JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES T H E P E N N S Y LVA N I A S T A T E U N I V E R S I T Y P R E S S JEMAHS VOL. 11 NO. 1 2023 JEMAHS EDITORS BOOK REVIEW EDITORS Ann E. Killebrew, !e Pennsylvania State University, University Park (USA) Sandra A. Scham, !e Catholic University of America (USA) Grace Erny, Stanford University (USA) Melanie Godsey, !e University of North Carolina at Chapel Hill (USA) A S S I S TA N T E D I T O R S E D I T O R I A L A S S I S TA N T Ina Berg, University of Manchester (UK) Hanan Charaf, Lebanese University (Lebanon) Gabriele Faßbeck, Independent Scholar (USA) EDITORIAL AND ADVISORY BOARD Salam Al-Kuntar, University of Pennsylvania (USA) Lorenzo d’Alfonso, New York University (USA) Jere L. Bacharach, University of Washington (USA) Reinhard Bernbeck, Freie Universität Berlin (Germany) Scott Bucking, DePaul University (USA) Eric H. Cline, !e George Washington University (USA) Anastasia Dakouri-Hild, University of Virginia (USA) Stella Demesticha, University of Cyprus (Cyprus) Elif Denel, American Research Institute in Turkey, Ankara (Turkey) Müge Durusu Tanrıöver, Bilkent University (Turkey) John D. M. Green, ACOR American Center of Oriental Research (Amman, Jordan) Joseph A. Greene, Harvard University (USA) Matthew Harpster, Koç University (Turkey) Rabei G. Khamisy, University of Haifa (Israel) Mark Leone, University of Maryland, College Park (USA) !omas E. Levy, University of California, San Diego (USA) Alexander Nagel, State University of New York, FIT (USA) Shelley-Anne Peleg, Independent Scholar (Israel) Susan Pollock, Freie Universität Berlin (Germany) On the cover: Rock art on the top of a volcanic mountain at the al Ha'it 4 site. (Photo by A. Nassr.) Rubina Raja, Aarhus University and Centre for Urban Network Evolutions (Denmark) Benjamin A. Saidel, East Carolina University (USA) Joshua Samuels, !e Catholic University of America (USA) Neil A. Silberman, University of Massachusetts Amherst (USA) Sharon R. Steadman, SUNY Cortland (USA) Margreet Steiner, Independent Scholar (!e Netherlands) James M. Weinstein, Cornell University (USA) Donald Whitcomb, !e University of Chicago (USA) Naama Yahalom-Mack, !e Hebrew University of Jerusalem (Israel) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES VOL. 11 v NO. 1 2023 From the Editors Sandra A. Scham, Ann E. Killebrew, and Gabriele Faßbeck ARTICLES ! Pathways to Know and Sidetracks to Forget: Walking and the Montane Cultural Landscape of Zagori (Northwestern Greece) Faidon Moudopoulos-Athanasiou "! #e Potential of Quantified Surface Data in Understanding the Rural Landscapes of Middle Byzantine Komana Mustafa Nuri Tatbul and Deniz Burcu Erciyas $% Monumental Stone Structure Archaeology in Northwest Saudi Arabia: New Investigations from the Al Ha’it Oasis Ahmed Nassr, Ahmed Elhassan, Ali Tueaiman, and Mohammed al-Hajj &" Funerary Artifacts from Roman Ashkelon and the Rockefeller Collections Tali Erickson-Gini and Alegre Savariego !'% Intrigue and Feud in Colonial Cyprus: Professor Talbot Rice’s Tendentious Report (!()*) on the New Antiquities Department Nicholas Stanley-Price !!( A Pot Bellows Fragment from Tel Megadim, Israel Samuel R. Wol! BOOK REVIEWS !"% Flint Procurement and Exploitation Strategies in the Late Lower Paleolithic Levant: A View from Acheulo-Yabrudian Qesem Cave (Israel), by Aviad Agam Reviewed by Christophe Delage !"* Iron Age Terracotta Figurines from the Southern Levant in Context, edited by E. D. Darby and I. J. de Hulster Reviewed by Sabine Fourrier !"& Untangling Blackness in Greek Antiquity, by Sarah F. Derbew Reviewed by Christopher Stedman Parmenter !)" Digging Up Jericho: Past, Present and Future, edited by Rachael #yrza Sparks, Bill Finlayson, Bart Wagemakers, and Josef Mario Bri+a Reviewed by Maura Sala , - .                                         (     ) is a peer-reviewed journal published by the Pennsylvania State University Press. JEMAHS is devoted to traditional, anthropological, social, and applied archaeologies of the eastern Mediterranean, encompassing both prehistoric and historic periods. "e journal’s geographic range spans three continents and brings together, as no academic periodical has done before, the archaeologies of Greece and the Aegean, Anatolia, the Levant, Cyprus, Egypt, and North Africa. As the journal will not be identi#ed with any particular archaeological discipline, the editors invite articles from all varieties of professionals who work on the past cultures of the modern countries bordering the eastern Mediterranean Sea. Similarly, a broad range of topics will be covered including, but by no means limited to: Excavation and survey #eld results; Landscape archaeology and GIS; Underwater archaeology; Archaeological sciences and archaeometry; Material culture studies; Ethnoarchaeology; Social archaeology; Conservation and heritage studies; Cultural heritage management; Sustainable tourism development; and New technologies/virtual reality. Appearing four times a year in February, May, August, and November, the journal will engage professionals and scholars of archaeology and heritage studies as well as non-practitioners and students, both graduate and undergraduate. In addition to combining traditional and theoretical archaeological data and interpretation, the journal’s articles may range from early prehistory to recent historical time periods. It also aims to publish accessible, jargon-free, readable, color-illustrated articles that will be informative for professional and non-professional readers. "e journal does not publish unprovenanced artifacts purchased on the antiquities market or objects from private collections. / 01 23//345 35 6 472 8, 345 Digital submissions should be sent to: www.editorialmanager.com/ JEMAHS. All correspondence should be sent to: Dr. Ann E. Killebrew ([email protected]). By submitting their work to JEMAHS, authors agree to editorial modi#cations of their manuscripts that are designed to help JEMAHS ful#ll its mission. Articles should be submitted as a MS Word #le together with all illustrations ($%&& dpi for black and white; '&& dpi for grayscale; and at least (&& dpi for color) referenced in the manuscript. Permissions to use photographs and copyrights for all illustrations are the responsibility of the authors and need to be included when the manuscript is submitted. (For more information regarding copyright issues for authors, go to: http://psupress.org/author/ author_copyright.html). Papers should be limited to not more than %&–%) manuscript pages or ca. ',&&&–*,&&& words. Shorter papers are welcome, but authors wishing to submit a paper longer than %) manuscript pages (including endnotes, references, and appendices) should consult with the editors in advance. For complete author submission guidelines, please visit: http://www.psupress.org/journals/jnls_JEMAHS.html / 01 / 973 : , 345 3 5 6 472 8, 345 "e Journal of Eastern Mediterranean Archaeology and Heritage Studies is published quarterly by the Pennsylvania State University Press, +%& N. University Dr., USB $, Suite C, University Park, PA $'+&%. Subscriptions, claims, and changes of address should be directed to our subscription agent, the Johns Hopkins University Press, P.O. Box $,,'', Baltimore, MD %$%$$, phone $-+&&-)-+-$*+- (outside USA and Canada: -$&-)$'',+*), [email protected]. Subscribers are requested to notify the Johns Hopkins University Press and their local postmaster immediately of change of address. All correspondence of a business nature, including permissions and advertising, should be addressed to the Pennsylvania State University Press, [email protected]. "e Pennsylvania State University Press is a member of the Association of University Presses. 73;- , / 85< :. 72 3/ / 345 JEMAHS is registered under its ISSN (%$''-()-+ [E-ISSN %$''-())']) with the Copyright Clearance Center, %%% Rosewood Drive, Danvers, MA &$,%( (www.copyright.com). For information about reprints or multiple copying for classroom use, contact the CCC’s Academic Permissions Service, or write to the Pennsylvania State University Press, +%& N. University Dr., USB $, Suite C, University Park, PA $'+&%. Copyright © %&%( by "e Pennsylvania State University. All rights reserved. No copies may be made without the written permission of the publisher. FROM THE EDITORS Sandra A. Scham, Ann E. Killebrew, and Gabriele Faßbeck Welcome to our 'rst issue of !"!#. Our readers will 'nd that the articles that follow are quite diverse in terms of the methods, theories, and regions they describe. $ey do, however, have one signi'cant thing in common. As archaeologists have come to realize that excavation is not the only way to address archaeological questions, they are turning, much more, to less destructive means to 'nd out about the past. Conducting surface surveys, examining museum collections, making better use of remote sensing data, reanalyzing previously excavated materials, and archival research are no longer seen as “auxiliary” methods to excavation but, rather, as stand-alone approaches to understanding earlier cultures. “Pathways to Know and Sidetracks to Forget,” is a discussion by Faidon Moudopoulos-Athanasiou about the interweaving of nineteenth-century travelers’ accounts and systematic modern surveys. Using the region of Zagori in northwestern Greece as a case study, the author describes how both historical and modern re(ections on cultural landscapes that are informed by walking those very landscapes facilitate the formation of a detailed view of the peoples who inhabited them. Promoting the idea of an “inclusive landscape archaeology,” Moudopoulos-Athanasiou suggests that the study of historical walks combined with a modern archaeologist’s survey expertise can inform appropriate heritage management of a region, which will bene't tourists as much as residents. Mustafa Nuri Tatbul and Deniz Burcu Erciyas, authors of “$e Potential of Quanti'ed Surface Data        , . , . ,  Copyright © !"