7 Conclusions
This is an online digital edition from ISAW Digital Monographs. The print edition of this work can be consulted at https://isaw.nyu.edu/publications/isaw-monographs/ain-el-gedida
7.1 Issues of Chronology
The chronology of ʿAin el-Gedida was established on the basis of a study of the architectural and material evidence gathered during the excavations carried out at the site. Although of significant value in many respects, the analysis of the architectural features investigated on the main hill, and also of those surveyed on the other mounds, does not provide particularly relevant information to establish, with any degree of precision, an absolute chronology of the site. Construction materials used at the site (mostly mud bricks and a few stones used as lintels), the construction techniques employed (often mixed and of rather poor quality), and the dimensions of wall and vault bricks (of standard Roman size with little variation) cannot be identified as proper to any specific geographic and chronological range within Late Antique Egypt. Both the materials and the techniques employed at ʿAin el-Gedida are, in fact, the same used for centuries at innumerable other sites in Dakhla, the nearby oases, the larger region of the Western Desert, and throughout Egypt. That noted, the spatial arrangement discerned at ʿAin el-Gedida, on mound I and partially on the smaller hills, is very complex and unusual, not resembling too closely the layout of other known village or monastery-like settlements.
The study of the architectural evidence has allowed us to discover the existence not only of limited alterations and/or restorations, involving features such as walls and doorways, but also of substantially different construction phases, identified in several rooms of mound I below floor level.1 The discovery, in 2008, of a large building near the west edge of the hill provided further information testifying to the multi-phased construction history for the site. As discussed above, the underlying layout of this structure, which served, at least in its latest occupational phase, as an industrial workshop for the production of pottery, seems to reflect that of a small-scale pagan temple made of mud bricks.2
Overall, the archaeological record about ʿAin el-Gedida, especially the temple and the church complex, points to a layered history that seems to extend back in time further than the chronological range, established through the study of the material evidence from occupational levels, would suggest. In particular, the identification of the west complex as a temple would support the existence of the settlement since probably at least the second century CE, certainly by the middle of the third century, when most building activity concerning pagan temples seems to end in Egypt.3 The absolute chronology of each phase is very difficult to reconstruct, however. The older architectural features were, in several instances, razed to the lowest courses of bricks. These were often laid directly on gebel without any foundation trenches, whose fill might have provided useful dating information. Furthermore, the material evidence that was gathered during the excavations at the site is rather homogeneous, not allowing us to distinguish among the different construction phases more precisely.
The numismatic evidence plays a valuable role in the establishment of the general chronology of ʿAin el-Gedida. Indeed, the coins retrieved in several contexts within the church complex and throughout all areas investigated on mound I, especially along the streets running to the east and south of rooms B5 and A46 (i.e., the church and the gathering hall) have allowed us to gather a considerable amount of information on the chronological framework of the settlement. Nonetheless, some limitations exist on the use of this numismatic evidence. As emphasized by G. Bowen in relation to the site of Kellis, one cannot rely exclusively upon coins found at any archaeological site to reconstruct its chronology.4 Indeed, a very large number of specimens retrieved at ʿAin el-Gedida come from contexts that are unreliable or of dubious reliability, although it is worthy of note that the chronological distribution of these does not substantially differ from the coins found in more secure contexts.5 Another limitation is caused by the relatively high percentage of coins whose state of preservation was so poor as to forbid any secure identification and/or dating. A rather broad chronological range could be assigned to several unreadable specimens, on the basis of their dimensions and weight, although corrosion caused, in many instances, substantial loss of thickness and weight.
At any rate, the overall pattern of coin loss at ʿAin el-Gedida testifies to an occupational period that undoubtedly covered the fourth century CE. The small number of coins dated to the second half of the previous century, found in highly disturbed contexts in the proximity of the church’s apse, are not necessarily evidence of an earlier construction phase of the rooms predating the complex (and incorporated into it).6 In fact, older issues could still be in circulation, or kept in hoards by private citizens, long after they had been withdrawn from the official pool, particularly following the currency reform of 296 carried out by Diocletian.7 Furthermore, it should be noted that the numismatic evidence from the church complex (matching that from the other excavated areas), fits with that from the Small East Church at Kellis, which is also typologically very similar to the church of ʿAin el-Gedida.8 Nevertheless, some degree of caution is in order, at least with regard to ʿAin el-Gedida, since, as already mentioned above, a significant number of coins found inside the church complex (and in the surrounding rooms) are either unreadable or only broadly datable to the fourth century CE.
The information gathered from the analysis of coins from ʿAin el-Gedida suggests that the site ceased to be inhabited sometime towards the end of the fourth century. There are several coins that have been assigned by D. Ratzan to the second half of the fourth century; most of them are datable up to the third quarter of the fourth century, with only a few specimens dated to the very end of that century, or more broadly assigned to the late fourth–fifth century.9 However, there is a degree of uncertainty in the readings of at least a few of these specimens; therefore, in the absence of more secure identifications, these cannot be used to prove a longer life-span, with regard to the whole settlement, extending beyond the late fourth century.
