Type III variety.

Lid Storage

Plain

195

2

RD: 24.0. Joining fragments of a large flat lid, likely for a pithos or other storage vessel. Handle attachment point preserved. Fragments possess evidence of sooting/thermal exposure. Likely indicates reuse in a cooking capacity.

*The remainder of the ceramic vessels identified from Room 3 can be identified as either Class 2 or Class 4.

Glass: Object Vessel Unknown

Color

Reg. #

Class

C'less

G133/ G134

2

Description Consolidated handle and body fragments of a glass vessel. handle, narrowing towards bottom.

Strap

*The remainder of the glass recovered in Room 3 can be identified as either Class 2 or Class 4.

Metal: Object

Metal

Reg. #

Class

Description

Cosmetic Spoon/ Kohl Stick

CA

M44

1d

Copper alloy object, likely a cosmetic spoon or khol stick.

Lead Object

PB

M48

2

L: 3.6; W: 2.1; H: 1.5. Fig. 6.17. Irregularly shaped lead object of unknown function. May be some type of lead plug or clamp for mending ceramic vessels.

Ring

CA

M55

2

L: 3.3; W: 2.1. Small, oval-shaped copper alloy ring. Not a finger ring. Lap and rivet joint.

120

Room 3 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Class

Coin

CA

C59

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C60

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C61

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C62

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C63

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C64

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C65

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C66

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C67

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C68

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C69

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C70

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C71

1c;1d;4

Copper alloy coin. Recovered in sift. GLORIA ROMANUM (?)

Coin

CA

C72

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C73

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C74

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C75

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C76

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C77

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C78

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C79

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C80

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C82

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C83

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C84

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C86

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C87

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C88

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C89

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C90

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C91

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C92

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C93

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C94

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C95

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C97

1c;1d;4

Copper alloy coin. Recovered in sift. Valens. Victory with wreath going left. SECURITAS.

Coin

CA

C98

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C99

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C100

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C101

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C102

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C103

1c;1d;4

Copper alloy coin. Recovered in sift.

Description

121

Room 3 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Class

Coin

CA

C104

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C105

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C106

1c;1d;4

Copper alloy coin. Recovered in sift. Valens: Sear 4018

Coin

CA

C107

1c;1d;4

Copper alloy coin. Recovered in sift. Constantius II. FEL TEMP.

Coin

CA

C108

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C109

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C110

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C111

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C112

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C113

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C114

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C115

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C116

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C117

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C118

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C119

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C120

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C121

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C122

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C123

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C124

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C125

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C126

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C127

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C128

1c;1d;4

Copper alloy coin. Recovered in sift. Valens: AD 365-68.

Coin

CA

C129

1c;1d;4

Copper alloy coin. Recovered in sift. Off Center Stamp. 4th c. AD.

Coin

CA

C130

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C131

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C132

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C137

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C138

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C139

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C140

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C141

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C142

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C143

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C150

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C153

1c;1d;4

Copper alloy coin. In situ. Found in base of vessel 215/237.

Coin

CA

C154

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C155

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C156

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C157

1c;1d;4

Copper alloy coin. Recovered in sift.

Description

122

Room 3 Assemblage (Cont.): Metal (Cont.): Object

Metal

Reg. #

Class

Coin

CA

C158

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C159

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C160

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C161

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C162

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C163

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C164

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C165

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C166

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C167

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C168

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C172

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C173

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C174

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C175

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C176

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C177

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C178

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C185

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C187

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C188

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C189

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C190

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C191

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C192

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C196

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C197

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C347

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C348

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C349

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C352

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C358

1c;1d;4

Copper alloy coin.

Coin

CA

C359

1c;1d;4

Copper alloy coin.

Coin

CA

C360

1c;1d;4

Copper alloy coin. In situ. In pot.

Coin

CA

C362

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C363

1c;1d;4

Copper alloy coin. In situ.

Coin

CA

C364

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C365

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C366

1c;1d;4

Copper alloy coin. Recovered in sift.

Coin

CA

C367

1c;1d;4

Copper alloy coin. Recovered in sift.

Description

Stone: No relevant stone objects were part of the room assemblage.

123

Room 3 Assemblage (Cont.): Bone: Object

Pin

Reg #

BI1

Class

Description

1d;2

L: 3.0; D (Head): 1.0. Fig. 6.18. Upper shank and head of a bone pin. Top of shank and head covered with thin metal identified as silver in the registry. If the metal covering the shank and head is silver, it is unlikely that the object, whether broken or not, would have been discarded without this material being removed. Seems to indicate that this object was part of the in use assemblage.

Room 6 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description

Amphora Kourion Type I

Amph

V478

1b

RD: 10.5; BD: 4.4; MaxD: 35.71; H: 67.1. Reconstructible vessel. Missing fragments likely in loose sherd material from room.

Amph

474-1 (523?)

2

RD: 12.8; BD: 5.1; MaxD: 37.36; H: 68.7. Fig. 6.23. Reconstructible vessel. Small missing fragments likely in loose sherd material from room. Bottom of the vessel has a circular hole cut in it, most likely to facilitate trans-vasing of original contents.

Amphora Kourion Type I

Amphora Kourion Type I

Amph

523 (474-1?)

2

RD: 12.2; BD: 4.7; MaxD: 35.30; H 67.5. Partially reconstructible vessel. Relatively large sections of the body missing. Vessel recovered adjacent to Locus 003, but some fragments recorded on plan in excavation notes as being from the passage from Room 6 and Room 7. Indicates that the vessel was broken prior to the earthquake.

Amphora Kourion Type I

Amph

V502

1b

RD: 11.1; BD: 4.1; MaxD: 35.8; H: 64.8. Reconstructible vessel. Small missing fragments likely in loose sherd material from room.

Amphora Kourion Type II

Amph

480/500

1b

RD: 11.8; BD: 4.9; MaxD: 33.96; H: 85.4. Reconstructible vessel. No significant missing fragments, except for single handle.

Amphora Kourion Type III

Amph

P522

1b

RD: 9.5; MaxD: 28.8; H: 71.0. Reconstructible vessel. No significant missing fragments.

Amphora Kourion Type III

Amph

V532

1b

RD: 10.2; MaxD: 28.2; H: 74.6. Reconstructible vessel. fragments likely in loose sherd material from room.

Amphora Kourion Type III

Amph

V542

1b

RD: 10.2; MaxD: 30.78; H: 74.8. Reconstructible vessel. Missing fragments likely in loose sherd material from room as well as small nonreconstructible pieces bagged with the vessel.

Amphora Kourion Type III

Amph

534/541

1b

RD: 11.2; MaxD: 31.83; H: 77.6. Reconstructible vessel. Missing fragments likely in loose sherd material from room as well as small nonreconstructible pieces bagged with the vessel.

Amphora Kourion Type III

Amph

474-2

1b

RD: 10.5; MaxD: 32.62; H: 82.9. Reconstructible vessel. No significant missing fragments.

Amphora Kourion Type III

Amph

521/ 533

1b

RD: 10.1; MaxD: 30.62; H: 74.8. Reconstructible vessel. No significant missing fragments.

1b;2(?)

RD: 10.1; MaxD: 32.69; H: 79.9. Reconstructible vessel. No significant missing fragments, except for single handle. Small missing areas likely in loose sherd material from room. Possible deliberate hole cut in vessel side towards the bottom, but cannot be determined due to postexcavation conservation.

2

RD: 11.1; MaxD: 33.96; H: 85.4. Partially reconstructible vessel. Relatively large sections of the body missing. Possible that some may be located in the sherd material from this room, but unlikely that all missing fragments would be located.

Amphora Kourion Type III

Amphora Kourion Type IV / Late Roman 4

Amph

Amph

V403

505

124

Missing

Room 6 Assemblage (Cont.): Pottery (Cont.): Object

Class

Reg. #

Type

Description

Amphora Late Roman 3 / Type BIV

Amph

508

2

RD: 3.5; MaxD: 18.93; H: 21.6 (Max). Partially reconstructible amphora top. Numerous body fragments, but not enough remain to fully reconstruct the vessel. No base present. High level of brokenness among the fragments likely indicates the vessel was broken prior to the earthquake.

Pithos

Plain

V443

1b

RD: 41.5; BD: 16.8; MaxD: 56.88; H: 67.7. Fig. 6.20. Profile similar to Kalavasos-Kopetra 215 and Kourion Survey 63n. Reconstructible vessel. No significant missing fragments.

Funnel

Plain

520

1b

Fig. 6.22. Reconstructible vessel. No significant missing fragments, except single small handle. Presence indicates that trans-vasing of some type occurred in this space.

Cookpot Two Handled

Cook

404/ 444

1c

RD: 16.5; MaxD: ca. 24.0; H: 15.2. Profile similar to Vasa Type 3B and Dhorios Form P111. Partially reconstructible vessel. Ca. 50% present, with single handle. Likely represents secondary use (dipper?; scoop?). Stratigraphic location seems to indicate it was stored on a shelf with other vessels.

Cookpot Single Handled Globular

Cook

554

1b

MaxD: 16.2; H: 12.0 (Max). Appears similar to Vasa Type 1F. Lower portion of a globular cooking vessel that was recovered from below floor level.

Amphora Vasa Type C

Amph

V476

1b

RD: 18.5; BD: 12.2; MaxD: 35.55; H: 43.5. Fig. 6.21. Reconstructible VTC amphora. Missing fragments most likely in loose sherd material from room.

Pitcher Single Handle

Cook

445

1c

Reconstructible vessel. No significant fragments missing.

Class

Reg. #

Type

Description

Bottle Cylindrical

Green

G138

1c

RD: 4.2; BD: 3.2; H: 11.5. Fig. 6.24. Largely reconstructible cylindrical glass bottle. Missing fragments most likely in loose glass recovered from the room.

Bottle Angular

Lt.Bl.

G139

1c

W: 4.2; H: 9.5. Half glass vessel preserved in situ. Missing portion of vessel at least partially recovered in remaining glass fragments recovered from the room.

Unguentarium (?)

C'less

UR

1c

Portion of vessel crushed below G139. Level of completeness unable to be determined with any confidence. Association with G139 suggests that the vessels were stored together.

Bottle (?) Cylindrical (?)

Green

UR

1c

Glass vessel fragments, which most likely can be reconstructed into a complete or nearly complete glass vessel similar to G138.

