Examining the Late Medieval Village from the Case at Ambroyi, Armenia

Abstract

Previous articleNext article FreeExamining the Late Medieval Village from the Case at Ambroyi, ArmeniaKathryn J. Franklin, Tasha Vorderstrasse, and Frina BabayanKathryn J. FranklinSchool of the Art Institute, Chicago Search for more articles by this author , Tasha VorderstrasseThe Oriental Institute of the University of Chicago Search for more articles by this author , and Frina BabayanNational Academy of Sciences Institute of Archaeology and Ethnography, Yerevan* Search for more articles by this author PDFPDF PLUSFull Text Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreIntroduction: By the Side of the Road and at the Edge of the ParadigmThis article examines the archaeological evidence from excavations at the medieval Armenian village of Ambroyi dating to the 13th–14th centuries ad (all dates throughout are ad). It focuses on reconstructing medieval life in the village and situates its analysis within wider trends of studying village archaeology in the medieval Near East. First, the article examines how villages have been approached in the wider Near East, before looking at the specific challenges of studying the village in Armenia in particular. It will then turn to evidence from archaeological excavations and what they reveal about villagers in medieval Armenia as participants in various social institutions, and in medieval life as a greater phenomenon. The data from Ambroyi contributes to an important work of integration, bringing studies of medieval Armenian and Near East society into conversation with each other. The research presented here also demonstrates the significance of medieval Armenia as a case study which bears upon wider discussions of medieval sociality, interaction, and complexity in Eurasia generally. A critical result of the research at Ambroyi is the empirical foundation for arguments regarding not only the continuation of social life in villages during periods of so-called “upheaval,” such as the 13th c. Ilkhanid period, but also for the participation of village inhabitants in interactions extending beyond the village site itself to towns, cities, and the passing travelers who slept and ate at the nearby caravan inn.Medieval Villages in the Near East: Discourses and DeconstructionsThe village as a site of social life has been historically marginalized in archaeological investigations of the medieval Near East. In general, with the exception of salvage excavations or cases of accidental discovery, archaeological excavations have followed the lead of medieval geography and focused on life in urban centers. This concentration on cities began with the first archaeological excavations in the Near East that targeted Islamic cities. Early excavations focusing on the medieval period in the Near East were frequently underwritten by museums and private individuals, and driven by the aim of procuring material for museum collections. As a result, these projects focused on major centers of medieval elite life such as Samarra, Nishapur, Rayy, and Fustat, where the results were expected to be the most spectacular and thus generate exemplary museum objects.1 Significant excavations of village sites, such as those carried out at Alishar Huyuk and Chatal Hoyuk in the Amuq, were undertaken in the course of investigations of earlier periods that underlay (or were intruded by) medieval contexts and materials.2A major influence on research has also been the latent presumption that while village life was socially important in the Byzantine and Christian medieval ecumene, the primary locus of social production within the so-called “world of Islam” was the “Islamic city,”3 which has therefore been the focus of historical and archaeological scrutiny to the exclusion of rural settlement. Emphasis has been placed on the construction of new Islamic cities, or the evolution of already-existing cities after the Islamic conquest in the 7th century.4 In the historical and geographical sources often used as the source of inspiration for these archaeologies, villages generally do not have names or are only mentioned in passing. Village names do appear in other sources, such as endowment inscriptions (see below), manuscript colophons,5 and lists of churches.6 In a similar trend, intense study on the part of historians, archaeologists, and scholars interested in Islamic period (i.e., post-7th c. medieval) architecture has focused primarily on the monumental forms which are both located in and categorically define cities. This focus on urban centers as the locus of Islamic social life, centered on monumental institutions, has only relatively recently been complemented by research on extra-urban social contexts as well as the architectural forms, such as castles and caravanserai, which have long attracted study as part of the larger monumental corpus.7In the Caucasus, this is also manifested in the excavation of cities (see Figure 1), namely the Armenian cities of Ani and Dvin,8 as well as the fortresses of Garni and Anberd.9 In Azerbaijan, attention has again largely focused on cities, such as Oren-Kala, Gabala, and Shamkir.10 The same is also true in Georgia, where again excavations focused on the fortress city of Dmanisi and cities including Mtskheta and Tbilisi, or on well-known monastery complexes such as David Gareja.11 There have been a number of recent publications of material from Georgia and Armenia that have largely focused on the medieval pottery of the region,12 again primarily coming from these city excavations, thus emphasizing the cities as important centers of wealth and commerce.Figure 1. Map of the Caucasus and adjoining regions, showing the locations of some of the medieval sites mentioned in this article.View Large ImageDownload PowerPointThe past twenty years, however, have brought an increasing recognition that medieval villages are a significant and productive object of study in their own right, as the loci of various forms of medieval daily life. This development has emerged as a result of shifts not only in Near Eastern archaeology, but as part of a general trend in history and archaeology toward an intensified focus on, in the former, histories of the everyday, and in the latter, on material culture and processes situated at the scale of the household or the community.13 In particular, work in Jordan has brought attention to Mamluk, Ottoman, and post-Ottoman villages14 as well as “Early Islamic period” (7th–9th centuries) single houses.15 Studies in the Islamic world have increasingly begun to look at settlements in marginal areas, such as marshes16 and deserts, where the spread of Islam into rural areas not long after the Islamic conquest has been studied in detail.17 In Byzantine Cappadocia, villages have been studied in considerable detail,18 including specific work of different aspects of domestic architecture such as courtyards,19 kitchens,20 and stables.21 Therefore, we are becoming increasingly better informed about villages in the Near East.Yet, while the village as a setting of social activity is receiving increased attention, archaeologists still must negotiate the lingering tendency to define the rural village in opposition to the urban, and persistent assumptions that are made about what kind of sociality is possible in a village. In particular, the implications of using the village as an unproblematic stand-in for the quotidian baseline in models of medieval sociality centered on (Christian or Islamic) cities is illustrated by the ways in which villages are “read” in archaeological analysis. For example, studies on village architecture in the Crusader southern Levant have applied ethnically deterministic models to argue for the presence of Crusaders and local Christians in planned (orthogonal) villages in contrast to local Muslims in more “agglutinative” villages.22 This distinction between religions is also observed in Egypt, where archaeologists who study the architectural remains associated with the Coptic Christian community do not study the contemporary “Islamic” houses from sites such as Fustat thought to be inhabited primarily by Muslims, or vice versa.23 Other studies of villages in Greece have made explicit the assumed direct connections between modern ethnography and archaeology,24 thereby mapping the social landscape of the present onto the past.Our work at Ambroyi both builds upon this corpus of published work on medieval villages, but also aims to continue challenging some of the discourses and assumptions about the role of the village in medieval sociopolitics that have structured archaeological interpretations of excavated villages. Most importantly for this article, a wave of development projects in Turkey, Azerbaijan, and Georgia in the last several decades (in the case of Azerbaijan and Georgia, specifically the construction of the Baku-Tbilisi-Ceyhan pipeline) has resulted in a series of rescue excavations, undertaken to record sites that would otherwise completely disappear due to rising dam waters or the construction of oil pipelines.25 This phenomenon meant that a number of villages were investigated in detail: village sites in the Republic of Turkey include Gritille, Aşvan Kale, Taşkun Kale, and Tille Höyük on the Euphrates; and various villages in Azerbaijan and Georgia located on the BTC pipeline.The lack of architectural plans found in Azerbaijan, perhaps because the architecture was mud brick and not preserved,26 means that it is not possible to make comparisons between these villages and those in Turkey. Further, these village sites have been dated by archaeologists to the 12th century or earlier, meaning that they are not totally contemporary with the villages in eastern Turkey, Armenia, and Georgia. Nonetheless, archaeologists working on the pipeline material in Georgia documented a number of villages that can be used for comparison with