!# $e Pennsylvania State University, University Park, PA https://doi.org/%".&#!&/jeasmedarcherstu.%%.%.v in Understanding the Rural Landscapes of Byzantine Komana,” describe their surveys at this site in Turkey and the late ninth- through twelfth-century churches and settlements they discovered. Scholars have reasoned that this was a time when building small churches became a trend, as rural settlements and monastic communities grew. $e authors use a “mixed methods” approach to survey methodologies against the theory of rural revival at a regional level. Utilizing qualitative data from previous surveys for information and comparison, they revisited the identi'ed sites to test whether the identi'cations of these churches could be veri'ed by quantitative analyses. In a very interesting geographic departure from the countries that are the journal’s usual focus, this issue contains an article about Saudi Arabia. Undeniably, the cultures of Arabia and the Ancient Near East co-existed, communicated, and traded in the very distant past. Unfortunately, unlike the Near East, many sites in Saudi Arabia remain unstudied, so it has been di)cult to establish its true place in the region’s history. Concentrating on “Monumental Stone Structure Archaeology” in the Al Ha’it oasis of northwest Saudi Arabia, authors Ahmed Nassr, Ahmed Elhassan, Ali Tueaiman, and Mohammed al-Hajj show how an intensive desktop archaeological survey using remote sensing, followed by traditional archaeological survey and excavations reveals the region’s archaeological richness. $e authors describe how among the monuments observed were previously undiscovered structures in the al Ha’it oasis. Tali Erickson-Gini and Alegre Savariego discuss excavated material from the site of Er-Rasm, located ca. ! km southeast of Tel Ashkelon, which revealed the existence of a cemetery of the second to mid-fourth century vi ! FROM THE EDITORS CE there. In “Funerary Artifacts from Roman Ashkelon and the Rockefeller Museum Collections,” they explain how recent and earlier twentieth-century sarcophagus 'nds illuminate the history of Middle and Late Roman Ashkelon, documenting the city’s signi'cant wealth at the time. “Intrigue and Feud in Colonial Cyprus” are explored by Nicholas Stanley-Price, who discusses a critical report submitted in %*#+ by Professor Talbot Rice to the British Government on the state of Cyprus’s Department of Antiquities. $e author contends that the structure of colonial rule in Cyprus at the time was such that it enabled the Governor of Cyprus, Sir Richard Palmer, to decide on matters concerning which he obviously had little knowledge. $e disposition and care of antiquities were de'nitely in that category. $e problems raised by Talbot Rice, the author further contends, were indicative of many things related to colonial rule—some of which are ongoing today, such as a continued reliance on the reports of outside experts with little knowledge of the country’s context. Stanley-Price shows in this article how archival research can shine a light on continuing problems in countries that are still su,ering from the systems created by colonialism. A short report by Samuel R. Wol,, which was written as a response to Naama Yahalom-Mack’s !"%- article entitled “Crucibles, Tuyères, and Bellows in a Longue Durée Perspective” (Journal of Eastern Mediterranean Archaeology and Heritage Studies, vol. -, issue %), brings our attention to a small 'nd with big implications. $e author explores a previously unpublished artifact from his %**. excavation in “A Pot Bellows Fragment from Tel Megadim, Israel” and reveals that this particular artifact signals a “turning point” between Middle Bronze Age stone pot bellows and ceramic pot bellows. $is issue concludes with four book reviews. $e 'rst, by Christophe Delage, is on Flint Procurement and Exploitation Strategies in the Late Lower Paleolithic Levant: A View from Acheulo-Yabrudian Qesem Cave (Israel), by Aviad Agam. Sabine Fourrier reviews Iron Age Terracotta Figurines from the Southern Levant in Context, edited by E. D. Darby and I. J. de Hulster. Christopher Stedman Parmenter reviews Untangling Blackness in Greek Antiquity, by Sarah F. Derbew, and Maura Sala reviews Digging Up Jericho: Past, Present and Future, edited by Rachael $yrza Sparks, Bill Finlayson, Bart Wagemakers, and Josef Mario Bri,a. THE POTENTIAL OF QUANTIFIED SURFACE DATA IN UNDERSTANDING THE RURAL LANDSCAPES OF MIDDLE BYZANTINE KOMANA       $e Middle Byzantine period in Anatolia is considered a recovery period after two centuries of instability at the end of the Early Byzantine period. From the late ninth through the twelfth centuries, building small churches became a trend as rural settlements and monastic communities expanded across the countryside. During extensive surveys at Komana, near Tokat in Turkey, a number of churches were identi'ed in rural contexts through scattered architectural fragments, especially brickwork typical of the Middle Byzantine period. $ese sites were later revisited for intensive collection in an attempt, 'rstly, to test whether identi'cation of Middle Byzantine churches through qualitative data could be veri'ed and, secondly, to improve our understanding of site types and their extent in the rural landscapes of Middle Byzantine Komana. In this article, the results of the intensive surveys will be discussed in the context of the new trends of the Middle Byzantine period.   : church, intensive survey, spatial analysis, GIS, medieval, Anatolia        , . , . ,  Copyright © !"!# $e Pennsylvania State University, University Park, PA https://doi.org/%".&#!&/jeasmedarcherstu.%%.%.""!% Mustafa Nuri Tatbul Deniz Burcu Erciyas Archaeological and historical evidence indicates an increase in urban and rural life in Anatolia of the Middle Byzantine period. Especially in the eleventh century, an economic boom resulted in the emergence of new domestic and industrial units at sites, a revival of monasteries, and reconstruction of new churches at the site of old ones or their reuse in some places (Holmes !""(: !)!). Archaeological data supports this boom through increasing visibility in the number of rural settlements across the empire lands from the tenth to the twelfth century (Harvey !""(: ##!). $e number of urban and rural churches increased from the ninth through the twelfth centuries, especially in Greece, Cyprus, Crete, and Anatolia, and the Byzantine pottery types became more common than before (Whittow !""(: *)#). In support of the archaeological and historical observations, environmental data attained through palynological studies con'rmed favorable climatic conditions and an increase in agropastoral practices during this period (see Izdebski !"%!; Haldon et al. !"%*; Xoplaki et al. !"%+; Roberts et al. !"%(). Extensive surveys conducted between !""* and !"", at Komana, in the central Black Sea region, Turkey, similarly indicated an increase in the number of settlements during the Middle Byzantine period within the territory of this ancient sanctuary site (Erciyas and Sökmen !"%"b). It was the decrease in the ability to identify churches through surface assemblages that encouraged the idea 2 2 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA to test survey methodologies against the theory of rural revival at a regional level at Komana, as well as to seek patterns of continuity of sacredness at church sites from the Early through the Middle Byzantine period (sixth– ninth century and tenth–thirteenth century, respectively). A special attempt to understand whether these highly visible churches were part of rural settlements (i.e., villages, hamlets, monasteries) was made. For these purposes, three distinct sites were revisited for intensive surface collection and geophysical survey where possible: Mezarlıktarla, Bula, and Pınarlı (Fig. %). Among the three church sites, Mezarlıktarla was the closest to the core of the Byzantine settlement and considered to have been in an urban context. It was also the closest to the Ye-ilırmak River (ancient Iris) and the modern highway, which overlaps with the ancient transportation route from Dokeia to Neocaesarea. Bula on the other hand, was in the periphery of Komana but within the immediate rural context. Pınarlı, which was a rural church site, was located at a considerable distance from Komana. Similar sites were recorded elsewhere in the larger territory of Komana (i.e., Ahmetalan, Hasanbaba, Kemalpa-a, and Ormandibi) during extensive surveys. $is article discusses survey methods applied at these three sites in order to better understand surface representation and subsurface anomalies. While variations in tile and ceramic densities are examined as indicators of site function, the potential of observation points for FIG. 1 Map of the study area and the territory of Komana. Inset shows the location of Komana at the inner part of the Black Sea, Turkey. (Courtesy of the Komana Archaeological Research Project Archive.) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 23 identi'cation of spatial patterning and the relationship between survey units are also evaluated. $e resolution of spatial data is utilized to interpret subsurface representation. Context of the Study Our study was conducted in the territory of Komana, which spreads across the fertile Ye-ilırmak valley between Tokat and Niksar and the terraces on the northern and southern hills. $is land was intensively exploited for agricultural production and animal husbandry throughout its history. Continuous habitation at the site is witnessed through archaeological data acquired during surveys as well as ongoing excavations at Hamamtepe, a mound site at a central position in the plain, since !"", (Erciyas !"%,) (Fig. !). Byzantine, Danishmend/Seljuk, and Ottoman layers provided a variety of archaeobotanical and zooarchaeological taxa, which indicated agropastoral activities supporting self-su.cient communities during these periods (Pi-kin !"%&; Pi-kin and Tatbul !"%&). $ese lands are still bountiful in terms of agricultural produce today. FIG. 2 Aerial picture of Hamamtepe mound, where the Komana excavations have been conducted since 2009. (Courtesy of the Komana Archaeological Research Project Archive.) 24 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA Archaeological evidence representing Middle Byzantine occupation at Komana is concurrent with the late ninth- and tenth-centuries economic recovery (Erciyas !"%,: %"–%+). $e cemetery with two chapels on Hamamtepe (Fig. #), habitation levels reached in a trench just below the mound indicating an expanding settlement over the plain, and several Middle Byzantine churches in the immediate vicinity may be considered testimony for a rural revival during the tenth through early twelfth centuries—a notion accepted for the period in general for Anatolia (see Niewöhner !"%): &+–&)). $is is further strengthened by the palynological data, with a slight decrease in agropastoral activities during the eleventh century (Izdebski !"%!; Haldon et al. !"%*; Xoplaki et al. !"%+; Roberts et al. !"%(). Komana falls comfortably into place with increasing evidence for a revival at the site and in its rural vicinity after a rather silent period between the late seventh to late ninth century. During the extensive surveys conducted within the territory of Komana between !""* and !""( (Erciyas !""+, !""); Erciyas and Sökmen !"",, !"%"a, !"%"b; Erciyas, Sökmen, and Kalaycı !""(), %( Byzantine sites were identi'ed based on qualitative data (Erciyas and Sökmen !"%"b). At eight sites, presence of Middle Byzantine churches was suggested through partly visible ruins of the structures and architectural remains on the surface, such as building blocks, tiles, bricks, terracotta rosettes, and mortar. Altitude of sites spreading across the rural territory ranged between &"* and %#)+ m.a.s.l. (Erciyas and Sökmen !"%"b). $e majority were found at remote locations with over ,"" m.a.s.l. (Erciyas and Sökmen !"%"b), while the kastron at Komana, on top of a mound (ca. +*" m.a.s.l.) in the middle of the fertile valley and near the bank of the Ye-ilırmak, dominated the landscape until the Middle Byzantine period. $ese sites were assessed in two groups according to elevation and size, and their chronological representations were identi'ed. $ey were also evaluated in accordance with the changing political and social dynamics between the Early and the Middle Byzantine periods (Erciyas and Sökmen !"%"b). As for Komana itself, on Hamamtepe, which must have been an abandoned kastron by this time, the inner part of the forti'ed settlement was transformed into a Middle Byzantine cemetery, where a sample from a skeletal remain (G!,) is dated to %"+( ±#& BP. Around %*" graves have been excavated associated with the two adjacent Middle Byzantine chapels. Anthropological study suggests a rural community with a limited diet (i.e., frequent tooth decays and cavities suggested a diet dominated by foods high in carbohydrates; Erdal, Erdal, and Koruyucu !"%&). $e rather compact settlement character at Komana during the Early Byzantine period may have changed (Tatbul and Erciyas !"%,) in the Middle Byzantine period. Communities spread across the /at and fertile valleys, yet at the same time inhabited eco-niches at remote high altitudes with suf'cient natural resources and suitable conditions for agropastoral practices, although di0erent than the valleys. Expansion to rural areas might also suggest that it had become safer during this period of relative political stability. $e climatic conditions were relatively favorable, while remote locations still provided natural protection isolated from main-road networks through the Byzantine period. Also, their isolated environments might have been suitable for monastic establishments, which /ourished especially by the end of Iconoclasm after the mid-ninth century (Brubaker and Haldon !"%%: &)"–)%). Starting from the second half of the twelfth century, Hamamtepe, the former forti'ed core of Komana, began to be resettled under the Danishmends followed by the Seljuks, a time for which intensive industrial production of glazed pottery, metal, and glass is attested, as well as a /ourishing bone-object industry and copious food-related activities (Erciyas !"%#, !"%,; Tatbul !"%); Tatbul !"!"; Erciyas and Tatbul !"%+, !"%(, !"%,; Erciyas et al. !"%&; Tatbul and Erciyas !"!"). $e amount and variety of commodities as well as zooarchaeological and archaeobotanical species from the twelfth to the fourteenth century indicate that Komana took on a pivotal role during the Seljuk period with a strong economy supported by its rural territory (Erciyas !"%,: &–%"; Tatbul !"!"). A recent thesis study on the ceramics of the Turkish layers of the settlement revealed that Komana had become an important FIG. 3 Architectural phases showing the tenth- to twelfth-century Middle Byzantine cemetery churches positioned at the center of Hamamtepe. (Courtesy of the Komana Archaeological Research Project Archive.) 2 6 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA ceramics producer and further contributed to pottery traditions in Anatolia as well as in Crimea (Karasu !"!"). Komana-style glazed ceramics are commonly found during surveys in the area at securely identi'ed Byzantine sites, suggesting continuity at small rural settlements in the Seljuk period. In this context, the Komana Project 'lls an important gap in scholarship, as it informs about the missing link between the Middle Byzantine sites and their fate after the spread of Seljuk domination. Patterns of continuity, modi'cation, or abandonment at the sites have the potential to understand this complex period in terms of social dynamics. $ere seems to be su.cient evidence for the matter at Komana, which is dealt with in another article by Tatbul (Tatbul !"!%). Komana continued as a nahiye (a small provincial administrative unit) during the Ottoman period as indicated in the Ottoman cadastral record books (tahrir) and court records of Tokat Province (!er’iye sicili) (Hanilce !"%#: *"–*%). Overall, the historical sources, the archaeological data, and the environmental studies suggest that there was a “rural revival” in Anatolia starting by the early tenth century. $e relatively higher numbers of Middle Byzantine sites in rural Komana documented during extensive surveys may represent re/ections of this revival in the valleys of the Ye-ilırmak. Understanding continuity at these sites from the Middle Byzantine through the Seljuk period could shed light on shifting settlement patterns in the transition period. A change in settlement patterning is evident in the archaeological and historical data obtained from Komana and its territory for this period. However, our research encountered a certain amount of di.culty in exactly identifying settlement types that accompanied the churches dotting the landscape. It has been suggested that churches in rural contexts, monasteries, and workshops were built during the Middle Byzantine period. By the end of Iconoclasm, widespread establishment of monastic organizations resulted in a boom in church constructions. $is is also when the new type of Middle Byzantine “cross in square” church model was introduced (see Ousterhout %,,,, !"%+; Akyürek %,,)). $ese new privatized and small-sized churches dominated the rural landscapes. During this period, investment was directed to rural churches in private estates instead of urban projects involving large public churches, which had been the norm for the Early Byzantine period (Niewöhner !"