Bottle Angular

C'less

UR

1c

Glass vessel fragments, which most likely can be reconstructed into a complete or nearly complete glass vessel similar to G139.

Juglet

C'less

G137

2

RD: 4.0. Partially reconstructible glass vessel including rim and handle.

Glass: Object

*The remainder of the glass from Room 6 can be identified as either Class 2 or Class 4.

Metal: Object

Class

Reg. #

Type

Description

Hook / Chain Link (?)

CA

M151

2;4

L: 1.9. Small copper alloy hook. Also possibly a link for use in a copper alloy chain.

No other metal objects were recovered as part of the room assemblage.

125

Room 6 Assemblage (Cont.): Stone: No stone objects were recovered as part of the room assemblage.

Bone: No bone objects were recovered as part of the room assemblage.

Room 7 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description

CRSW Bowl Kourion Type I

Fine

438/ 430/ 418

4

RD: 21.5; BD: 8.5; H: 6.3. Partially reconstructible vessel. Overall completeness does not allow the vessel to be in use. Post-earthquake discard.

Amphora Kourion Type I

Amph.

477

1b;2

BD: 5.1; MaxD: ca. 35.0; H: 57.0 (Max.). Partially reconstructible vessel. Neck and one handle missing.

Amphora Kourion Type I

Amph.

499a

2

RD: 11.7.Partially reconstructible vessel.

Amphora Kourion Type I

Amph.

499b

2

RD: 10.4; BD: 4.6; H: 36.2 (Max.). Partially reconstructible vessel.

Amphora Kourion Type I

Amph.

535

2

BD: 4.8. Reconstructible neck and base.

Amphora Kourion Type I

Amph.

P510

1b;2

RD: 11.2; BD: 4.3; MaxD: 38.96; H: 66.3. Largely reconstructible vessel.

Amphora Kourion Type I

Amph.

535

1b;2

RD: 12.6; BD: 4.7; MaxD: 35.49; H: 65.6. Largely reconstructible vessel.

Amphora Kourion Type I

Amph.

551

1b

RD: 12.2; BD: 3.9; MaxD: 37.71; H: 68.3. Reconstructible vessel.

Amphora Kourion Type I

Amph.

V475

2

Partially reconstructible vessel.

Amphora Kourion Type I

Amph.

UR

2

RD: 13.4. Reconstructible neck.

Amphora Kourion Type III

Amph.

469

2;4

Amphora Kourion Type III

Amph.

V503

2

RD: 10.8; MaxD: 31.67; H: 36.5 (Max). Partially reconstructible vessel. Good probability for reuse (funnel).

Amphora Kourion Type III

Amph.

V536

1b

RD: 8.3; MaxD: 29.73; H: 76.7. Reconstructible vessel.

Amphora Kourion Type III

Amph.

V519

1b

RD: 9.8; MaxD: 34.05; H: 81.1. Reconstructible vessel.

Amphora Kourion Type V/ Late Roman 1

Amph.

545

2

RD: 6.4; H: 53.2 (Max). Partially reconstructible vessel.

Amphora Kourion Type V/ Late Roman 1

Amph.

V539

2

RD: 6.9; MaxD: 34.21; H: 44.5 (Max). Partially reconstructible vessel.

Amphora Kourion Type V/ Late Roman 1

Amph.

V546

2

RD: 6.1; MaxD: 34.21; H: 32.4 (Max).

Amphora Kourion Type V/ Late Roman 1

Amph.

V538

2

RD: 6.9; MaxD: 33.16; H: 51.0 (Max). Reconstructible except for bottom.

Amphora Kourion Type V/ Late Roman 1

Amph.

V552

1b

RD: 6.4; MaxD: 31.22; H: 52.7. Fig. 6.26. Reconstructible vessel. Only vessel of this type that preserves bottom. Missing fragments likely in loose sherd material.

RD: 11.5. Partially reconstructible vessel.

126

Partially reconstructible vessel.

Room 7 Assemblage (Cont.): Pottery (Cont.): Object

Class

Reg. #

Type

Amphora Kourion Type V/ Late Roman 1

Amph.

V501

2

Description Partially reconstructible vessel.

*The remainder of the ceramic vessels from Room 7 can be identified as Class 2 or Class 4.

Glass: No Class 1 glass vessels were part of the room assemblage. The glass from Room 7 can be identified as Class 2 or Class 4.

Metal: Object Coin

Metal

Reg. #

CA

C371

Class

Description

4

Single copper alloy coin of unknown denomination. Identified by E. Lane in notes as belonging to Probus (AD 276-282). Recovered at an elevation well above the occupational surface. Material deposited/redeposited post-earthquake.

Stone: No stone objects were recovered as part of the room assemblage.

Bone: Two bone objects (BI9 and BI10) were recovered from this space. Based on their findspots both appear to be Class 4 objects.

Room 8 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description RD: 49.0; BD: 28.0; H: 17.0. Fig. 6.28. Largely reconstructible two handled plainware basin. Remaining missing fragments likely in the loose sherd material from room. RD: 9.4. Largely reconstructible Kourion Type III amphora. Not reconstructed. Vessel appears to have been intact and fell during the earthquake.

Basin

Plain

V560

1b

Amphora Kourion Type III

Amph

UR BSK 119

1b

Amphora Kourion Type III

Amph

UR

1b;4

RD: 10.7. Largely reconstructible Kourion Type III amphora. Not reconstructed. Context uncertain, possibly post-earthquake discard.

Amphora Vasa Type C

Amph

594

2

BD: 10.5; H: ca. 45.5. Partially reconstructible VTC amphora. Rim appears to have been removed as a result of maintenance processes.

CRSW Bowl Form 1

Fine

579/ 601

1b

RD: 14.5; BD: 4.8; H: 4.6-4.2. Fig. 6.30. Reconstructible fineware bowl. Missing fragments account for ca. < 2% of the vessel.

Cookpot Dhorios Type P132

Cook

567

1c;4

RD: 12.0; H: 11.0. Largely reconstructible. Some fragments may be present in the loose sherd material but not enough for complete reconstruction. Possibly deposited post-earthquake.

Cookpot Two Handle

Cook

584

2

Basin / Mortarium

Plain

578

1c;4 (?)

RD: 22.0; H: 11.0 (Max). Partially reconstructible cookpot. RD: 44.5; BD: 19.0; H: 12.4. Rim profile similar to Kourion Basilica L14. Reconstructible except for a section of the base. Some fragments possibly in loose sherd material, but unlikely that all could be located. Recovered high in the stratigraphic sequence, which may indicate postearthquake discard.

127

Room 8 Assemblage (Cont.): Pottery (Cont): Object

Ware

Reg. #

Class

Description RD: 41.6; BD: 17.0; MaxD: 57.45; H: 67.8. Fig. 6.29. Profile Similar to Kalavasos-Kopetra 215 and Kourion Survey 63n. Largely reconstructible pithos with a drain hole in the base. Base of the vessel recovered from the occupation surface. Not post-earthquake discard.

Pithos

Plain

V568

1b (?);2

CRSW (LRC?) Cup

Fine

V550

1b

Egg-Shaped Bottle

Plain

V561

1b

CRSW PlateForm 1

Fine

555

4

Lamp Vessberg 18

Lamp

597

2

Reconstructible fineware cup. On display in museum. Same type as Agora P25137. Reconstructible bottle with a wheel made egg-shaped body and separately added wheel made neck. No base. Likely of Egyptian origin. Largely reconstructible fineware plate. Rim complete. Large sections of vessel floor and base missing. Likely post-earthquake material. Rosette type. Fragment has discus and handle.

*The remainder of the ceramic vessels identified from Room 8 can be identified as Class 2 or Class 4.

Glass: Object

Color

Reg. #

Class

Description

Glass Fragments

C'less

UR

1b (?); 1c (?)

Fragments consolidated in situ. Seven pieces minimum, but total number unknown at present. All body fragments. Unknown if the fragments represent a complete vessel or only part of a vessel.

*The remainder of the glass from Room 8 can be identified as Class 2 or Class 4.

Metal: Object

Metal

Reg. #

Class

Description

Rod(?) / Bar(?) / Spike(?)

FE

M174

1b

L: 15.4; D: 2.6-1.5. Large, heavy cylindrical iron object. Spike? Diameter tapers towards bottom of the object. Could also possibly be a chisel, punch, or awl. Recorded as recovered in situ and associated with a sherd concentration.

Fork

FE

M175

1b

L: 10.3; W: 7.7; H: 1.5. Fig. 6.33. Asymmetrical fork-shaped iron object. Four tines. Exterior tines formed by the u-shaped frame of the object, with the two interior ones added separately. No evidence of a handle or hafting socket. Small size. Not a pitchfork. Possibly has an agricultural function. Tempting to see the object as a quasitrident for fishing, but cannot be positively identified as such since it does not possess barbs.

Lead Sheet

PB

UR

1b(?);2

L: 2.5; W: 2.5. Small fragment of lead sheet. Folded slightly. Possibly raw material.

Nail(?) / Pin(?)/ Needle(?)

CA

UR BSK 166

1b;2;4

L: 1.95; D: 0.2. Small cylindrical copper alloy fragment. Appears to taper to a point at one end. May be the shank of a nail, but could also possibly be part of a pin or needle. Size conducive to infiltration. Possibly post-earthquake.

Iron Object

FE

M191

1b

L: 24.1; W: 4.1-3.5. Thin, slightly concave, rectangular iron object. Tapers slightly towards one end. Rounded corners. Does not appear to be a knife blade. Note on tag reads "Lg. Iron Piece w/ Table w/ Ass. Charcoal." Some sort of tool, but function unknown at present (pry-bar?; agricultural implement?).

Copper Alloy Object

CA

M179

1b;2

L: 3.3; W: 1.3-1.0. Fig. 6.31. U-Shaped copper alloy object of unknown function. Note on tag records fibers being attached, but no notation of the type. Possibly tweezers, but more likely some type of pin, probably for a garment. Not a nail.

128

Room 8 Assemblage (Cont.): Metal (Cont): Object

Class

Reg. #

Type

Description

Copper Alloy Object

CA

M185

1b;2

L: 2.6; W: 1.6-1.2. Fig. 6.31. U-shaped copper alloy object of unknown function. Ends relatively sharp. Similar to other objects of this type from the room.

Copper Alloy Object

CA

M192

1b;2

L: 2.5; W: 1.3-0.8. Fig. 6.31. U-Shaped copper alloy object of unknown function. Note on tag reads "In sieve. In strip of dirt from column to Wall N." Appears to be a pin, as the ends are relatively sharp. Does not appear to be a nail.