the villages in eastern Anatolia and Armenia. The material from the BTC pipeline, as well as surveys in Nakhchivan, Azerbaijan, and southern Georgia supplement our own recent work, demonstrating that village settlements were part of a broad and complex landscape of sites, infrastructure, and land use during the late medieval period throughout the South Caucasus. The work in Georgia and Azerbaijan, because it was concentrated in particular geographical areas, suggests the existence of a rich hinterland beyond the urban centers of the late medieval south Caucasus which remains to be studied.27In Georgia, the villages documented by archaeologists working on the BTC pipeline (see Figure 2) were all built from stone and documented extensively. Some of these villages had been investigated previously, but—thanks to the pipeline investigations—these have now been published in an accessible way and dated. Village sites in Georgia are markedly similar to the case documented by Sharon Gerstel in Greece,28 in that they are focused on a church in the center of the village, with any additional churches located at the fringes of the main settlement.29 The centrality of the church to the village, as documented in medieval Georgian and Greek villages, can also be observed at the Armenian village of Mren in eastern Turkey (see below), where the Mren cathedral was at the center of the settled area. Villages in Georgia were also often fortified, and many round structures can be identified in them, which could be towers attached to the houses, suggesting they might have been for storage or animals. Additionally, there were buildings made up of a series of small rooms, many of which had rounded corners; archaeologists excavating these rooms did not document the activity areas which can be observed in eastern Turkish and Azerbaijani villages (see below).Figure 2. Comparison of architecture footprints of late medieval villages in Georgia: A. Tkemlara. B. Nachivchavebi. C. Sakire. D. Takhtiskaro. E. Tskhratskaro. F. Chivchavi. All plans after Heritage Protection Department of Georgia, 2003.View Large ImageDownload PowerPointAt Nachivchavebi village (11th–16th centuries; see Figure 2B), the excavators observed that the domestic houses typically had five to six rooms, including kitchens, wine cellars, bakeries, and places for storing cattle; the structures were associated with fields nearby, surrounding the church in the center of the village. The church in the center of the village was surrounded by a wall and had a cemetery, while there was also another church on the fringes of the village that also had a cemetery. This is similar to Tkemlara village (11th–16th centuries) (see Figure 2A), which had similar patterns of domestic settlement and a church in the village. Chivchavi Gorge village (10th–16th centuries; see Figure 2F) also consisted of groups of settlements with a church and a cemetery, and a fortification wall. Takhtiskaro (see Figure 2D) village (10th–16th century) was a large settlement that had houses which varied in size from one to seven rooms and a fortification wall. There was also a small church on the fringes of the village. At Tskhratskaro Gorge village (medieval period; see Figure 2E), the village consists of houses between three to five rooms, and a round room that was separate from the rest of the house. There were also places for cattle that had been dug into the ground. Sakire Village (see Figure 2C) was associated with the 11th–13th century Sakire fortress, and was built below the fortifications.30 Additional village types included houses built from mud (e.g., Narli Dara in Georgia) and buildings with paved floors that were used to store cattle.31 Other structures built apart from the primary clusters of village structures suggest that certain specialized production, such as wine making, may have been distributed through space, tying the landscape of fields, buildings and other spaces together into a larger landscape of village activity.32The body of rescue research in Turkey demonstrates an interesting pattern of interpreting the medieval landscape, in that ethnic or social evolutionary logics are frequently (and tacitly) applied to explain perceived patterns in village architectural forms. Excavated villages in eastern Turkey and the Amuq (see Figure 3) are frequently described as “agglutinative” in form, implying that the lack of apparent clear distinctions between rounded “houses,” or evidence for additive construction over time, indicates a lack of town planning on the part of the village occupants.33 At the root of analytical oppositions between planned and unplanned villages is a linked conceptual polarization between villages as prepolitical communities and villages as productive units integrated within a scalar political economy.Figure 3. Comparison of Abroyi village excavated area with excavated architectural footprints of late medieval villages in Turkey: A. Taşkun fortress (after