%): &+). In this context, identi'cation of rural sites accompanying the churches, whether they were modest villages, monasteries, or other sites, emerges as a signi'cant objective for this study, which may contribute to the discussion around changing socioeconomic and political circumstances. In order to achieve this goal, pilot studies were designed at three church sites, where a number of di0erent methods were applied. In the following section, these case studies will be discussed in detail. Survey Methodology $e aim of the survey was to create quantitative surface data from three previously visited sites (Mezarlıktarla, Bula, Pınarlı) and to test whether the identi'cation of Middle Byzantine churches through qualitative data could be veri'ed. Another purpose of intensive collecting was to improve our understanding of site types and their extent with the hopes of illuminating the question regarding rural settlement patterns during the Middle Byzantine period. $e analysis was also supported by geophysical prospection at two sites; Mezarlıktarla and Bula, where two di0erent techniques (ground penetrating radar [GPR] and magnetic resonance respectively) were tried. $e third site, Pınarlı, was only analyzed through quanti'ed surface data. Pınarlı proved especially signi'cant for understanding the position of the previously identi'ed church within its rural setting and the potential for identifying a rural settlement with domestic quarters. Units of the survey areas were de'ned by the modern 'eld boundaries, and they were given numbers when more than one 'eld was surveyed at each site, otherwise called with names (i.e., Mezarlıktarla and Bula). In the 'elds, the surface was divided into grids where a transect was set in the middle lines. $ree di0erent grid dimensions were used depending on JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 27 the surveyed 'elds. Accordingly, Mezarlıktarla was divided into & × ! m, Bula into & × & m, and Pınarlı into %" × %" m grids, respectively. Length of the transects varied depending on the 'eld size and each transect was walked and sampled in ! m width. $is was based on the ability of the walker to collect materials at % m range at each side without leaving the transect line. $us, a uniform sampling strategy and limit was set to all team members. Materials collected were left in observation points every & or %" m depending on the established grid dimensions. A similar method of quantifying observation points was used at the Avkat and Göksu surveys (Newhard et al. !"%#). Every artifact on the transect was collected, identi'ed, quanti'ed, photographed, and recorded on site. Data on the sheet was transferred to Excel tables and converted to CSV (comma-separated values) format and embedded into QGIS software as attribute tables. Observation points were de'ned as point coordinates. High-resolution geographically referenced aerial photographs were taken at every surveyed 'eld from )& m altitude by drone. Photographs were processed using Metashape Pro software and geographically referenced orthomosaic photos were prepared in geoti0 format to be used as canvas for the survey area in QGIS software. Surface material distributions were queried and analyzed in QGIS by interpolation (kernel density heatmap) method, and material densities were displayed in graduated circles in each sampling grid created on each transect line. At Pınarlı, two di0erent quanti'cation methods were used: tiles and ceramics were both counted and weighed to check whether the distribution results varied between sampled 'elds. To be more speci'c, size difference among the tile fragments gives di0erent values when counted and weighed. Case Study : Mezarlıktarla At a distance of &"" m, Mezarlıktarla is the church site closest to Hamamtepe. It is located on the /at valley bottom, ca. +"" m.a.s.l., immediately next to the modern Tokat–Niksar road. $e ancient road may have followed a similar route, but possibly passing by the church on the east. Based on the high density of architectural materials spotted during initial observations on the 'eld surface, the presence of a church was considered. In !"%*, geophysical prospection was conducted at the site using GPR. Systematic surface collection was also performed, establishing ! × & m grids—a system adapted from the ! m wide geophysical transects. At Mezarlıktarla, an area of !(+" m! was systematically surveyed, with !(+ grids sampled and GPR readings recorded at #( transects (Fig. *A–D). Due to the bushes located at the western part of the 'eld, an area of %( × %& m could not be sampled, and geophysical surveys could not be performed. A total number of &,*", tiles, !*, ceramics, #& terracotta rosettes, and %# opus sectile fragments were recorded. It was observed that the distribution of tiles was dense around the bushy spot where neither archaeological nor GPR survey could be conducted. $e qualitative character of the architectural materials as well as their surface distribution quantitatively supported the initial proposition of the presence of a Middle Byzantine church at Mezarlıktarla. $e distribution pattern clearly pointed toward one speci'c area (underneath the aforementioned bushes) for the location of the church (Fig. *A). In !"%(, when the owner of the 'eld cut the bushes in order to put the 'eld up for sale, some mortar and wall remains became visible on the surface. Immediately thereafter, an aerial photo was taken by drone, con'rming the presence of a three-apsed church at the spot where based on the survey data we had expected such 'nd (Fig. &). Since the church was covered with thorny bushes during the surface survey and geophysical prospection, the spot could not be included in the study. However, the rest of the 'eld was studied in its entirety. Geophysical prospection showed some very fragmentary anomalies. Another structure on the northwest corner of the 'eld and a number of possible sarcophagi were identi'ed. $e ceramic distributions map showed a cluster adjacent to this GPR anomaly. One of the tile clusters corresponded to the anomaly together with the ceramic density. $e fact that ceramics were found not around FIG. 4 Spatial data obtained in the Mezarlıktarla survey: (A) heatmap showing tile densities; (B) heatmap showing ceramic densities; (C) interpretation of GPR results indicating a structure within the blue circle; (D) pie chart showing the total ratio among the surface find types. (Courtesy of the Komana Archaeological Research Project Archive.) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 29 FIG. 5 Aerial picture of the Middle Byzantine church explored at Mezarlıktarla. The church ruin became visible after the cleaning of the bushes, which coincides with the unsampled spot (see Fig. 4A–B) during the surveys. (Courtesy of the Komana Archaeological Research Project Archive.) the now visible church but near another possible building indicated the presence of other secular activities. Lack of ceramics in the immediate surrounding of identi'able Middle Byzantine churches has been a pattern observed throughout the surveys in the area. Among the ceramics were terra sigillata, red slip wares, and glazed ceramics—an assemblage very familiar from the twelfthto fourteenth-century levels at Komana (Fig. +). $e ceramic sequence indicated that Mezarlıktarla was continuously occupied from the Roman to the Seljuk period. $e distribution of the terracotta rosettes and opus sectile fragments coincided with the expansion of the tiles, creating a uniformity in the distribution of architectural elements. While this site may have been part of a separate settlement, it may also have been a neighborhood of Komana itself, given its proximity to Hamamtepe. $e inhabitants of Gümenek village, whose houses spread across the area between Hamamtepe and Mezarlıktarla, often speak about pithoi, hollows in the ground, passages, and more, which can be taken as continuity of structures between the two locations. FIG. 6 Surface finds from the Mezarlıktarla survey: (A) terracotta rosettes; (B) overfired tile fragments; (C) marble opus sectile fragments; (D) glazed Komana wares; (E) coarse ware handles. (Courtesy of the Komana Archaeological Research Project Archive.) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 31 Case Study : Bula $e church site Bula, taking its name from the nearby village, is located %,)"" m north of Hamamtepe. It is situated on a /at terrace slightly above the valley bottom at ca. +&" m.a.s.l. During the extensive surveys in !""+, the surface materials as well as the big trench dug by looters (ca. +"" m! and ! m deep) suggested the presence of a Middle Byzantine church in the 'eld, where agricultural activity continues despite its status as a registered site. In the same survey season, geophysical prospection was conducted using magnetic resonance technique. $irty-one grids with !" × !" m dimensions were established and magnetometer readings were recorded every ".& m in % m wide transects. In !"%(, the team revisited the site of the church, and another +,*"" m! of the 'eld surface were systematically surveyed and !&+ grids of & × & m sampled. A total number of !+( tiles, (# ceramics, one terracotta rosette, and four opus sectile fragments were recorded within the survey area (Fig. )A–D). $e looters’ trench in the northern section of the surveyed area could not be sampled since the surface was not suitable for the application. However, mortared brick-construction walls were recorded. $e artifact distributions were analyzed and interpreted separately in two halves of the 'eld, imagining a dividing line on the east–west axis because the upper (northern) half of the 'eld was ca. % m higher than the southern part. According to the distribution analysis of the surface materials, a cluster of tiles was observed immediately to the southwest of the looters’ trench (located inside the striped polygon in Fig. )A). $e structure had been identi'ed as a Middle Byzantine church based on the tiles, terracotta rosettes, and opus sectile fragments observed during the extensive surveys. Furthermore, the quanti'ed data acquired during the intensive collecting in !"