Copper Alloy Object

CA

M178

1b;2

L: 3.0; W: 1.2-1.0. Fig. 6.31. U-Shaped copper alloy object of unknown function. Fibers recorded as being preserved on surface. Presence of material identified as cloth fibers seems to confirm the identification of this object as some type of pin rather than a nail or tweezers. Ends blunted.

Pitcher

CA

M164

1b;1c

Complete copper alloy pitcher recovered along Wall K.

Pitcher

CA

M165

1b;1c

Complete copper alloy pitcher recovered along Wall E.

Cup

PB

M176

1b; 1c (?)

RD: 4.5-3.3; BD: 4.5; H: 3.6. Fig. 6.34. Small, complete lead vessel. Small size does not indicate drinking vessel. Top slightly broken and bent. Solid bottom, slightly concave. Unknown function.

Spearhead

FE

M177/M1 80

1b

L: 26.5; W (Blade): 4.0 (Max); D (Socket): 3.0-2.4. Fig. 6.32. Shape appears similar to Manning Type V105. Fragments heavily corroded and friable.

Proportional Dividers

CA

M183

1b

Copper alloy dividers. Discovered on Locus 005 close to Wall J.

Lamp

CA

M161

1c

Large copper alloy hanging lamp. chains intact.

Coin

CA

C386

1b;2

Stone

Reg. #

Type

Description

Mortarium

LS

S75/ S86/ S87

1b

RD: 34.0-32.0; H: 7.5. Fig. 6.35. Three joining fragments of a complete stone mortarium. Roughly round. Rim well-worn in some places which appears to indicate relatively extensive use. Two lug handles on opposite sides.

Basin / Trough

LS

S84

1b

L: 31.7; W: 27.1; H: 17.6. Largely complete limestone basin recovered upside-down on the occupation surface. Roughly one half of the object walls are heavily damaged, broken off in some cases all the way down to the floor of the basin. No joining fragments recovered.

Tabletop & Legs

Alab/ LS

S88/ S89/ S90/ S91

1b

L: 86.0; W: 80.0; H: 7.0. Large relatively complete alabaster tabletop. Fragmented in ca. 40 pieces, three of which are large. Separate limestone legs from a different object. Only three legs present. Fourth leg appears to have been of ephemeral material or is missing altogether. Located on the west side of the central column.

Hanging hook and suspension

Copper alloy coin. Recovered in situ on floor.

Stone: Object

Bone: No relevant bone objects were recovered as part of the room assemblage.

129

Room 14 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description

Amphora Vasa Type C

Amph

P665

1b;2

BD: ca. 10.0; MaxD: 40.0; H: 25.0. Largely reconstructible lower half of a VTC amphora. Edge appears very regular, and the top appears to have been removed intentionally as a result of maintenance processes.

Amphora Vasa Type C

Amph

V619

1b

BD: 13.8; MaxD: 50.7; H: 26.3 (Max). Fig. 6.40. Largely reconstructible VTC amphora. Rim and neck, as well as one handle completely missing. Neck section connecting to other handle preserved. Fragments of the neck and rim do not appear to have been recovered, which indicates breakage prior to the earthquake.

Amphora Vasa Type C (?)

Amph

P1301934

1b

Fig. 6.38.B. Probable VTC amphora recovered during the 1934 excavation. Unknown how many handles vessel possessed.

CRSW Plate Form 1

Fine

V622

1b

RD: 22.8; BD: 10.4; H: 5.6. Fig. 6.41. Complete unbroken CRSW Form B1 dish found in the center of the room near the oven. Pronounced fingerprint marks on interior and exterior of vessel.

1b

RD: 18.5(Long)-11.5(Short); BD: 4.1; MaxD: 40.2; H: 84.0. Fig. 6.39. Reconstructible vessel. Small missing fragments most likely in sherd material from the room. Three dipinti: two on shoulders, one on neck. Significant manufacturing defects in the form of warped vessel walls, oval rim and elongated neck.

Amphora Kourion Type II

Amph

P615

Amphora Kourion Type IV / Late Roman 4

Amph

V621

1b

RD: 11.2; MaxD: 28.8; H: 62.0. Nearly complete reconstructible KTIV/LR4 amphora of the single handled variety. Small missing fragments likely in loose sherd material from the room. Red dipinto on shoulder.

Amphora Kourion Type IV / Late Roman 4

Amph

P1291934

1b

Fig. 6.38.A. In situ KTIV/LR4 amphora of the two handled variety recovered during the 1934 excavations. Located in northeast corner of the room. Rim area appears to have been broken, with a relatively small section of the shoulder missing.

Cookpot Two Handle

Cook

P1311934

1b

Fig. 6.38.C. Largely reconstructible two handled, hemispherical bottomed cookpot. Recovered during the 1934 excavation in the northeast corner of the room. Only intact cookware vessel recovered from this room.

Pithos

Plain

V617

1b

RD: 46.2; MaxD: 66.14; H: 80.6. Fig. 6.37. Similar in profile to Kalavasos-Kopetra 206. Complete unbroken, moderately-sized pithos from the northwest corner of the room. Small solid pedestal base. Held in place by three stones around the base. Hole in vessel wall has ancient repair in plaster. Smaller hole in lower part of vessel also appears to be ancient.

Jug Vasa Type 3G

Plain

V618

1b

RD: 7.5; BD: 8.1; MaxD: 18.74; H: 25.8. Fig. 6.42. Largely reconstructible Vasa Type 3G jug. Ring base, round body. Rounded, slightly thickened rim. Slight grooves around center of body. Much of rim missing. Likely broken prior to earthquake.

Bowl

Fine

P1331934

1b;1c

Fig. 6.38.D. In situ vessel recovered during the 1934 excavation. Described in notes as "red bowl," which seems to indicate a fineware vessel.

Basin

Plain

V620

1b

RD: 38.2; H: 14.6. Completely reconstructible plainware basin with thumb impressed handles on either side.

2

RD: 17.5; MaxD: 21.5; H: 11.4 (Max). Fig. 7.10. Partially reconstructible two-handled, hemispherical bottomed cookpot. Recovered from the discard area east and south of the oven. Joining fragments recovered from Room 12A, Locus 003 and Room 19, Locus 004.

Cookpot Two Handle

Cook

V657

130

Room 14 Assemblage (Cont.): Pottery (Cont.): Object

Ware

Reg. #

Class

Description

African Red Slip Plate - Form 67

Fine

UR

2

RD: 18.0. Rim fragment of ARS Hayes' Form 67. Recovered from the discard area located east and south of the oven. Joining rim fragment recovered from Room 19, Locus 004.

Cookpot Globular

Cook

V658

2

RD: 15.0; MaxD: 18.5; H: 12.3 (Max). Partially reconstructible cookpot. Recovered from the discard area located east and south of the oven.

Cookpot Two Handle

Cook

V656

2

RD: 27.0; MaxD: 36.0; H: 16.2 (Max). Partially reconstructible moderately sized two handled cookpot. Recovered from the discard area located east and south of the oven.

Lid Storage

Plain

P637

2

D: 26.0. Joining fragments of a probable flat plainware lid. Sooting on bottom appears to indicate reuse as some type of cooking base. Recovered from the discard area located east and south of the oven.

2

RD: 31.0-32.0; H: 12.0 (Max). Joining fragments of a casserole. Heavy sooting. Non-joining, but probable fragments of this vessel recovered from Room 12, Locus 003. Recovered from the discard area located east and south of the oven.

Casserole

Cook

P636

Bowl

Plain

V659

2

MaxD: 12.5; H: 6.5 (Max). Partially reconstructible vessel, most likely a bowl. Majority of fragments recovered from the discard area south of the oven, but some recovered in the northern part of the room. Dispersal as well as size of fragments indicates material in discard for an extended period of time.

Storage Vessel

Plain

UR

2

RD: 22.0. Rim fragment of a probable storage vessel. Recovered from the discard area located east and south of the oven. Joining rim fragment recovered from Room 19, Locus 004.

Lamp Vessberg 18

Lamp

P635

2

RD: 8.5 (Max)-6.2(Min); BD: 3.0; H: 2.5. Fig. 7.11. Intact top half of a Vessberg 18 lamp of the "galloping horse" type. Recovered from the discard area located east and south of the oven. Joining fragments belonging to the lower half recovered from Room 19, Locus 004.

Lamp Vessberg 18

Lamp

UR

2

RD: 6.7; H: 2.3 (Max). Joining fragments that form nearly the complete upper part of a Vessberg 18 lamp of the "galloping horse" type. Recovered from the discard area located east and south of the oven.

Lamp Vessberg 18

Lamp

P690

2

RD: 7.5; H: 2.5 (Max). Joining fragments that form part of the discus and nozzle of a Vessberg 18 lamp of indeterminate type. Recovered in the northwest corner of the room close to the pithos.

Lamp Unknown

Lamp

P638

2

RD: 7.5; H: 1.7 (Max). Fragment of discus and shoulder that appear to belong to a Vessberg 18 of the "rosette" type. Recovered from the discard area located east and south of the oven.

Lamp Vessberg 18

Lamp

P689

2

RD: 8.0-7.0; H: 1.5 (Max). Joining fragments that form part of the discus and the handle of a Vessberg 18 lamp of the "galloping horse" type. Fragments recovered in the northwest quadrant close to the room center.

*The remainder of the ceramic vessels from Room 14 can be identified as Class 2 or Class 4.

131

Room 14 Assemblage (Cont.): Glass: Object

Bowl

Bowl

Color

C'less

C'less

Reg. #

G147

G148

Class

Description

1c

Fig. 6.43. Rim of a relatively large glass open form vessel, most probably a bowl. Fragments consolidated in situ. Remainder of glass completely fragmentary and unable to be reconstructed. Wide, outturned rim. Based on the number of fragments, if they could be reconstructed, its overall completeness would be high.

1c

Fragments of an open form vessel, most probably a bowl. Rim fragment indicates a flared rim. Overall level of completeness is moderate, but the fragments were recovered in proximity to the boundary of the 1934 excavation. Proximity to G147 would seem to indicate that the vessels were likely stored together.

*The remainder of the glass from Room 14 can be identified as Class 2 or Class 4.