Redford, Archaeology of the Frontier, 69); B. Lidar 1; C. Lidar 2 (both after ibid., 71–72); D. Gritille (after ibid., 40–43); E. Tille Höyuk (after ibid., 74); F. Aşvan Kale 1 (after Mitchell, Asvan, Fig. 19); G. Aşvan Kale 2 (after Mitchell, Asvan, Fig. 16); H. Ambroyi; I. Chatal Hüyük (after Richard C. Haines, Excavations in the Plain of Antioch II: The Structural Remains of the Later Phases: Chatal Hüyük, Tell al-Judaidah, and Tell Tayinat (Chicago, 1971), pl. 28); J. Sos Höyük (after Sagona and Sagona, “Upper Levels at Sos Höyük, Erzurum,” Fig. 2).View Large ImageDownload PowerPointThis seemingly straightforward opposition between forms of village sociality (planned and controlled vs. agglutinative and communal/organic) is already called into question by the particulars of excavated villages. At Tille Höyük (see Figure 3E), excavators found buildings of mud brick built on stone with staircases, constructed around a courtyard that seems to have had specific activity areas. The rooms were altered through time as the complex was rebuilt and may have been dwelling houses of families.34 Similar organic dwelling houses were found at Korucutepe, where the excavators found an alleyway, ovens, and buildings;35 and at Sos Höyük (see Figure 3J), where excavators found rectangular stone houses with cobblestone areas and paved alleyways.36 The exposure at the site of Aşvan Kale (see Figures 3F–G) was much smaller but nonetheless illuminating for our understanding of the different types of settlement that occurred in villages; the settlement seems to have incorporated both living and production spaces, thus further challenging categorical oppositions of the domestic and industrial for late medieval Anatolia. The initial phase of settlement was not well-preserved, but the subsequent settlement was marked by walls of stone and mortar 80 cm thick. The excavators found signs of pottery-making in the form of kilns; in subsequent phases, a large building (identified as a medrese) was built over the top of this industrial area.37The existence of orthogonal architecture or regular plans at some sites has resulted in a solidification of village formal types across the greater Anatolian and Iranian region for this period, 38 especially at sites presumed to have been under state control or influence. For instance, the excavations at Taşkun Kale (see Figure 3A), centered on a regular mudbrick oval “fort” with towers built on stone foundations. Yet even within this “planned” military structure, excavators noted that the population made certain choices about the placement of installations, such as ovens, in rooms. There was also a settlement outside the fortress that had an associated church and graveyard.39 The juxtaposition of various building forms, placement choices, and layout strategies at these settlements—including Aşvan Kale, Tille Höyük, Taşkun Kale, and Ambroyi—suggest that the ultimate influence on site form is a complex of historically and locally particular factors rather than polar socio-evolutionary dynamics. In other words, seemingly “unplanned” or “agglutinative” villages could be and were integrated within late medieval economic structures and political systems. These type of “planned”—in the sense of orthogonal as opposed to organic or agglutinative—villages have not been documented in Azerbaijan or Georgia. There is an indication of planned villages, such as the village that apparently housed the servants who worked at Queen Tamar’s summer palace—but significantly, in this case, the categorization of “planned” refers to intentionality only, and not to the form of the structures themselves.40For the remainder of this article we will narrow our regional focus to the case study of the social life of the village in late medieval (1200–1500) Armenia. The question is how to develop the role of the rural village settlement in regional and even global social life in the late medieval period. As we will discuss, to do this in the Armenian context means breaking apart paradigms that maintain the village as central to ethnic memory but peripheral to social action—if not altogether absent—in the late medieval period.Late Medieval Villages in the Republic of ArmeniaThe spatial pattern of archaeological research in the Republic of Armenia developed, on the one hand, out of a focus on named urban centers (itself inherited from medieval archaeology’s strong dependence on historical sources) and, on the other, from an operational understanding of medieval sociality derived from Marxist historical materialism and urban-centered European traditions.41 The Marxist model prescribes a historical trajectory whereby only medieval urbanism enabled the increased divisions of labor which were necessary precursors to capital accumulation; the parallel tenets of early 20th-century economic historians like Henri Pirenne stipulated that medieval market economy and urban life were inextricable and mutually dependent. Therefore, pioneering