%( supported the identi'cation of the building and its location. $e natural slope toward the south and southwest contributed to the direction of the surface scatter, which was also observed on the magnetometer images. $e magnetic survey results suggested an anomaly immediately to the southwest of the church. It was interpreted as scattered building materials resulting from agricultural activity. $e 'eld on the slight slope (ca. &–%" degree angle) also favored the downward movement of materials. $e tile cluster as well corresponded to the subsurface anomaly interpreted in the geophysical prospection. $e geophysics and the surface clutter suggest that the tiles might have originated from the church ruins; they were typically large, rather thick, and /at tiles used in wall construction as observed in situ. However, the assemblage also included molded examples with grill-like decoration similar to examples from the chapels on Hamamtepe (Vorderstrasse !"%,: *,, 'g. !). Geophysical prospection also suggested that there were wall lines on the northern part of the church spot, which separates the 'eld into two halves (Fig. )C). $is actually explains why there is a terrace dividing the modern-day 'eld. Another wall line extended in north–south direction separating the church spot from the rectangular building. Additionally, two anomalies indicative of burnt spots were detected: one on the west of the northern half and the other on the east of the looters’ trench. An oblique photo taken in June !"%,, just before the harvest, showed that there were crop marks exactly at the same spots where geophysics indicated other structures and activity areas, which may have served di0erent functions in addition to the church. A pattern was also observed in the distribution of the surface ceramics. $ey formed a cluster in the northern half of the site. $e assemblage included glazed pieces, typical of the Komana group dating to the twelfth to fourteenth century, and coarse ware including lids and jar handles representing daily ware (Fig. (). $is assemblage is characteristic of the area around Komana as well as Hamamtepe itself (the core of the site) and did not necessarily represent the stratigraphic depth that the excavations demonstrated. $e wall between the higher and lower sections of the 'eld (north and south) may have encircled the church, separating it from the rest of the settlement. $is could be the reason for the lower number of ceramics on the south of the circumference wall where the church was located. Patterns of tile and ceramic clusters suggested that the site might have been a multifunctional settlement / a rural parish with a church at its center. Across the 'eld, eight burnt tiles were also observed among the !+( collected. $ese were over'red production wasters rather than tiles from structures which were exposed to 're. $is observation, in addition to the spots FIG. 7 Spatial data obtained in the Bula survey: (A) heatmap showing tile density; (B) heatmap showing ceramic density; (C) magnetometer results showing subsurface anomalies around the church spot; (D) pie chart showing the total ratio among the surface find types. (Courtesy of the Komana Archaeological Research Project Archive.) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 33 FIG. 8 Ceramic finds from the Bula survey. (Courtesy of the Komana Archaeological Research Project Archive.) with higher anomalies detected in the magnetic survey (especially a large round area on the northern section), suggested presence of furnaces, possibly temporarily used to produce tiles during the construction of the church and any other structures. Test excavations in the future have potential to clarify this assumption. Case Study : Pınarlı Pınarlı is a village located ).) km from Komana toward the east, on the modern Tokat–Almus road. $e survey area is about %.& km north/northeast of the village in the agricultural land between Pınarlı and Çöre1ibüyük. $e site is on the eastern edge of the alluvial plain, opening up like a fan from Komana toward the east and ending at mountains. In the north, Ye-ilırmak continues through a narrow pass toward Niksar, and in the south a second route climbs the mountains to reach Almus. $e site is also considerably far from Ye-ilırmak. $e signi'cant distance of the site to Komana, its isolated, almost hidden rural location and surface representation distinguish it from the sites of Mezarlıktarla and Bula as we will show. $e previously identi'ed church is located in the /at bottom of an isolated gorge at the bottom of a rocky formation named Çatalkaya. $e area including the church as well as four neighboring 'elds toward the south was systematically surveyed. Field % is where the Middle Byzantine church was detected; its location must have been intentionally chosen. During the extensive surveys in !""), the Middle Byzantine church site was identi'ed based on tiles, terracotta rosettes, and a large number of mortar fragments on the surface of the survey area, which were exposed as a result of destruction caused by heavy machinery in !""). In !"%,, 've 'elds were systematically surveyed, covering an area of %*,#)& m!. Grids of %" × %" m were set at all 've 'elds. A total number of )#& tiles and %)(* ceramics were recorded from ! m wide walking transects of the 've 'elds (Fig. ,A). Among the 've 'elds, the highest density of tile fragments was observed in Fields % and ! (Figs. ,–%"). FIG. 9 Charts showing surface data collected in the Pınarlı survey: (A) distribution of ceramic and tile numbers among fields; (B) distribution of ceramic and tile weights among fields; (C) distribution of material densities calculated count per m2; (D) distribution of material densities calculated weight per m2. (Courtesy of the Komana Archaeological Research Project Archive.) F I G . 10 Spatial data obtained in the Pınarlı survey: (A) heatmap showing tile counts among units and fields; (B) heatmap showing ceramic counts among units and fields; (C) heatmap showing tile weights among units and fields; (D) heatmap showing ceramic weights among units and fields. (Courtesy of the Komana Archaeological Research Project Archive.) 36 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA $e rest of the 'elds were poor in tiles. $e density in Field % came as no surprise since it was known that a building was present. $e density of tiles was analyzed for all 'elds, taking both fragment counts and weights into account (Figs. ,A–D, %"A, %"C). When the fragment counts are considered, signi'cant clusters were seen in Field %, Field !, and Field #. In Field %, the church 'eld, the tile/brick density dominated when weights were considered. Ceramic clusters were seen at Field #, Field *, and Field & when the fragment counts were calculated (Figs. ,A, ,C, %"B). Field # stood out, when the ceramic weights were taken into account (Figs. ,B, ,D, %"D). In contrast to tiles (Fig. %%), very few ceramics were recovered at the transects of Field %, which has been a pattern we observed at church sites during the earlier surveys. $e rest of the 'elds had a signi'cant number of ceramics, indicating domestic structures and related activities. $ese two distinctive patterns indicated that the site was of a church and a village. Dating of the site, especially of the settlement section, can only be done in a broad sense. $ere are glazed sgra.to ware (types very similar to Komana pottery from the twelfth to fourteenth century), high numbers of coarse ware (including storage vessels and pithoi), and nonglazed 'ne ware. Diagnostic pieces are very few and the pieces are often heavily FIG. 11 Tile finds from Field 1 (church) of the Pınarlı survey. (Courtesy of the Komana Archaeological Research Project Archive.) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 37 abraded. $e general characteristics of this pottery coincide with the twelfth- to fourteenth-century pottery from Komana (Fig. %!). One glazed piece with the face of a siren is an exact copy of an example from Komana and must have been made there. Another group of ceramics consists of sigillata, indicating a pre-Byzantine occupation phase. A piece of glass and the hand of a terracotta 'gurine must be from that phase. Broadly speaking, it would be possible to suggest that the pottery assemblage recovered in Fields * and & included more of the 'ner ware with sigillata and smaller pieces (as well as the terracotta 'gurine and the glass piece) compared to Fields ! and #, possibly indicating the location of the pre-Medieval site on the /at land. $e glazed and unglazed group together with the coarse ware indicate that the church had an interrelated settlement, a rural village located closer to the church (Fields ! and #) with a predecessor in the Roman period on the 'elds to its west (Fields * and &). Discussion Two main issues emerged from our revisits to church sites. One is the complex relationship between the so-called rural revival of the Middle Byzantine period, the emergence of church sites across the Komana landscape, and recognition of these churches as possible signals for villages/hamlets. $e second is an evaluation of the potential of quanti'ed surface data as an e0ective way for identifying these church sites. $ese issues will be covered in more detail below under contextual and methodological discussions. Contextual Interpretation Historical sources indicate turbulences throughout the Byzantine era as Anatolia shifts between peaceful periods and periods of unrest. $e impact of these turbulences is FIG. 12 Glazed ceramics from the Pınarlı survey. (Courtesy of the Komana Archaeological Research Project Archive.) 