Metal: Object

Metal

Reg. #

Class

Description

Pitcher

CA

M193

1c

Complete copper alloy pitcher. Recovered on the east side of the oven. Flared rim. Single handle. Surface around neck of vessel engraved with decorative design. Object likely stored on a shelf, either on Wall D, or on Wall T.

Fishhook

CA

M198

1c (?); 2 (?); 4 (?)

Complete copper alloy fishhook. Recovered from the area of the room excavated in 1934. Object may be out of context contamination.

Coin

CA

C396

1b;2

D: 1.1. Unidentified. Recorded as being found on occupation surface.

Coin

CA

C399

1b;2

D: 1.6. Unidentified. Recorded as being found in situ.

Coin

CA

C400

1b;2

D: 1.4. Unidentified. Recorded as being found in situ under Locus 004.

Coin

CA

C401

1b;2

D: 1.25. Unidentified. Recorded as being found under Locus 004.

Object

Stone

Reg. #

Class

Description

Mortarium

Unk

ST13 1934

1b;1c

Fig. 6.38.E. Complete stone mortarium from the northwest corner of the room during the 1934 excavation. Appears relatively small. Rounded lug type handle similar to the larger Room 8 example.

Gab

UR BSK 118

1c

L: 10.9; W: 6.4; H: 6.7 (Max). Fig. 6.81. Groundstone tool. Made from a heavy, dense, fine grained stone, probably Gabbro. Roughly oval in shape. Primary work surface located on the bottom of the object. Additional work areas that display abrasion due to probable "pecking" on the upper and side surfaces.

Object

Reg #

Type

Description

Needle(?) / Punch(?) /Awl(?)

B58

1b (?); 1c;2

L: 3.0; W: 1.0 (Max); H: 0.8 (Max). Fig. 6.44. Single complete bone object. Identified as a needle but more likely an awl or punch. Note on tag reads "Found N of Wall Q."

Stone:

Groundstone

Bone:

132

Room 20 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description

Amphora Unknown

Amph

P687

1b (?);2

RD: 12.3. Largely reconstructible amphora/storage vessel of uncertain type. Not reconstructed. Generally ovoid body with very short spike.

Jug Vasa Type 1H

Plain

UR

2

RD: 17.5. Largely reconstructible rim and neck of a probable Vasa Type 1H jug. Possible dipinto present.

Storage Vessel(?) / Jug(?)

Plain

UR

2

RD: 17.5. Partially reconstructible rim of a small storage vessel, probably a jug of some type. Fragments found throughout the room.

Lamp Vessberg 18

Lamp

UR

2

Lamp fragments, some of which join. Located in the northeast part of room.

Mortarium

Plain

UR

2

RD: 40.0; BD: 18.0. Fragments of a mortarium, some of which join. Appears similar to the example from Room 8. Completely reconstructible base. Single large rim fragment.

Egg-Shaped Bottle

Plain

UR

2

Fragments of an egg-shaped bottle similar to the example from Room 8. Some joining pieces.

Storage Vessel Jug

Plain

UR

2

RD: 10.0; MaxD: 11.6; H: 10.5 (Max). plainware jug.

Partially reconstructible

*The remainder of the ceramic vessels from Room 20 can be identified as either Type 2 or Type 4.

Glass: Object

Color

Reg. #

Class

C'less

UR BSK 1407

1b;1c

Object

Metal

Reg. #

Class

Description

Metal Disc / Coin (?)

CA

UR

1b (?);2

D: 1.8-1.6. Corroded and uncleaned. Definitive identification not possible at present. Possibly a small coin.

Tintinnabulum

CA/FE

UR BSK 1412

1c

D: 2.5; H: 2.4. Fig. 6.47. Object heavily corroded and uncleaned. Conical copper alloy piece with attached iron suspension loop and clapper. Very fragile.

Ring (Finger) Chi Rho

CA

UR

1a

Copper alloy ring with round bezel. Inscribed with the Chi-Rho symbol as well as the alpha on one side. Found in association with the human remains.

Ring (Finger)

FE

UR BSK 566

1a

D: 1.6. Fig. 6.46. Round bezel of an iron finger ring which preserves fragments of the iron band on either side. No traces of decoration on upper or lower surfaces of the bezel. Found in association with the human remains.

Lion Paw

CA

UR

2

Copper alloy lion paw. examined.

Amphoriskos / Juglet

Description BD: 4.5. Small, single handle juglet. Most reconstructible. High degree of brokenness.

likely

completely

Metal:

Possibly a lamp foot.

Not located. Not

Stone: No relevant stone objects were recovered as part of the assemblage of the room.

Bone: Object

Reg #

Type

Description

Hair Pin

UR

1a

Bone hair pin found next to the cranium of the female victim. In place in the victim’s hair at the time of death. Object preserved in situ as part of block lift of victims. Unable to be examined. Carved cylindrical head. Shaft tapered to a point.

133

Room 25 Assemblage: Pottery: Object

Ware

Reg. #

Class

Description

CRSW Jar Hayes Form 12A

Fine

V677

2

RD: 14.2; BD: 8.5; H: 14.0 (Max). Fig. 6.49. Largely reconstructible decorated Hayes Form 12A fineware jar. Found associated with the Dhorios Type P132 and KTIII amphora in the northwest corner of the room. Rim intentionally removed prior to the earthquake as a result of maintenance processes. Large hole in side.

Cookpot Dhorios Type P132

Cook

V678

2

RD: 12.1; H: 10.4. Largely reconstructible small Dhorios Type P132 cookpot. Vessel does not appear to have been used, as no sooting or evidence of thermal exposure is present on the exterior or interior surfaces. Significant section of the vessel missing.

Amphora Kourion Type III

Amph

P683

1b

RD: 10.6. Reconstructible vessel. sherd material from room.

Amphora Kourion Type III

Amph

UR BSK 561

1b;2

RD: 10.7; MaxD: 31.4; H: 79.0. Largely reconstructible KTIII amphora. Relatively large section of the vessel shoulder, neck, and rim are missing. Some fragments may be present in the remaining sherd material found in the room.

Amphora Kourion Type III

Amph

V679

1b

Complete and unbroken KTIII amphora. Found in the northwest corner of the room with the Hayes Form 12A jar (V677) and Dhorios Type P132 cookpot (V678).

Amphora Vasa Type C

Amph

UR

2

RD: 19.0. Partially reconstructible Vasa Type C amphora.

Pithos/Amphora (?)

Plain

P684

1b

RD: 26.0; BD: 10.8; MaxD: 40.7; H: 46.5. Fig. 6.50. Completely reconstructible two handled storage vessel that could either be classified as an amphora or a pithos with handles. Small missing fragments likely in loose sherd material.

Pithos

Plain

UR BSK 236, 252, 564

2

RD: ca. 60.0; H: ca. 97.0 (Min). Partially reconstructible large pithos. Solid pedestal base.

Amphora Kourion Type III

Amph

UR

2

Partially reconstructible KT III amphora.

Missing fragments likely in loose

*The remainder of the ceramic vessels from Room 25 can be identified as Class 2 or Class 4.

Glass: No relevant glass objects were recovered as part of the room assemblage.

Metal: No metal objects were recovered as part of the room assemblage.

Stone: Object

Stone

Reg. #

Type

Description

Mortar

Gran

S109

1b

L: 33.6; W: 32.1; H: 12.5. Fig. 6.51. Large, irregularly shaped heavy granite mortar with a deep, hemispherical basin. Basin D: 17.0.

Pestle

Gran

S110

1b;1c

D: 5.8-3.1; L: 13.8. Fig. 6.52. Heavy, granite groundstone pestle. Same type of stone as the mortar.

Pestle

Unk

S111

2;4

D: 5.2-2.5; L: 15.6. Groundstone pestle. Different material than other two groundstone objects from this room. Fine-grained, green-grey stone. Object not complete, with the larger end being broken.

Bone: No relevant bone objects were recovered as part of the room assemblage.

134

Appendix B: Kourion Amphora Typology and Capacity Estimates

135

136 After Williams 1987.

Appendix C: Tables

137

Table 1: Earthquake House Room Summary Room Number

Dimensions (m: N-S X EW)

Total Area (m2)

Pavement Type

Roofing Type

1 (Prior to Division)

3.25 X 6.25

21.12

Compact Earth

Wood & Reed Matting (Upper Floor)

None

1

3.25 X 2.4

7.8

Compact Earth

Wood & Reed Matting (Upper Floor)

None

2

3.2 X 3.0

9.6

Compact Earth

Wood & Reed Matting (Upper Floor)

Semi-Perm:Trough

3

3.5 X 2.75

9.62

Compact Earth

Wood & Reed Matting (Upper Floor)

None

6

3.7 X 3.05

11.28

Compact Earth

Sloped/Flat(?): Wood, Reed Matting, Mud

Perm:

Sloped/Flat(?): Wood, Reed Matting, Mud. Some Tile.

None

7

3.24 X 2.9

9.39

Compact Earth

8

7.65 X 6.4

48.96

Compact Earth

Pitched: Tile

None

Fixtures (Permanent & Semi-Permanent)

Perm:

Stair Foundation (?)/Bench (?)

Central Column

Semi-Perm: Stone Table, Grinding Station

11

7.7 X 13.4

103.18

Earth; Compact Earth

12A

2.2 X 1.25

2.75

Hydraulic Cement

None (?)

14

3.14 X 2.82

8.85

Compact Earth

Flat (?): Timber, Reed Matting, Mud or None

Perm:

Oven

15

4.1 X 7.6

31.16

Stone Paving

None

Perm:

Staircase, Basin (?)

18

4.0 X 2.6

10.4

Compact Earth

Sloped: Wood, Reed Matting, Mud

Perm:

Ovens (2)

19

2.96 X 1.16

3.43

Compact Earth

Stairway; Wood (?)