research into the nature of medieval life in Armenia (as elsewhere) was essentially a search for cities. The foundations of this tradition were laid in many ways by the investigations of Nicolai Marr in Shirak and at Ani in the late 19th and early 20th centuries. Marr argued—as other scholars such as James Breasted and Franz Boas did in this period—that not only Europe, but also nations outside it, contributed to the development of world civilization, and Marr was looking for native Armenian contributions to this developmental trajectory. Marr’s research was focused, in part, on a search for an essential Armenian culture which was expressed in megalithic monuments, and which reached an apex in the form of medieval urban life.42Subsequent students of medieval archaeology continued to concentrate, as Marr had, on the cities that were known from historical accounts—indeed, these historical accounts continue to be used by archaeologists to date features found at urban sites. Historians and archaeologists after Marr (such as Babken Arakelyan, Karo Ghafadaryan, and Aram Kalantarian) excavated urban sites because they believed such centers would yield evidence for cross-cultural interactions that they clearly believed could not be found in villages. 43 They did investigate other types of sites such as castles (e.g., Anberd) and churches (e.g., Zvartnoc), but this was because they were important historical sites or had impressive standing architecture.44 This research served to confirm the representations of Armenian social structure in the medieval period presented within historical accounts—specifically, the social dominance of Christian princes or naxarars and their military exploits, and the importance of cities as centers of economic and political life.A combination of historical biases regarding the dynamics of social and cultural change in the medieval period have resulted in the medieval Armenian village being, paradoxically, simultaneously marginalized as an object of study and fetishized as the locus of primordial Armenian ethnos. A longstanding approach casts the city as an alien element in Armenia; conversely, the mountain village is implicated as the natural type of aggregated settlement in the Armenian highlands. Yet after the Islamic conquest in the 7th century, the Arab settlers are thought by scholars to have been solely Muslim urban dwellers in cities.45 This historical assumption comes, on the one hand, from accounts of Arab tribes settling in cities such as Amida, Arzn, and Npʾrkert,46 and on the other, from evidence such as the treaty between Theodore Rshtuni and the caliph Muʾawiya, in which the caliph states: “I shall not send emirs into your fortresses, nor a Tačik officer, nor a single horseman,” thus rhetorically evoking the Armenian naxarar aristocracy and their loyal Christian subjects as ensconced in extra-urban mountain strongholds.47 This oppositional characterization of Armenian villagers and urban Arabs persists despite evidence such as an inscription at Zvartnoc that clearly points to the presence of Arabic speakers in the countryside.48 An implication of this historical narrative is that villagers in their “natural” Armenian environment remained Armenian and Christian, thereby preserving native Armenian culture in the face of “Islamization” efforts by the new invaders. Historians and archaeologists have imagined the early medieval Armenian landscape as populated by “alien” cities filled with foreigners, with native culture and heritage preserved in villages, monasteries, and churches. The primeval Armenian landscape has therefore been characterized as occupation on a small scale. For the 12th century and later, however, this narrative is contradicted by the narratives of the primary historical record, in which cities are the seats of Armenian culture and the exclusive locations of cultural production, conceptualized in opposition to the nomadic invaders—Seljuk Turks and Mongols—who would buffet the walls of these cities with their mounted armies. For example, the 12th century account of Matteos Urhayeci (Matthew of Edessa) describes the Armenian inhabitants of the city of Ani holding out against repeated Seljuk attempts to conquer it.49This latent argument for the localization of an original Armenian culture in upland villages was made more explicit in syntheses of the history of Armenian architecture. Drawing in part from the perceived genetic relationship between Armenian medieval architecture and gothic architectural forms in Europe, as argued by analysts like Ernst Herzfeld, architectural historians such as Ohannes Xalpakhchyan and Anatoly Yakobson argued for the roots of “high” medieval church and civic architecture in the form of the Armenian village house.50 This set of arguments cast medieval Armenians as a whole as active rather than passive in medieval cultural exchanges, but ironically removed agency from that medieval Armenian village and