38 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA partially observable in changing features of larger urban settlements, which either shrink into forti'cations at times of distress or spread out over the adjacent arable lands during peaceful periods (for revival of forti'cations, see Niewöhner !"%): &)–&(; Foss %,,(: #+#, #+&–++; Niewöhner !"%+: !),; Stroth !"%): ##"; Wright !""): %*(– *,; Belke %,,": %+"–+#; Barnes and Whittow %,,(: #&%). $e most signi'cant series of events for the Anatolian lands are, 'rstly, the Persian wars in the sixth and early seventh centuries followed, secondly, by the Arab raids beginning with the seventh century and lasting through the end of the ninth century (Decker !"%+). $is period is marked by an interruption of communication between Constantinople and various cities of Anatolia, reorganization of the land into themata (Haldon !"%"), and resettlement in remote rural areas (Brubaker and Haldon !"%%) if not in forti'ed towns. $e impact on the rural lands is less easy to identify, partly due to lack of systematic surveys (especially in the Black Sea region) and partly due to the di.culty of reading surface material (in addition to the di.cult vegetation cover throughout the Black Sea) and recognizing continuity and abandonment (for the systematic surveys in the Black Sea region, see Düring and Glatz !"%&; Matthews and Glatz !"",). It had been a challenge to identify Early Byzantine sites in the territory of Komana, but toponyms such as gavur mezarlı"ı (cemetery of the in'dels) and kilise yeri (church site), used by the villagers in reference to locations near their villages in addition to various Armenian and Greek oral records (Erciyas and Ivanova, forthcoming), indicated a persistent and deep-rooted Christian tradition in the region. Erciyas and Sökmen published an overview of the Early Byzantine sources related to the introduction and spread of Christianity at and around Komana (Erciyas and Sökmen !"%"b: %!#–!+). From the tenth century onward, scholars describe a revival that lasted until the early thirteenth century. $is revival is re/ected in the Anatolian landscape by the increasing number of sites on /at arable lands. $is may be the reason why it has been much easier to recognize Middle Byzantine sites in the territory of Komana during surveys. $e ease with which churches can be recognized through surface material has been essential for discovering rural settlements. Identi'cation of Middle Byzantine sites during the surveys at Komana strongly depends upon the presence of architectural materials such as tiles, bricks, terracotta rosettes, decorated terracotta tiles, and mortar densities visible on the 'eld surface, which indicate churches (Erciyas and Sökmen !"%"b: %#%). At all three sites presented in this article, in addition to many others from surveys, architectural fragments that could be associated with a church have been found in abundance, while it has been impossible to suggest the presence of domestic architecture. $is must be due to the durability of materials such as tiles, which were exclusively used on public or religious buildings in rural settings. $e terracotta rosettes widely found at church sites and on Hamamtepe are rather peculiar objects that have assisted the team in identifying the structures as churches and dating them to the Middle Byzantine period (Fig. %#). Vorderstrasse described in detail the decorative elements of the Komana chapels and especially the rosettes, comparing them to the survey examples and those on the Bibi Hatun Türbesi in Tokat, a comparison that suggests common workshops (Vorderstrasse !"%,: *,–&!). Very few examples of rosettes were published from Late Byzantine churches in Sardis (Hanfmann %,(#: !"!; Buchwald !"%&: ),), Constantinople, and the Balkans (Trkulja !"%!: %*+, %&", %&&–&+). $ere is one example from a Middle Byzantine church in Amastris (Crow and Hill %,,&: !+"), and others from Neocaesarea (Niksar) were studied through museum inventories without contextual information (Ermi- !"%,). Regardless, the terracotta rosettes have been taken as important indicators of Middle Byzantine churches in the Komana region. Less durable or recognizable materials (irregular stones as foundations and mudbrick for superstructure) must have been used for houses, making it very di.cult to identify them without excavation. Tiles may have been used exclusively on the roofs, which may explain the limited presence on 'elds where domestic activity is recognized. Similarly, at the Sagalassos survey, durable materials helped identify churches such as those at Yassıgüme and A1lasun (tenth–eleventh century), while lack of diagnostic pottery and building materials resulted in the invisibility of Middle Byzantine and Ottoman sites (Vanhaverbeke and Waelkens !""#: #"*–)). $e Avkat FIG. 13 Terracotta rosettes from Komana. (After Vorderstrasse 2019: 74, fig. 3.) 4 0 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA survey team recognized a strong association between less mobile roof tiles and subsurface remains based on an assessment with GPR readings (Bikoulis et al. !"%&: %%*). During the Avkat survey, some over'red roof tiles were recovered, which were interpreted as evidence for on-site production (Bikoulis et al. !"%&: %%%). At Bula, Mezarlıktarla, and Pınarlı, over'red tiles were also found within the sampled areas. $is may similarly suggest on-site production rather than 're exposure, especially because of the large areas with high magnetism detected at Bula during the magnetometer survey. $e team previously identi'ed one of them in the north as the site of a large kiln. Tracing on-site tile production is a very important pursuit, and the topic needs further investigation. $e terracotta rosettes could also have been produced on-site together with tiles. $eir forms are various, careless, and non-uniform. To sum up, the visibility of churches in surface assemblages is both good and unique among the surveyed sites at Komana, and, so far, they have been identi'ed to be an integral part of a domestic compound with the presence of pottery evidence. Also, the intensive survey results showed that the tiles stay close to the building of their origin and do not move far easily (on the interpretation of tiles in surface surveys, see Poulter, Beckmann, and Strange %,,(). Methodological Issues $e function of the sites can be identi'ed through the amounts of scattered pottery and architectural materials. $erefore, ceramics and tiles were compared. In the Komana survey, the density of ceramics within the Pınarlı church 'eld (Field %) was low, while the density of tiles was high. Conversely, the other surveyed 'elds had more ceramics, while tiles decreased in density. $is, 'rst and foremost, indicated lack of domestic activities in the immediate vicinity of the church, as well as limited use of tiles in domestic architecture further away from the church. Since the tiles were observed in situ both at Pınarlı and Bula, they could be more securely dated to the Middle Byzantine period. $e pottery at the site, on the other hand, represented a broader chronological bracket and varied between domestic coarse wares with large pieces and 'ne ware with small pieces, with a slight di0erence in representation over the surveyed 'elds as described above. While the church could con'dently be dated to the Middle Byzantine period through its architectural details, the closest group of pottery both physically and chronologically included the non-diagnostic coarse ware and the glazed pottery of the twelfth to fourteenth century. In light of these on-site observations, two alternatives for Pınarlı may be suggested: (%) $e church was an isolated site without a corresponding settlement, abandoned with the arrival of the Danishmend/Seljuk Turks, and later a village was established to the north of the church. (!) $e church was part of a Christian village or a hamlet (possibly dating to the pre-Seljuk period) and continued during the Seljuk rule. Several other projects paid special attention to the differences between tile and ceramic densities. At Avkat’s super-intensive survey, a high density of ceramics was recorded together with tiles in the 'eld where geophysics indicated a church (Bikoulis et al. !"%&: %%%, %%#). During the survey at the village of A1lasun, which combined both geophysical prospection and surface collecting, it was concluded that the church was accompanied by other secular buildings, such as dwellings, storage spaces, and workshops (Talloen et al. !"%): #(+). Here, the spatial patterning of the building materials and ceramics con'rmed the subsurface features detected through geophysical prospection. In the absence of any forti'cation, the site was de'ned as an open rural site and the location (Field %)%) where the church was identi'ed was suggested to have been the core of the site (Talloen et al. !"