None

20 (Prior to Division)

3.1 X 6.85

21.23

Compact Earth

Flat(?): Timber, Reed Matting, Mud

None

20

3.1 X 3.3

10.23

Compact Earth

Flat(?): Timber, Reed Matting, Mud

None

25

3.1 X 2.8

8.68

Compact Earth

Flat (?): Timber, Reed Matting, Mud

None

138

Perm:

Oven

None (Cistern)

Table 2: Individual Vessel Capacities† Room

Vessel Pithos (P1)

1

Storage Vessel (P134-1934)

3

Vasa Type 1H Jug (187)

6

7

8

20 25



10 - 15 5-7

Pithos (V443)

97 - 98

VTC Amphora (V467)

10 - 15

KT I Amphora (V478)

35 - 36

KT I Amphora (V502)

35 - 36

KT II Amphora (480/500)

35 - 36

KT III Amphora (V403)

25 - 26

KT III Amphora (474-2)

25 - 26

KT III Amphora (V521/533)

25 - 26

KT III Amphora (P522)

25 - 26

KT III Amphora (534/541)

25 - 26

KT III Amphora (V542)

25 - 26

KT I Amphora (V551)

35 - 36

KT III Amphora (V519)

25 - 26

KT III Amphora (V536)

25 - 26

LR1 Amphora (V546) KT III Amphora (UR: BSK 119)

13 - 14

Pithos (V617)

14

Capacity (lt.) 34 - 35

25 - 26 151 - 152

KT II Amphora (V615)

35 - 36

LR4 Amphora (V621)

28 - 29

LR4 Amphora (P129 - 1934)

28 - 29

VTC Amphora (V619)

15 - 20*

VTC Amphora (P130 - 1934)

10 - 15

Amphora - Unknown (P687)

20 - 25

Pithos (P684)

35 - 36

KT III Amphora (V679)

25 - 26

KT III Amphora (P683)

25 - 26

All capacities approximate. Volumes calculated using the "stacked cylinder" methodology (c.f. Nelson 1985, 313). * Reflects diminished capacity due to breakage.

139

Table 3: Foodstuff Weight/Volume Equivalencies Commodity

Weight

Olive Oil

Volume 3

.81 kg/liter

3

.85 kg/liter

3

1.0 kg/liter

.81 gr/cm

Lentils/Pulses

.85 gr/cm

Water

1.0 gr/cm

Wheat

.72-.83(.775 avg.) gr/cm3

.775 kg/liter

Wine

.978 gr/cm3

.978 kg/liter

Table 4: Yearly Individual Foodstuff Requirements Commodity Olive Oil Lentils/Pulses

Quantity 17 kg (20 liters)/person/yr. 16 kg/person/yr.

Wheat/Cereals

166 kg/person/yr.*

Wine

38 kg/person/yr.

* This is the quantity derived by Christakis (2008). acknowledge that this figure is most likely too high.

Foxhall and Forbes propose an amount of 212 kg/person/yr., but

Table 5: Earthquake House Yearly Foodstuff Calculations Population: 5 Adults; 1 Juvenile‡ Commodity Olive Oil Lentils/Pulses Wheat/Cereals Wine

Calculation (20 X 5) + (20 X .70) = 100 + 14 (16 X 5) + (16 X .70) = 80 + 11.2 = 91.2 kg (166 X 5) + (166 X .70) = 830 + 116.2 = 946.2 kg 38 X 5 = 190 kg

Total 114.0 liters 107.29 liters 1220.9 liters 194.27 liters

Total olive oil, lentils/pulses, wheat/cereals and wine consumed by the Earthquake House inhabitants: 1636.46 liters/year

Total Type A storage in the Earthquake House for olive oil, lentils/pulses, wheat/cereals and wine: 268.8 - 285.6 liters.††

‡ Foxhall and Forbes (1982, 72) estimate that the yearly requirements for a female child aged 10-12 years are 70% of the adult requirements.

††

Based on the estimate by Christakis (2008, 32) that 80% of total storage space is utilized for these basic commodities.

140

Bibliography Adam, J.P. 1989. “Osservazione tecniche sugli effetti del terramotto di Pompei del 62 d.C.” In I terramoti prime del Mille in Italia e nell’area Mediterranea, edited by E. Guidaboni, 224-43. Bologna: SGA Storis-Geofisica-Ambiente.

Ault, B., and L. Nevett. 1999. “Digging Houses: Archaeology of Classical and Hellenistic Domestic Assemblages.” In The Archaeology of Household Activities, edited by P.M. Allison, 43-56. London: Routledge.

Allison, P.M. 1992. “Artefact Assemblages: Not the Pompeii Premise.” In Papers of the Fourth Conference in Italian Archaeology Vol. 3.1, edited by E. Herring, R. Whitehouse, and J. Wilkins, 49-56. London: Accordia Research Center.

Badain, E. 1965. “M. Porcius Cato and the Annexation and Early Administration of Cyprus.” JRS 55. 110-21.

Allison, P.M. 1995. “House Contents in Pompeii: Data Collection and Interpretative Procedures for a Reappraisal of Roman Domestic Life and Site Formation Processes.” Journal of European Archaeology 3.1: 145-76. Allison, P.M. 1997a. “Why Do Excavation Reports Have Finds’ Catalogues?” In Not So Much a Pot, More a Way of Life: Current Approaches to Artefact Analysis in Archaeology. Monograph 83, edited by C.G. Cumberpatch and P.W. Blinkhorn, 77-84. Oxford: Oxbow. Allison, P.M. 1997b. “Artefact Distribution and Spatial Function in Pompeian Houses.” In The Roman Family in Italy: Status, Sentiment, and Space, edited by B. Rawson and P. Weaver, 321-54. Oxford: Clarendon Press. Allison, P.M. 1999. “Labels for Ladles: Interpreting the Material Culture from Roman Households.” In The Archaeology of Household Activities, edited by P.M. Allison, 57-77. London: Routledge. Allison, P.M. 2004. Pompeian Households: An analysis of the Material Culture. Monograph 42. Los Angeles: Cotsen Institute of Archaeology. Ames, K.L. 1981. “Meaning in artifacts: hall furnishings in Victorian America,” in Schlereth 1981: 206-21. Arthur, P. 1997. “Pottery in Structural Contexts,” in Ling 1997: 325-31. Ault, B.A. 1994. “Classical Houses and Households: An Architectural and Artifactual Case Study from Halieis, Greece.” Ph.D. Dissertation: Indiana University. Ault, B.A. 1999. “Koprones and Oil Presses at Halieis: Interaction of Town and Country and the Integration of Domestic and Regional Economies.” Hesperia 68.4: 549-73. Ault, B.A. 2005. The Excavations at Ancient Halieis Vol. 2: The Houses – The Organization and Use of Domestic Space. Bloomington & Indianapolis: Indiana University Press.

Bagnall, R.S. 1976. The Administration of the Ptolemaic Possessions Outside Egypt. Leiden: E.J. Brill. Bagnall, R.S. and Drew-Bear, T. 1973a. “Documents from Kourion: A Review Article Part 1: Principles and Methods.” Phoenix 27:2, 99-117. Bagnall, R.S. and Drew-Bear, T. 1973b. “Documents from Kourion: A Review Article Part 2: Individual Inscriptions.” Phoenix 27:3, 213-44. Bagnall, R.S. and Drew-Bear, T. 1974. “Notes on the History of Kourion.” ChrÉg 49, 179-95. Baurain, C. 1984. Chypre et la méditerranée orientale au Bronze Récent. Paris. Bekker-Neilson, T. 2004. The Roads of Ancient Cyprus. Copenhagen: Museum Tusculanum Press, University of Copenhagen. Beloch, J. 1923. Griechische Geschichte 3.2. Berlin and Leipzig: W. de Gruyter & Co. Benson, J.L. 1972. Bamboula at Kourion: The Necropolis and Finds. University Museum: Philadelphia. Benson, J.L. 1973. The Necropolis of Kaloriziki. Studies in Mediterranean Archaeology (SIMA) 36. Berger, A. 1992. Reading Matter: Multidisciplinary Perspectives on Material Culture. New Brunswick: Transaction Publishers. Berry, J. 1997. “Household Artefacts: Towards a Re Interpretation of Roman Domestic Space.” In Domestic Space in the Roman World: Pompeii and Beyond, edited by R. Laurence and A.F. Wallace Hadrill, 183-96. JRA Supplementary Series No. 22. Portsmouth: JRA. Bonifay 2004. Etudes sur la céramique romaine tardive d’Afrique. BAR-IS 1301. Oxford: BAR Publishing. Bosio, L. 1983. La Tabula Peutingeriana, una descrizione pittorica del mondo antico. Rimini: Maggioli. Bradley, R. and M.G. Fulford. 1980. “Sherd Size in the Analysis of Occupation Debris.” BIALond 17: 85-94.

141

Buell, D.M., Mavromatis, C. and Parks, D.A. 2009. “The Kourion Mapping Project: a contextual spatial analysis of the Kourion acropolis.” RDAC: 261-294. Buitron, D and Soren, D. 1982. “Sanctuary of Apollo Hylates.” In An Archaeological Guide to the Ancient Kourion Area and the Akrotiri Peninsula edited by H. W. Swiny, 58-67. Nicosia: Department of Antiquities, Cyprus. Buitron-Oliver, D. 1996. The Sanctuary of Apollo Hylates at Kourion: Excavations in the Archaic Precinct. Studies in Mediterranean Archaeology Vol. CIX. Jonsered, Sweden: Paul Ǻströms förlag. Buitron-Oliver, D. 1997. “Kourion: The Evidence for the Kingdom from the 11th to the 6th Century B.C.” In The City-Kingdoms of Early Iron Age Cyprus in Their Eastern Mediterranean Context. BASOR 308. 27-36. Bullard, R. 1987. “Environmental Geology.” In Soren, 1987a. 53-71. Cahill, N. 2002. Household and City Organization at Olynthus. New Haven: Yale University Press. Catling, H.W. 1972. “An Early Byzantine Pottery Factory at Dhiorios in Cyprus.” Levant: 4: 1-82. Cesnola, L.P. 1877. Cyprus: Its Ancient Cities, Tombs, and Temples. London: John Murray. Chase, P.G. 1985. “Whole Vessels and Sherds: An Experimental Investigation of their Quantitative Relationships.” JFA 12: 213-218.

Christou, D. 2001. Kourion: Its Monuments and Local Museum (8th Edition). Nicosia: Filokipros. Cohen, D. 1912. de magistratibus Aegyptiis externas Lagidarum Regni provincias administrantibus. The Hague. Connelly, J.B. and Wilson, A. 2002. “Hellenistic and Byzantine Cisterns on Geronisos Island.” RDAC: 26986. Costello IV, B. 2011. “The Waste Stream of a Late Roman House: Case Study of the Commissary Block in the Earthquake House at Kourion.” In: “Pottery in the Archaeological Record: A View from the Greek World.” A Workshop on J. Theodore Peña, Roman Pottery in the Archaeological Record edited by M. Lawall and J. Lund. The Danish Institute at Athens; June 20-21 2008. Proceedings of the Danish Institute at Athens. 75-84. Crawford, J.S. 1990. The Byzantine Shops at Sardis. Cambridge: Harvard University Press. Daniel, J.F. 1937. “Two Late Cypriot III Tombs from Kourion.” AJA 41. 56-85. Daniel, J.F. 1938. “Excavations at Kourion: The Palace.” UPMB 7:2, 4-10. Darnay, A. and Franklin, W.E. 1972. Salvage markets from materials in solid wastes. Washington D.C.: U.S. Government Printing Office. Davis, T., and Soren, D. 1985. “Seismic Archaeology at Kourion: The 1984 Campaign.” RDAC: 293-306.