rural landscape as a locus of cultural production, in that it was relegated to a primordial past and immobilized as a non-evolving ideal.This phenomenon has relegated village life in medieval Armenia to a condition of stasis, with the implication that the experience and lifeways of villagers did not change from the 4th century to the present, other than in exogenous terms such as invasion or heavy taxation.51 The Islamic conquest ushered in what is believed by Armenian historians to be a time of economic deprivation and intense taxation. That approach that has been applied to the late medieval period as well, when once again the region is thought to have been in decline after the Mongol invasions, and under the heavy yoke of Ilkhanid administration. Whether framed by the feudal parameters of a historical materialist tradition (by Armenian history and archaeology) or through the focus on artists, patrons, and elite actors (by art history), the medieval village landscape has been imagined as timeless and unchanging, much in keeping with longstanding modes of imagining the medieval world generally.52 Armenian village life in the late medieval period is construed as a precursor to the socioeconomic developments of the urban, proto-capitalist late medieval world.The developments in Armenia, however, reflect what was happening in other parts of the south Caucasus under the Ilkhanid administration. Armenian historians frequently refer not only to events in Armenia itself, but also in Georgia and what is now modern Azerbaijan. Indeed, from the perspective provided by archaeological data, it has been observed that medieval villages in the South Caucasus show remarkable stability in settlement despite the seeming disorder that would have been caused by the documented historic events.53Ambroyi Village and the Kasakh Valley in the Late Medieval PeriodThe archaeological datasets which form the core of this analysis were excavated on the eastern slope of Mt. Aragats, on the western side of the Kasakh River Valley near the contemporary village of Arai (known historically as Bazarjuł, hereafter designated as Arai-Bazarjuł). The Kasakh Valley is located on the edge of the highland plateau between the broad volcanic peak of Mt. Aragats and the curving Tsaghkunyats range (See Figure 4). The Kasakh River is fed by tributaries running out of both mountain systems, and cuts a deep canyon as it descends out of the highlands towards the Arax River valley to the south.Figure 4. Map of the general research area of the Kasakh Valley and Mt. Aragats, showing the site of Ambroyi and nearby late medieval sites.View Large ImageDownload PowerPointOngoing research in the region has demonstrated the intensive occupation of the landscape of the Kasakh Valley and Tsaghkahovit Plain for the past several millennia; the medieval period was no exception to the pattern of active construction of its social landscape.54 The ethnohistorical records identified the general area of ruins south of Arai-Bazarjuł village as Ambroyi; this area of surface remains was dated to the late medieval period (12th–15th cs.) using surface ceramic material collected by Kathryn Franklin in 2010.55 The specific region of architecture investigated in 2013 and 2014 was designated Hin Bazarjuł, or “Old Bazarjuł.”56 Our investigations at Ambroyi were driven by an array of questions directed at late medieval social life in the Kasakh Valley, and particularly at how people living in villages in the Kasakh were integrated into the wider worlds of Armenia, the Near East, and the Silk Road cultural ecumene. In this aim, our research departed from previous work in the Kasakh Valley focused on monastic life in late medieval monasteries, or on princely life in fortified castles on the slopes of Mt. Aragats.57 Our excavations were also motivated by the close spatial association of the settlement remains with the Arai-Bazarjuł karavanatun (“caravan inn”), located 500m to the east and dated to 1213. Franklin has previously argued that the karavanatun was a site for the mediation of long-distance travel both by localized political projects undertaken by Armenian merchant princes, as well as by local cultures of hospitality, particularly in the form of locally-provisioned road food.58In excavating a settlement which was initially dated to the 13th–15th centuries and spatially associated with the karavanatun, we were testing a number of premises put forward in Armenian medieval historiography regarding the temporality and spatiality of social life—especially economic life—in the post-Seljuk period. As will be further discussed below, we were critically interested in the argument advanced by Armenian archaeologists and historians that social life in Armenia contracted to cities or stopped altogether in the period following the 1236 Mongol invasions of the South Caucasus. As the following sections examine, travel and trade during the Ilkhanid period in Armenia not only stimula