%): #()). At Neoklaudiopolis, geophysical prospection was combined with intensive surface collection in the exploration of a Byzantine martyrium complex, and a correlation between the two data was con'rmed (Winther-Jacobsen !"%&: ()). $e presence of domestic pottery indicated that the martyrium complex was associated with a small settlement (Winther-Jacobsen !"%&: ((–(,). Identifying the church as part of a rural settlement at Pınarlı is debatable due to the absence of ceramics at the church site and the broad chronological representation at the site in general; however, it may still be argued based on the surface pottery scattered across the neighboring JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 41 'elds. $e distinct character of the landscape where and how the site is located (i.e., the church being nestled in a narrow valley) must be the main determinant for the character of the surface scatter. Also, geophysical prospection at the site in the future may improve our understanding. For Mezarlıktarla and Bula, the survey assemblages presented a more mixed picture of tiles and ceramics, safely suggesting the existence of domestic units around the churches. $is observation has been con'rmed with geophysics at both sites, similar to the examples mentioned above. During the extensive surveys at Komana, presence of Middle Byzantine churches was suggested, evaluating the qualitative data found on the surface. In particular, mortar remains together with tiles and terracotta rosettes have been key indicators of churches during the Komana survey. Intensive systematic collection at these preidenti'ed locations was implemented. $e spatial distribution of tiles concentrated around the previously identi'ed church locations. Geophysics at both Bula and Mezarlıktarla indicated other structures around the churches. $is observation was con'rmed by the quantitative surface data. Mortar scatters on the surface seem to be a very important determinant for identifying subsurface structures. Similarly, at Avkat and A1lasun (Sagalassos survey) a relationship between the surface 'nds and subsurface structures was recognized. At the Avkat survey, GPR suggested the presence of a church, and tile and brick concentrations on the surface supported the GPR results (Bikoulis et al. !"%&: %"+). Also, emphasis was laid on using a combination of multiple artifact types in identifying subsurface structures through surface representations (Bikoulis et al. !"%&: %%*). At A1lasun, identi'cation of a church site through high concentrations of ceramics and tiles was con'rmed by geophysical prospection (Talloen et al. !"%)). Overall, we can suggest a strong relationship between surface representation and subsurface anomalies. During the intensive surveys at Komana, observation points were used to analyze the distribution patterns of the surface materials around already suggested subsurface structures. $is was done also to see the clusters within the survey units rather than simply using transect lines for quanti'cation. $ese patterns were consistent for both the architectural materials clustering around the expected church spot and the ceramic clusters, which were clearly isolated from the churches, even though a slight number of ceramics was present at church spots, and, conversely, tiles were found at the neighboring areas of possible domestic function. Large-scale systematic surveys generally count the total amount of surface 'nds per 'eld. $is enables teams to make comparisons between hundreds of 'elds and to observe the signi'cant densities. Collecting observation point data is time- and energy-consuming and is more bene'cial when applied at strategically chosen and previously observed dense 'elds. $erefore, at Komana such collection was conducted at speci'cally chosen 'elds and was structure oriented. Projects such as Göksu and Avkat appealed to observation points continuously at all 'elds walked in order to create a common ground and be free of 'eld boundaries in the analyses (Newhard et al. !"%#). Conclusion Systematic intensive surveys in Turkish archaeology are not as widespread as they are in Greece and Italy; however, they prove to be vital to further understand various issues in Anatolian archaeology, especially those relating to rural life. At the same time, the focus of Byzantine archaeology has been mostly on urban settlements and important centers of religion and trade, resulting in only fragmentary knowledge of signi'cant transformative processes that a0ected the complete geography of the empire. One of these is certainly the economic boom in the tenth to eleventh century, which led to the emergence of new domestic and industrial units at sites, a revival of monasteries, as well as an intensi'ed rural life. $at phenomenon can best be observed in rural contexts, not so much in the urban landscape. In this study, evidence pertaining to the phenomenon was discovered in the apparent increase of churches across the Komana rural landscape in the Middle Byzantine period. $e study also attempted to position these churches within the site typologies proposed for this period in order to 42 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA identify the rural settlements. Revisiting the three previously identi!ed churches for intensive collecting and geophysics where possible indicated that two of the churches were part of small-size villages, and one was most probably part of Komana as indicated by the presence of domestic pottery and subsurface structures identi!ed through geophysics. Finally, our study aimed to test surface collection methods. Di"erences in the density of tiles and ceramics was recognized as an indicator for public (churches in this case) versus domestic units. We concluded that there is a strong relationship between surface representation and subsurface anomalies. Use of observation points on the !eld transects was bene!cial for recognizing di"erentiated clusters of tiles and ceramics, matching the suggested function of the site. Note Acknowledgments: We would like to thank numerous students and colleagues who participated in surface survey and surface collection, counting and visual documentation of the materials. In particular, we would like to thank Atakan Yüklü, who conducted and interpreted the georadar results at Mezarlıktarla. Our thanks also go to Dr. Tuna Kalayci for the implementation of the magnetometer survey at Bula, as well as to David Monsees for his contribution to the interpretation of the magnetometer results. Finally, we are grateful to the anonymous reviewers for their constructive comments since the beginning of the publication process of the article. References Akyürek, E. #$$%. Bir Ortaça& Sanatı Olarak Bizans Sanatı. In Sanatın Ortaça!ı: Türk, Bizans ve Batı Sanatı Üzerine Yazılar, ed. E. Akyürek, %#–''. Kabalcı Yayınevi $(, Sanat Dizisi %. Istanbul: Kabalcı Yayınevi. Barnes, H., and M. Whittow. #$$'. )e Survey of Medieval Castles of Anatolia (#$$*–$+): )e Maeander Region. In Ancient Anatolia: Fifty Years’ Work by the British Institute of Archaeology at Ankara, ed. R. Matthews, ,(%–-'. London: British Institute of Archaeology at Ankara. Belke, K. #$$.. Phrygia between Byzantines and Seljuks. Byzantinische Forschungen #+:#-$–+-. Bikoulis, P., H. Elton, J. Haldon, and J. Newhard. *.#-. Above as Below: )e Application of Multiple Survey Techniques at a Byzantine Church at Avkat. In Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Byzantine Period, ed. K. Winther-Jacobsen and L. Summerer, #.#–#%. Geographica Historica ,*. Stuttgart: Steiner. Brubaker, L., and J. Haldon. *.##. Byzantium in the Iconoclast Era c. "#$–#%$: A History. Cambridge: Cambridge University Press. Buchwald, H. *.#-. Churches EA and E at Sardis. Archaeological Exploration of Sardis, Report +. Cambridge, MA: Harvard University Press. Crow, J., and S. Hill. #$$-. )e Byzantine Forti!cations of Amastris in Paphlagonia. Anatolian Studies (-:*-#–+-. Decker, J. M. *.#+. &e Byzantine Dark Ages. London: Bloomsbury Academic. Düring, B. S., and C. Glatz, eds. *.#-. Kinetic Landscapes: &e Cide Archaeological Project: Surveying the Turkish Western Black Sea Region. Warsaw: De Gruyter Open. doi: #..#-#-/ $%',##.((($%#. Erciyas, D. B. *..+. Tokat /li Komana Antik Kenti Yüzey Ara0tırması *..(. Ara'tırma Sonuçları Toplantısı *, (*):#,–**. ———. *..%. Komana Yüzey Ara0tırması *..-. Ara'tırma Sonuçları Toplantısı *( (*):#--–++. ———. *.#,. Komana/Sisiyye’de Bir Ortaça& /0li&i: Bizans’dan Dani0mendliler’e Tokat’ın De&i0en Çehresi. In Güne' Karadeniz’den Do!ar—Sümer Atasoy’a Arma!an Yazılar, ed. 1. Dönmez, #,,–-.. Ankara: Hel Yayıncılık. ———. *.#$. Archaeology at Komana. In Komana Small Finds, ed. D. B. Erciyas and M. Acara Eser, #–(+. Settlement Archaeology Series %. Monograph *. Istanbul: Ege Yayınları. !"#$%&% '"() $%$*"+ (PhD ,-./, Middle East Technical University) is a faculty member of the Art History Department at Bartın University, Turkey. He participated in the Komana Archaeological Research Project (KARP) for .- years. In his PhD thesis, he dealt with intrasite spatial analysis of multidisciplinary data within Seljuk contexts at Komana. His main research interests are the medieval archaeology of Anatolia, field survey methods, environmental archaeology, and GIS applications. (Bartın Üniversitesi, Kutlubeyyazıcılar Kampüsü, Edebiyat Fakültesi, Merkez/Bartın /0.