Chiaramonte Treré, C. 1986. Nuovi contribute sulle fortificazioni pompeiane. Quaderne di Acme 6. Milan: Cisalpino-Goliardica.

Davis, T.W. 1989. “A History of American Archaeology on Cyprus.” BiblArch 52:4, 163-9.

Christakis, K.S. 2008. The Politics of Storage: Storage and Sociopolitical Complexity in Neopalatial Crete. Philadelphia: INSTAP Academic Press.

Davis, T.W. 2003. Shifting Sands: The Rise and Fall of Biblical Archaeology. Oxford: Oxford University Press.

Christou, D. 1983. “Excavations at Kourion, First Preliminary Report, 1975-1982.” RDAC. 266-74.

Deal, M. 1985. “Household Pottery Disposal in the Maya Highlands: An Ethnoarchaeological Interpretation.” JAnthArch 4: 243-91.

Christou, D. 1985. “Excavations at Kourion, 1975-1984.” In Archaeology in Cyprus 1960-1985 edited by V. Karageorghis, 269-76. Nicosia: A.G.Leventis Foundation. Christou, D. 1994. “Chronique des Fouilles et Découvertes archéologiques à Chypre en 1993.” BCH 118: 799-838. Christou, D. 1997. “Chronique des Fouilles et Découvertes archéologiques à Chypre en 1996.” BCH 121,2: 891-993.

Deal, M. and Hagstrom, M. 1995. “Ceramic Reuse Behavior Among the Maya and the Wanka: Implications for Archaeology.” In Expanding Archaeology, edited by J. Skibo, W. Walker, and A. Nielson, 111-25. Salt Lake City: University of Utah Press. De Caro, S. 1994. La villa rustica in località Villa Regina a Boscoreale. Pubblicazioni Scientifiche del Centro di Studi della Magna Grecia dell’Università degli Studi di Napoli Federico II. Terza serie, vol. 1. Rome: Bretschneider.

142

Deetz, J. 1977. In small things forgotten: the archaeology of early American life. New York: Anchor Press/Doubleday. Dever, W.G., Lance, H.D. and Bullard, R.G. 1978. A Manual of Field Excavation. Cincinnati: Hebrew Union College Dever, W.G., Lance, H.D. and Wright, G.E. 1970. Gezer I: Preliminary Report of the 1964-66 Seasons. Annual of the Hebrew Union College Biblical and Archaeological School in Jerusalem, Vol. 1. Jerusalem: Hebrew Union College Biblical and Archaeological School. Dias, L.F., Soren, D. and Vaz Pinto, I. 1988. “A Guide to Oil Lamps Used in Kourion in 365 A.D.” RDAC. 17984.

Eaton, S. and Robertson, A. 1993. “The Miocene Pakhna Formation, Southern Cyprus, and its relationship to the Neogene tectonic evolution of the Eastern Mediterranean.” Sedimentary Geology 86: 3-4. 27396. Ellis, S. 2000. Roman Housing. London: Gerald Duckworth & Co. Ltd. Ellis, S. 1998. “The End of the Roman House.” AJA 92: 565-76. Fales, D.C. Jr. 1950. “Kourion: The Amusement Area.” UPMB 14.4: 27-35. Fejfer, J. (ed.). 1995. Ancient Akamas, Vol. 1. Settlement and Environment. Aarhus: Aarhus University Press.

Diederichs, C. 1980. Salamine de Chypre IX: Céramiques Héllenistiques, Romaines et Byzantines. Paris: de Boccard.

Fejfer, J. and Mathiesen, H.E. 1995. “The Site of Ayios Kononas” in Ancient Akamas, Vol. 1. Settlement and Environment edited by J. Fejfer, 73-86. Aarhus: Aarhus University Press.

Dikaios, P. 1961-62. “Archaeology in Cyprus, 1959-61.” AR 8, 32-46.

Fleming, E.M. 1981. “Artifact study: a proposed model,” in Schlereth 1981: 162-73.

Du Plat Taylor, J. 1958. “Roman Tombs at “Kambi,” Vasa.” RDAC, 1940-1948. 10-76.

Fokaefs, A. & Papadopoulos, G. 2007. “Tsunami hazard in the Eastern Mediterranean: strong earthquakes and tsunamis in Cyprus and the Levantine Sea.” Natural Hazards 40:3, 503-26.

Dupré Raventόs, X. and Remolà, J.A. (eds.), 2000. Sordes Urbis: La eliminación de residuos en la ciudad romana. Boblioteca Italica 24. Rome: Bretschneider. Dyson, S.L. 1989. “The Role of Ideology and Institutions in Shaping Classical Archaeology in the 19th and 20th Centuries.” In Tracing Archaeology’s Past: The Historiography of Archaeology, edited by A.L. Christenson. 127-35. Carbondale: Southern Illinois University Press. Dyson, S.L. 1993. “From New to New Age Archaeology: Archaeological Theory and Classical Archaeology – A 1990s Perspective.: AJA 97:2. 195-206. Dyson, S.L. 1995. “Is There a Text in this Site?” In Methods in the Mediterranean: Historical and Archaeological Views on Texts and Archaeology edited by D.B. Small, 25-44. Mnemosyne; Supplementa, 135. Leiden; New York: E.J. Brill. Dyson, S.L. 1998. Ancient Marbles to American Shores: Classical Archaeology in the United States. Philadelphia: University of Pennsylvania Press. Dyson, S.L. 1999. “Brahmins and Bureaucrats: Some Reflections on the History of American Classical Archaeology.” In Assembling the Past: Studies in the Professionalization of Archaeology edited by A.B. Kehoe and M.B. Emmerichs, 103-16. Albuquerque: University of New Mexico Press.

Foxhall, L., and Forbes, H.A. 1982. “Σιτομετρεία: The Role of Grain as a Staple Food in Classical Antiquity.” Chiron 12, 41-90. Gallant, T.W. 1991. Risk and Survival in Ancient Greece: Reconstructing the Rural Domestic Economy. Cambridge: Polity Press. Garnsey, P. 1988. Famine and Food Supply in the Graeco-Roman World: Responses to Risk and Crisis. Cambridge: Cambridge University Press. Garnsey, P. 1999. Food and Society in Classical Antiquity: Essays in Social and Economic History. Cambridge: Cambridge University Press. George, M. 1997a. “Repopulating the Roman House.” In: The Roman Family in Italy: Status, Sentiment and Space, edited by B. Rawson and P. Weaver, 299-319. Oxford: Clarendon Press. Grahame, M. 2000. Reading Space. Social Interaction and Identity in the Houses of Roman Pompeii. Oxford: Oxford University Press. Green, H.J.M. 1961a. “An Analysis of Archaeological Rubbish Deposits.” Archaeological News Letter 7: 5154.

143

Green, H.J.M. 1961b. “An Analysis of Archaeological Rubbish Deposits: Part Two.” Archaeological News Letter 7: 91-93.

Hoffman, M.A. 1974. “The Social Context of Trash Disposal in an Early Dynastic Egyptian Town.” AmerAnt 39: 35-50.

Greene, J. 1987. “Methods of Excavation Recording.” In Soren, 1987a, 3-23.

and

Hurcombe, L.M. 2007. Archaeological Artifacts as Material Culture. New York: Routledge.

Guidobaldi, F., Pavolini, C., and Pergola, P. (eds.) 1998. I materiali residui nello scavo archeologico, CÉFR 249. Rome: École Française de Rome.

Ionas, I. 1988. La Maison Rurale de Chypre (XVIIe-XXe siècle). Publications du Centre de recherché scientifique de Chypre. Nicosia: Centre de recherché scientifique de Chypre.

Guidoboni, E., Comastri, A. and Traina, G. 1994. Catalogue of Ancient Earthquakes in the Mediterranean Area up to the 10th Century (1994). A Supplemented and Corrected Translation of the Catalogue in I terremoti prima del mille in Italia e nell' area mediterranea edited by E. Guidoboni. Translated by B. Phillips. Rome: Instituto nazionale di geofisica. Gunnis, R. 1936. Historic Cyprus: A Guide to its Towns and Villages, Monasteries and Castles. Methuen & Co. Ltd.: London. Halstead, P., Hodder, I., and Jones, J. 1978. “Behavioral Archaeology and Refuse: A Case Study.” Norwegian Archaeological Review 11: 118-31.

Jameson, M. 1990. “Domestic Space in the Greek City State.” In Domestic Architecture and the Use of Space: An Interdisciplinary Cross-Cultural Study edited by S. Kent. 92-113. New Directions in Archaeology. Cambridge: Cambridge University Press. Jensen, R. 1985. “The Kourion Earthquake: Some Possible Literary Evidence.” RDAC. 307-11. Kamp, K. 1991. “Waste Disposal in a Syrian Village.” In The Ethnoarchaeology of Refuse Disposal, edited by E. Staski and L. Sutro, L., 23-31. Arizona State University Anthropological Research Papers 42. Tempe: Arizona State University. Karageorghis, V. 1962. “Chroniques des fouilles à Chypre en 1961.” BCH 86. 327-414.

Harris, A. 2004. “Shops, retailing and the local economy in the early Byzantine world. The example of Sardis.” In Secular Buildings and the Archaeology of Everyday Life in the Byzantine Empire, edited by K. Dark. 82122. Oxford: Oxbow.

Karageorghis, V. 1963. “Chroniques des fouilles à Chypre en 1962.” BCH 87. 325-87.

Hayden, B. and Cannon, A. 1983. “Where the Garbage Goes: Refuse Disposal in the Maya Highlands.” JAnthArch 2: 117-63.

Karageorghis, V. 1964. “Chroniques des fouilles et découvertes archéologiques à Chypre en 1963.” BCH 88. 289-379.

Hayes, J.W. 1975. Roman and pre-Roman Glass in the Royal Ontario Museum: A Catalogue. Toronto: The Royal Ontario Museum.

Karageorghis, V. 1966. “Chroniques des fouilles et découvertes archéologiques à Chypre en 1965.” BCH 90. 297-389.

Hayes, J.W. 2008. The Athenian Agora Vol. XXXII: Roman Pottery Fine Ware Imports. Princeton: The American School of Classical Studies at Athens.

Karageorghis, V. 1970. “Chroniques des fouilles et découvertes archéologiques à Chypre en 1969.” BCH 94a. 191-300.

Herscher, E. 1995. “Archaeology in Cyprus.” AJA 99:2, 257-94.

Karageorghis, V. 1977. “Chroniques des fouilles et découvertes archéologiques à Chypre en 1976.” BCH 101. 707-99.

Hill, G. 1972. A History of Cyprus I: To the Conquest by Richard Lion Heart. Cambridge. Hillier, B. and Hanson, J. 1984. The Social Logic of Space. Cambridge: Cambridge University Press. Hodder, I. 1986. Reading the Past: Current Approaches to Interpretation in Archaeology. Second edition. Cambridge: Cambridge University Press.

Kelletat, D. 1998. “Geologische Belege katastrophaler Erdkrustenbewegungen 365 AD im Raum von Kreta.” In Naturkatastrophen in der antiken Welt: Stuttgarter Kolloquium zur historischen Geographie des Altertums 6, 1996 edited by E. Olhausen and H. Sonnabend. Franz Steiner Verlag GMBH: Stuttgart. 156-61. Kelly, G. 2004. “Ammianus and the Great Tsunami.” JRS 94, 141-67.

144

Kent, S. 1984. Analyzing Activity Areas: An Ethnoarchaeological Study of the Use of Space. Albuquerque: University of New Mexico Press. Kent, S. 1987. Method and Theory for Activity Area Research: An Ethnoarchaeological Approach. New York: Columbia University Press. Kent, S. 1999. “The Archaeological Visibility of Storage: Delineating Storage from Trash Areas.” AmAnt 64-1: 79-94. Keay, S. J. 1984. Late Roman amphorae in the Western Mediterranean. A typology and economic study: the Catalan evidence. BAR Int. 196. Oxford. Kingery, W.D. 1996. Learning from Things: Method and Theory of Material Culture Studies. Smithsonian Institute Press: Washington D.C. Last, J.S. 1975. “Kourion: The Ancient Water Supply.” PAPS 119.1: 39-72. Lavan, L., Swift, E., Putzeys, T. (eds). 2008a. Objects in Context, Objects in Use: Material Spatiality in Late Antiquity. Late Antique Archaeology Vol. 5. Leiden: Brill. Lavan, L., Swift, E., Putzeys, T. (eds). 2008b. “Material Spatiality in Late Antiquity: Sources, Approaches, and Field Methods” In Objects in Context, Objects in Use: Material Spatiality in Late Antiquity edited by L. Lavan, E. Swift, and T. Putzeys, 1-42. Late Antique Archaeology Vol. 5. Leiden: Brill. Leonard Jr., A. 1987. “A Brief Survey.” In Soren, 1987a, 80-116. Leonard, J.R. 2005. “Roman Cyprus: Harbors, Hinterlands, and ‘Hidden Powers’.” Ph.D. Dissertation: University at Buffalo, SUNY. Libeschuetz, W. 2000. “Rubbish Disposal in Greek and Roman Cities.” In: Dupré Raventόs and Remolà 2000: 51-61. Lichocka, B. and Meyza, H. 2001. “Seismic Events and the Evidence of Coins and Pottery. The Case of Destruction of the House of Aion in Paphos.” ÉtTrav 19. 145-208. Lister, F. and Lister, R. 1981. “The Recycled Pots and Potsherds of Spain.” Historical Archaeology 15: 6678. Loulloupis, M.C. 1971. “ANAΣΚΑΦΑΙ ΕΙΣ ΚΟΥΡΙΟΝ, 1967-70.” RDAC. 86-116. Luckenbill, D.D. 1927. Ancient Records of Assyria and Babylonia Volume II: From Sargon to the End. New York: Greenwood Press.

Maiuri, A. 2002. L’ultima fase edilizia di Pompei. (Reprint) Naples: Arte Tipographica. Manning, W.H. 1985. Catalogue of the Romano-British iron tools, fittings and weapons in the British Museum. London: British Museum Publications. Marangou, A.G. 2000. The Consul Luigi Palma di Cesnola 1832-1904: Life and Deeds. The Cultural Center of the Popular Bank Group: Nicosia. Masson, O. 1990. “Un vieux problème: Alasia = Chypre?” RÉG 103. 231-5. McFadden, E. 1971. The glitter and the gold; a spirited account of the Metropolitan Museum of Art’s first director, the spirited and high-handed Luigi Palma di Cesnola. New York: Dial Press. McFadden, G.H. 1935. “The Cyprus Expedition.” UPMB 5:4, 11-2. McFadden, G.H. 1938a. “Excavations at Kourion: The Sanctuary of Apollo.” UPMB 7:2, 10-17. McFadden, G.H. 1938b. “Excavations at Kourion: The Basilica.” UPMB 7:2, 3-4. McFadden, G.H. 1940. “The Sanctuary of Apollo.” UPMB 8:4: 22-8. McFadden, G.H. 1946. “A Tomb of the Necropolis of Ayios Hermoyennis (sic) at Kourion.” AJA 50. 449-89 McFadden, G.H. 1954. “A Late Cypriot III Tomb from Kourion.” AJA 58. 131-42. Megaw, A.H.S. 1951. “Archaeology in Cyprus, 19491950.” JHS 71, 258-60. Megaw, A.H.S. 1953. “Archaeology in Cyprus, 1952.” JHS 73, 133-37. Megaw, A.H.S. 1956. “Archaeology in Cyprus, 1956.” AR 3, 24-31. Megaw, A.H.S. 1976. “Excavations at the Episcopal Basilica of Kourion in Cyprus in 1974 and 1975: A Preliminary Report.” DOP 30: 345-71. Megaw, A.H.S. 1979. “The Atrium of the Episcopal Basilica at Kourion: A Preliminary Report.” RDAC, 358-65. Megaw, A.H.S. 1985. “Progress in Early Christian and Medieval Archaeology: The Episkopal Basilica at Kourion.” In Archaeology in Cyprus 1960-1985 edited by V. Karageorghis, 292-95. Nicosia: A.G. Leventis Foundation.

145

Megaw, A.H.S. 2007. Kourion: Excavations in the Episcopal Precinct. Dumbarton Oaks Studies, Vol. Washington, D.C.: Dumbarton Oaks XXXVIII. Research Library and Collection. Meltzer, D.J. 1981. “Ideology and Material Culture.” In Modern Material Culture Studies: The Archaeology of US, edited by R.A. Gould and M.B. Schiffer. 113-125. New York: Academic Press. Merrillees, R.S. 1987. Alashia Revisited. Paris: J. Gabalda. Meyers, E.M., Strange, J.F. and Meyers, C.L. 1981. Excavations at Ancient Meiron, Upper Galilee, Israel 1971-72, 1974-75, 1977. Meiron Excavation Project, Vol. III. Cambridge: The American Schools of Oriental Research. Meyza, H. 2007. Cypriot Red Slip Ware: Studies on a Late Roman Levantine fine ware. Nea Paphos V. Warsaw: Centre D’Archéologie Méditerranéenne De L’Académie Polonaise Des Sciences et Centre Kazimierz Michałowski D’Archéologie Méditerranéenne De L’Université De Varsovie. Mitford, T.B. 1966. “Three Milestones of Western Cyprus.” AJA 70.2: 89-99. Mitford, T.B. 1971. The Inscriptions of Kourion. Philadelphia: American Philosophical Society. Mitford, T.B. 1980. “Roman Cyprus.” In ANWR II.7.2, 1285-1384. Berlin and New York: Principat. Molinari, P., Leonard, J.R., and Soren, D. 1988. “University of Arizona Excavations at Kourion 1984 1987.” RDAC 2. 171-78. Moore, H. 1990. “Paul Ricoeur: Action, Meaning and Text.” In Reading Material Culture: Structuralism, Hermeneutics, and Post-Structuralism edited by C. Tilley, 85-120. Basil Blackwell, Inc.: Oxford. Nappo, S. 2004. Pompeii: A Guide to the Ancient City. New York: Barnes & Noble Books. Needham, S. and Spence, T. 1997. “Refuse and Formation of Middens.” Antiquity 71: 77-90. Nelson, B.A. 1985. “Reconstructing Ceramic Vessels in their Systemic Context.” In Decoding Prehistoric Ceramics, edited by B.A. Nelson. 310-29. Carbondale: Southern Illinois University Press. Nevett, L. 1994. “Separation or Seclusion? Towards an Archaeological Approach to Investigating Women in the Greek Household in the Fifth to Third Centuries BC.” In Architecture and Order: Approaches to Social Space edited by M.P. Pearson and C. Richards, 98-112. London: Routledge.2010.

Nevett, L. 1999. House and Society in the Ancient Greek World. New Studies in Archaeology. Cambridge: Cambridge University Press. Nevett, L. 2008. “Artefact Assemblages and Activity Area Analysis. A Comparison from Greek Domestic Contexts.” In Thinking About Space. The Potential of Surface Survey and Contextual Analysis in the Definition of Space in Roman Times, edited by H. Vanhaverbeke, J. Poblome, F. Vermeulen, M. Waelkens, and R. Brulet, 153-160. Leuven: Brepols. Nevett, L. 2010. Domestic Space in Classical Antiquity. Cambridge: Cambridge University Press. Nicolaou, K. 1973a. “Archaeological News from Cyprus, 1971.” AJA 77:1, 51-60. Nicolaou, K. 1973b. “Archaeological News from Cyprus, 1972.” AJA 77:4, 425-33. Nicolaou, K. 1975. “Archaeological News from Cyprus, 1973.” AJA 79:2, 125-34. Nicolaou, K. 1976. “Archaeological News from Cyprus, 1974.” AJA 80:4, 361-75. Nicolaou, K. 1978. “Archaeological News from Cyprus, 1976.” AJA 82:4, 521-35. Nur, A. and Burgess, D. 2008. Apocalypse: Earthquakes, Archaeology, and the Wrath of God. Princeton and Oxford: Princeton University Press. Oost, S.I. 1955. “Cato Uticensis and the Annexation of Cyprus.” CP 50:2. 98-112. Orton, C. 1989. “An Introduction to the Quantification of Assemblages of Pottery.” Journal of Roman Pottery Studies 2: 94-7. Orton, C. 1993. “How Many Pots Make Five? – An Historical Review of Pottery Quantification.” Archaeometry 35.2: 169-84. Orton, C. R. and Tyers, P. A. 1992. “Counting Broken Objects: The Statistics of Ceramic Assemblages.” ProcBritAc 77: 163-84. Orton, C., Tyres, P., and Vince, A. 1993. Pottery in Archaeology. Cambridge Manuals in Archaeology. Cambridge: Cambridge University Press. Oswald, D.B. 1987. “The Organization of Space in Residential Buildings: A Cross Cultural Perspective.” In: Method and Theory for Activity Area Research: An Ethnoarchaeological Approach, edited by S. Kent, 295-344. New York: Columbia University Press. Owens, E.J. 1983. “The koprologoi at Athens in the fifth and fourth centuries BC.” CQ 33:1, 44-50.