--, Turkey; [email protected]) 12')3 *"(4" 2(4)5%# (PhD ,--., University of Cincinnati) is a professor at the Graduate Program in Settlement Archaeology at Middle East Technical University, Turkey. She has been directing the Komana Archaeological Research Project (KARP) since ,--0. Her research interests include archaeology of the Black Sea, the Pontic Kingdom, Byzantine and Seljuk Anatolia, survey archaeology, public archaeology, and history of archaeology. (Orta Do6u Teknik Üniversitesi Mimarlık Fakültesi 7ehir ve Blg. Pln. Böl Dumlupınar Bulvarı No.. -89-- ODTÜ Ankara, Turkey; [email protected]) JOURNAL OF EASTERN MEDITERRANEAN ARCHAEOLOGY AND HERITAGE STUDIES ! 43 Erciyas, D. B., and P. Ivanova. Forthcoming. Bizeri Monastery near Komana in the Light of Oral Historical Records. In Tios/ Tieion on the Southern Black Sea in the Broader Context of Pontic Archaeology, ed. R. G. Tsetskhladze and 2. Yıldırım. Oxford: Archaeopress. Erciyas, D. B., and E. Sökmen. !"",. Komana Antik Kenti ve Çevresi Yüzey Ara-tırması !""). Ara!tırma Sonuçları Toplantısı !+ (%):!(,–#"+. ———. !"%"a. Komana Antik Kenti Arkeolojik Ara-tırma Projesi !""( Yılı Raporu. Ara!tırma Sonuçları Toplantısı !) (!):#&&–)*. ———. !"%"b. An Overview of Byzantine Period Settlements around Comana Pontica in North-Central Turkey. Byzantine and Modern Greek Studies #*:%%,–*%. Erciyas, D. B., E. Sökmen, and T. Kalaycı. !""(. Tokat 3li Komana Antik Kenti Yüzey Ara-tırması !""+. Ara!tırma Sonuçları Toplantısı !& (!):%,)–!%!. Erciyas, D. B., and M. N. Tatbul. !"%+. Anadolu’da Ortaça1 Kazıları ve Komana. Kazı Sonuçları Toplantısı #) (!):+%%–!+. ———. !"%(. Komana’da !"%+ Yılı Arkeolojik Ara-tırmaları. Ara!tırma Sonuçları Toplantısı #& (%):!&&–+(. ———. !"%,. Komana !"%) Yılı Kazı Çalı-maları. Kazı Sonuçları Toplantısı *" (%):*!&–#+. Erciyas, D. B., M. N. Tatbul, E. Sökmen, C. Kocabıyık, and R. Sünnetçi. !"%&. Komana’da !"",–!"%* Yılları Arasında Yapılan Kazı Çalı-malarının Ön De1erlendirmesi. In Komana Ortaça" Yerle!imi / #e Medieval Settlement at Komana, ed. D. B. Erciyas and M. N. Tatbul, !%–+!. Settlement Archaeology Series &. Monograph %. Istanbul: Ege Yayınları. Erdal, Y. S., D. Erdal, and M. M. Koruyucu. !"%&. Ortaça1’da Nüfus De1i-imi Öncesine Ait Bir Bizans Toplulu1u: Komana 3nsan Kalıntılarının Antropolojik Analizi. In Komana Ortaça" Yerle!imi / #e Medieval Settlement at Komana, ed. D. B. Erciyas and M. N. Tatbul, (#–%%*. Settlement Archaeology Series &. Monograph %. Istanbul: Ege Yayınları. Ermi-, Ü. M. !"%,. Niksar’dan Bizans Dönemine Ait Keramoplastik ve Çini Örnekleri. Art-Sanat Dergisi %!:%,#–!"*. Foss, C. %,,(. $e Survey of Medieval Castles of Anatolia %,(!–%,(*. In Ancient Anatolia: Fifty Years’ Work by the British Institute of Archaeology at Ankara, ed. R. Matthews, #&,–++. London: British Institute of Archaeology at Ankara. Haldon, J. F. !"%". #e Palgrave Atlas of Byzantine History. Houndmills, Basingstoke, UK: Palgrave Macmillan. Haldon, J. F., N. Roberts, A. Izdebski, D. Fleitmann, M. McCormick, M. Cassis, O. Doonan, W. Eastwood, H. Elton, S. Ladstätter, S. Manning, J. Newhard, K. Nicoll, I. Telelis, and E. Xoplaki. !"%*. $e Climate and Environment of Byzantine Anatolia: Integrating Science, History, and Archaeology. #e Journal of Interdisciplinary History *&:%%#–+%. Hanfmann, G. M. A. %,(#. Sardis: From Prehistoric to Roman Times; Results of the Archaeological Exploration of Sardis $%&'–$%(&. Cambridge, MA: Harvard University Press. Hanilce, M. !"%#. $'$)–$'$* Tarihli ($+ Numaralı) Tokat ,er’iye Sicili’nin Transkripsiyonlu Metni ve De"erlendirmesi. Ankara: Türk Tarih Kurumu Yayınları. Harvey, A. !""(. $e Village. In #e Oxford Handbook of Byzantine Studies, ed. E. Je0reys, J. Haldon, and R. Cormack, #!(–#*. New York: Oxford University Press. Holmes, C. !""(. Political-Historical Survey, (""–%!"*. In #e Oxford Handbook of Byzantine Studies, ed. E. Je0reys, J. Haldon, and R. Cormack, !+*–),. New York: Oxford University Press. Izdebski, A. !"%!. $e Changing Landscapes of Byzantine Northern Anatolia. Archaeologia Bulgarica %+ (%):*)–++. Karasu, Y. E. !"!". Komana Anadolu Selçuklu Ça"ı Seramikleri. PhD diss., Hacı Bayram Veli University. Matthews, R., and C. Glatz, eds. !"",. At Empire’s Edge: Project Paphlagonia; Regional Survey in North-Central Turkey. BIAA Monograph **. London: British Institute of Archaeology at Ankara. Newhard, J. M. L., N. S. Levine, A. D. Phebus, S. Craft, J. D. Little'eld. !"%#. A Geoinformatic Approach to the Collection of Archaeological Survey Data. Cartography and Geographic Information Science *":#–%). Niewöhner, P. !"%+. $e Byzantine Settlement History of Miletus and Its Hinterland—Quantitative Aspects: Stratigraphy, Pottery, Anthropology, Coins, and Palynology. With contributions by A. Demirel, A. Izdebski, H. Sancaktar, N. Schwerdt, and H. Stümpel. Archäologischer Anzeiger !"%+ (!):!!&–,". doi: %".#*)("/dyap-)%*e. ———. !"%). Urbanism. In #e Archaeology of Byzantine Anatolia from the End of Late Antiquity until the Coming of the Turks, ed. P. Niewöhner, #,–&,. New York: Oxford University Press. Ousterhout, R. G. %,,,. Master Builders of Byzantium. Princeton: Princeton University. ———. !"%+. Bizans’ın Yapı Ustaları. Koç Üniversitesi Yayınları )). Istanbul: Koç Üniversitesi Yayınları. Pi-kin, E. !"%&. Byzantine and Ottoman Animal Husbandry at Komana. In Komana Ortaça" Yerle!imi / #e Medieval Settlement at Komana, ed. D. B. Erciyas and M. N. Tatbul, %%&–#(. Settlement Archaeology Series &. Monograph %. Istanbul: Ege Yayınları. Pi-kin, E., and M. N. Tatbul. !"%&. Archaeobotany at Komana: Byzantine Plant Use at a Rural Cornucopia. In Komana Ortaça" Yerle!imi / #e Medieval Settlement at Komana, ed. D. B. Erciyas and M. N. Tatbul, %#,–++. Settlement Archaeology Series &. Monograph %. Istanbul: Ege Yayınları. Poulter, A. G., M. Beckmann, and P. Strange. %,,(. Field Survey at Louloudies: A New Late Roman Forti'cation in Pieria. #e Annual of the British School at Athens ,#:*+#–&%%. Roberts, N., M. Cassis, O. Doonan, W. Eastwood, H. Elton, J. Haldon, A. Izdebski, and J. Newhard. !"%(. Not the End of the World? Post-Classical Decline and Recovery in Rural Anatolia. Human Ecology *+:#"&–!!. Stroth, F. !"%). Aezani. In #e Archaeology of Byzantine Anatolia from the End of Late Antiquity until the Coming of the Turks, ed. P. Niewöhner, #!)–#!. New York: Oxford University Press. Talloen, P., R. Vandam, M. Broisch, and J. Poblome. !"%). A Byzantine Church Discovered in the Village of A1lasun 4 4 ! T H E P O T E N T I A L O F Q U A N T I F I E D S U R FA C E D ATA (Burdur): Some More Light on Dark Age Pisidia. ADALYA !":#)&–*"*. Tatbul, M. N. !"%). Identifying Medieval Komana in the %!th–%#th Centuries through Spatial Analysis of Archaeological Data with a Multidisciplinary Approach. PhD diss., Middle East Technical University. ———. !"!". Arkeolojik Verilerin I!ı"ında Ortaça" Komana’sı: Çok Disiplinli Bir Yakla!ım. ODTÜ !"%) Yılın Doktora Tezi Kalbiye Tansel Yayın Ödülü. Ankara: ODTÜ Yayıncılık. ———. !"!%. Abandonment, Continuity, Transformation: Setting Komana into Archaeological Context through the Middle Byzantine and Early Turkish Periods. ADALYA !*:#!&–&#. doi:%".*)&(,/adalya.%"#+)#". Tatbul, M. N., and D. B. Erciyas. !"%,. Evaluation of the Recent Finds at Komana from the Early and Middle Byzantine Period. In Settlements and Necropoleis of the Black Sea and Its Hinterland in Antiquity, Select Papers from the #ird International Conference “#e Black Sea in Antiquity and Tekkeköy: An Ancient Settlement on the Southern Black Sea Coast,” )(–)% October )-$(, Tekkeköy, Samsun, ed. G. R. Tsetskhladze and S. Atasoy, !)!–(". Oxford: Archaeopress. ———. !"!". Komana Kazısının %".Yılı: !"%( Yılı Çalı-maları. Kazı Sonuçları Toplantısı *% (%):&#)–&". Trkulja, J. !"%!. $e Rose Window: A Feature of Byzantine Architecture? In Approaches to Byzantine Architecture and Its Decoration: Studies in Honor of Slobodan .ur/i0, ed. M. J. Johnson, R. Ousterhout, and A. Papalexandrou, %*#–+*. Farnham: Routledge. Vanhaverbeke, H., and M. Waelkens. !""#. #e Chora of Sagalassos: #e Evolution of the Settlement Pattern from Prehistoric until Recent Times. Studies in Eastern Mediterranean Archaeology &. Turnhout: Brepols. Vorderstrasse, T. !"%,. Experiencing the Medieval Churches of Komana. In Komana Small Finds, ed. D. B. Erciyas and M. Acara Eser, *)–)+. Settlement Archaeology Series ). Monograph !. Istanbul: Ege Yayınları. Whittow, M. !""(. $e Middle Byzantine Economy (+""–%!"*). In #e Cambridge History of #e Byzantine Empire c. &--–$1%), ed. J. Shepard, *+&–,!. New York: Cambridge University Press. Winther-Jacobsen, K. !"%&. Contextualizing Neoklaudiopolis: A Glimpse at Settlement Dynamics in the City’s Hinterland. In Landscape Dynamics and Settlement Patterns in Northern Anatolia during the Roman and Byzantine Period, ed. K. Winther-Jacobsen and L. Summerer, (#–,,. Geographica Historica #!. Stuttgart: Steiner. Wright, G. R. H. !""). Beyce Sultan: A Forti'ed Settlement in Byzantine Phrygia. Anatolica ##:%*#–+(. doi: %".!%*#/ANA. ##.".!"!%)&). Xoplaki, E., D. Fleitmann, J. Luterbacher, S. Wagner, J. F. Haldon, E. Zorita, I. Telelis, A. Toreti, and A. Izdebski. !"%+. $e Medieval Climate Anomaly and Byzantium: A Review of the Evidence on Climatic Fluctuations, Economic Performance and Societal Change. Quaternary Science Reviews %#+:!!,–&!.