146

Parks, D.A. 1996. “Excavations at Kourion’s Amathus Gate Cemetery, 1995.” RDAC, 127-33. Parks, D.A. 1997. “Excavations at Kourion’s Amathus Gate Cemetery, 1996.” RDAC, 271-6.

Renfrew, C. 1980. “The Great Tradition versus the Great Divide: Archaeology as Anthropology?” AJA 84: 3. 287-98.

Parks, D.A., Given, M.J.M., and Chapman, N.E.M. 1998. “Excavations at Kourion’s Amathus Gate Cemetery 1997.” RDAC, 171-85.

Robinson, O. 1993. Ancient Rome: city planning and administration. London and New York: Routledge. Ross, L. 1852. Reisen nach Kos, Halikarnassos, Rhodos und der Insel Cypern (Inselreisen IV). C.A. Schwetschke & Sohn: Halle.

Parks, D.A. and Chapman, N.E.M. 1999. “Preliminary Report of the 1998 Excavations at Kourion’s Amathous Gate Cemetery.” RDAC, 259-67.

Rupp, D.W. 1978. “A Hellenistic Black and White Pebble Mosaic from the Acropolis of Kourion.” RDAC, 25465.

Parks, D.A., Mavromatis, C.M., and Harper, N.K. 2000. “Preliminary Report of the 1999 Excavations at Kourion’s Amathous Gate Cemetery.” RDAC, 305-16.

Russell, J. 1982. “Byzantine Instrumenta Domestica from Anemurium: The Significance of Context.” In City, Town and Countryside in the Early Byzantine Era edited by R.L. Hohlfelder, 133-64. East European Monographs. No. CXX. Byzantine Series, No. 1. New York, Columbia University Press.

Peña, J.T. 2007. Roman Pottery in the Archaeological Record. New York: Cambridge University Press. Peristianes, H.K. 1910. Γενική Ιστορία της νήσου Κύπρου. Lefkosia. Preziosi, D. 1979. The Semiotics of the Built Environment: An Introduction to Architectonic Analysis. Bloomington: Indiana University Press. Prown, J.D. 1982. “Mind in Matter: An Introduction to Material Culture Theory and Method.” Winterhur Portfolio 17:1. 1-19. Putzeys, T., Waelkens, M., Poblome, J., Van Neer, W., De Cupere, B., Van Thuyne, T., Kellens, N. and Bes, P. 2008. “Shops and Retail in Late Antiquity – A Contextual Approach to the Material Evidence from Salagassos.” In Thinking About Space. The Potential of Surface Survey and Contextual Analysis in the Definition of Space in Roman Times, edited by H. Vanhaverbeke, J. Poblome, F. Vermeulen, M. Waelkens, and R. Brulet, 161-217. Leuven: Brepols. Rautman, M.L. 2003. A Cypriot Village of Late Antiquity: Kalavasos-Kopetra in the Vasilikos Valley. JRA Supp. 52. Rautman, M.L. 2004. “Valley and Village in Late Roman Cyprus.” In Recent Research on the Late Antique Countryside, edited by W. Bowden, L. Lavan, and C. Machado, 189-218. Leiden-Boston: Brill. Renfrew, C. 2004. “Towards a Theory of Material Engagement.” In Rethinking Materiality: The Engagement of Mind with the Material World, edited by E. DeMarrais, C. Gosden, and C. Renfrew, 23-31. MacDonald Institute Monographs. Cambridge: MacDonald Institute of Archaeological Research.

Schiffer, M.B. 1996. Reprint. Formation Processes of the Archaeological Record. Salt Lake City: University of Utah Press. Original edition, Albuquerque: University of New Mexico Press, 1987. Schlereth, T.J. 1982. “Material Culture Studies in America: 1876-1976.” In Material Culture Studies in America edited by T.J. Schlereth, 1-79. Nashville: The American Association for State and Local History. Schlereth, T.J. 1985. “Material Culture and Cultural Research.” In Material Culture a Research Guide edited by T.J. Schlereth, 1-34. Lawrence: University of Kansas Press. Scobie, A. 1986. “Slums, Sanitation, and Mortality in the Roman World.” Klio 68: 399-433. Scranton, R. 1967. “The Architecture of the Sanctuary of Apollo Hylates at Kourion.” TAPS (New Series) 57:5, 3-85. Sinos, S. 1990. The Temple of Apollo Hylates at Kourion and the Restoration of Its South-West Corner. A.G. Leventis Foundation: Athens. Slane, K. 2004. “Amphoras -used and reused- at Corinth.” Transport amphorae and trade in the eastern Mediterranean. Acts of the International Colloquium at the Danish Institute at Athens, September 26-29, 2002 edited by J. Eirin and J. Lund. Monographs of the Danish Institute at Athens 5. Aarhus: Aarhus University Press. 361-69. Small, A., B. Roe, J. Hayes, C. Simpson, G. Guzzetta, M. MacKinnon, and S. Monckton. 1994. “A pit group of c. 80-70 B.C. from Gravina di Puglia,” BSR 62: 197260.

147

Smith, M.F. 1983. The study of ceramic function from artifact size and shape. Ph.D. dissertation (Eugene: University of Oregon).

Structuralism, Hermeneutics, and Post-Structuralism edited by C. Tilley, 3-84. Basil Blackwell, Inc.: Oxford.

Soren, D. (ed.). 1987a. The Sanctuary of Apollo Hylates at Kourion, Cyprus. University of Arizona Press: Tucson.

Toombs, L.E. n.d. Pella Field Manual (Privately Circulated).

Soren, D. (ed.). 1987b. “The Excavation and Its Significance.” In Soren 1987a. 24-52. Soren, D. (ed.). 1988. “The Day the World Ended at Kourion: Reconstructing an Ancient Earthquake.” NatGeogRes 174: 30-53. Soren, D., Gardiner, R., and Davis, T. 1986. “The Archaeological Excavations and Analyses of Earthquake-Affected Strata at Roman Kourion, Cyprus.” RDAC. 202-11. Soren, D. and James, J. 1988. Kourion: The Search for a Lost Roman City. New York: Doubleday. Stampolidis, N.C. 2004. Eleutherna: Polis, Acropolis, Necropolis. Athens : Museum of Cycladic Art. Stillwell, R. 1961. “Kourion: The Theater.” PAPS 105: 1. 37-78. Stiros, S. 2001. “The AD 365 Crete Earthquake and Possible Seismic Clustering During the Fourth to Sixth Centuries AD in the Eastern Mediterranean: A Review of Historical and Archaeological Data”. Journal of Structural Geology 23: 545-62. Sullivan, A. 1989. “The Technology of Ceramic Reuse: Formation Processes and Archaeological Evidence.” WorldArch 21: 101-14. Swiny, H.W. (ed.) 1982. An Archaeological Guide to the Ancient Kourion Area and the Akrotiri Peninsula. Nicosia: Department of Antiquities, Cyprus. Swiny, S. and Mavromatis, C.M. 2000. “Land Behind Kourion: Results of the 1997 Sotira Archaeological Project Survey.” RDAC 433-51. Thür, H. 2006. Antrag auf Förderung des Einzelprojekts “Wohneinheit 6 des Hanghauses 2 in Ephesos. Fonds zur Förderung der Fundkontexte.” wissenschaftlichen Forschung.

van der Leeuw, S.E. 1990. “Archaeology, Material Culture and Innovation.” SubStance 19:62/63, 92-109. Vessberg, O. 1953. “Hellenistic and Roman Lamps in Cyprus.” OpAth I-II. Lund. 115-129. Vessberg, O and Westholm, A. 1956. The Hellenistic and Roman periods in Cyprus. The Swedish Cyprus Expedition Vol. IV, Pt. 3. Stockholm: The Swedish Cyprus Expedition. Vidua di Conzano, C. 1826. Inscriptiones antiquae a comite Carolo Vidua in Turcico itinere collectae. Lutetiae Parisiorum: Dondey-Dupré. Vince, A.G. 1977. “Some Aspects of Pottery Quantification.” Medieval Ceramics 1: 63-74. Walmsley, A. 2007. “Households at Pella, Jordan: Domestic Destruction Deposits of the Mid-8th c.” In Objects in Context, Objects in Use: Material Spatiality in Late Antiquity edited by L. Lavan, E. Swift, and T. Putzeys, 239-72. Late Antique Archaeology Vol. 5. Leiden: Brill. Watkin, H.J. 1988. “The Development of Cities in Cyprus from the Archaic to the Roman Period.” Ph.D. Dissertation: Columbia University. Weber, E. 1989. “Zur Datierung der Tabula Peutingeriana.” In Labor omnibus unus. Gerold Walser zum 70. Geburtstag dargebracht von Freunden, Kollegen und Schülern. Stuttgart. 113- 17. Wild, J.P. 1987. “The Roman Horizontal Loom.” AJA 91:3, 459-71. Williams, D.F. 1987. “Roman Amphorae from Kourion, Cyprus.” RDAC. 235-8. Wright, G.R.H. 1992. Leiden: Brill.

Ancient Building in Cyprus.

Zemer, A. 1978. Storage Jars in Ancient Sea Trade. Haifa: National Maritime Museum Foundation.

Tilley, C. 1990. “Claude Lévi-Strauss: Structuralism and Beyond.” In Reading Material Culture:

148