Ceramics are another highly significant category of evidence for the establishment of the site’s chronology. The analysis and classification of pottery finds was started by Gillian Pyke and carried to completion by Delphine Dixneuf.10 A catalogue of forms and fabrics from the site was created on the basis of the repertoire from the nearby site of Kellis, with which the material from ʿAin el-Gedida is very consistent.11 The pottery assemblages from Kellis were extensively studied by C. Hope and A. Dunsmore, whose work has become a standard reference source for the study of Late Roman pottery in Dakhla.12 Strong parallels, with regard to fourth-century types and materials, can be established with the evidence from several other sites in Dakhla, including Amheida, and also from Kharga.13
The range of the ceramic repertoire from the church complex and its neighboring rooms (B10–B15) is consistent with the evidence gathered in the northern half of mound I, in particular within rooms B1–B4, where two significant ceramic deposits were unearthed.14 Indeed, the complete vessels and the diagnostic sherds found in the northern and central part of mound I, particularly in the two assemblages from rooms B1 and B4, include many of the types that were collected in area A, i.e., in the southern part of the hill, during the 2006–2007 survey.15 Among them are jars, flasks, cooking-pots, craters, plates, and several bowls, often painted with white slip on the outer surface and red dots along the rim.
Overall, the ceramic repertoire from ʿAin el-Gedida is rather limited in its number of fabrics, with a large majority of polished ware, made with iron- or calcium-rich clay and mostly of local production, while imports from the Kharga Oasis and the Nile Valley are rare.16 The treatment of the surfaces is generally poor and their decoration is simple and quite standard, often limited, as mentioned above, to red dots painted along the rim. There is also a prevalence of small objects, such as cups, bowls, and plates, over larger containers, although the latter are represented within most units. Although they were found in different contexts and rooms, all these vessels, forming the corpus of ʿAin el-Gedida, share a character that is consistent with a domestic assemblage and a fairly poor rural environment.17
The complete vessels and the large quantity of pottery sherds largely confirm the chronological range established by the numismatic analysis, that is to say, an occupational phase extending to the third quarter of the fourth century CE.18 Limited evidence of Early Roman forms was found in one room (kitchen B10), besides some third-century pieces likely used as chinking sherds for the construction of vaults. The Early Roman vessels from room B10 and the third-century coins from the church allow for the possibility that the site might have gone through occupational phases dating back to the third century or even earlier. It is true that the coins come from unreliable units and the Early Roman pottery was found in contexts mixed with fourth century material. Furthermore, no ceramic or numismatic evidence earlier than the fourth century CE was found in any other excavated room. However, it is unlikely that the pottery sherds datable to the third century came from another site or that third-century vessels were still in use during the fourth century. Therefore, they are valid indicators of earlier phases of occupation at the site.
Providing similar information to the pottery and coins are the ceramic lamps uncovered in several rooms of mound I. They can be generally dated to the Late Roman/Byzantine period and are commonly found in other fourth and fifth-century contexts in Dakhla.19
Another type of material evidence that is used to date archaeological deposits consists of ostraka. Twelve of them were found at ʿAin el-Gedida, two in Coptic and the rest in Greek. Roger Bagnall and Dorota Dzierzbicka dated the creation of eleven of them to the fourth-century, largely on the basis of content and palaeography. One ostrakon was dated, instead, to the third century based on content.20 Those with more precise information come from after 350; some have connections to Kellis documents from the period 350–370. This analysis confirms and supplements the data provided by the other categories of material evidence listed above.
Aside from ceramics (including ostraka and lamps) and coins, other categories of small finds collected at ʿAin el-Gedida do not provide significant information that can be used to establish (or confirm beyond doubt) the site’s chronology.21 They can hardly be assigned to a specific time frame, and only their association with otherwise datable contexts allows their tentative dating. Although of a limited chronological value, these objects offer valuable data for the study of material culture at ʿAin el-Gedida and at other sites in the oasis. Indeed, the typology of the evidence from ʿAin el-Gedida is consistent with the range found at several other sites in Dakhla, such as nearby Kellis, both from domestic and public contexts, and also Amheida, in particular from the excavation of large domestic units.22
A puzzling question, directly related to the issue of chronology of ʿAin el-Gedida, concerns the abandonment of the site. The archaeological record, gathered during the 2006–2008 excavation seasons and the survey of the structures investigated in the mid-1990s, has not provided, thus far, any evidence suggesting episodes of violent destruction, which might have led the inhabitants to leave the site abruptly. Indeed, no clue pointing to extensive fires was detected in any of the excavated rooms, either on their walls or floors or in their stratigraphy. The numerous layers consisting of vault and/or wall collapses seem to have formed at various times and resulted from natural factors, such as prolonged exposure to the elements after their abandonment, rather than human action. Deposits of ash, charcoal, and smoke on the walls were identified in several contexts. However, these were all related to cooking activities and indeed were found, for the most part, in kitchens, domestic middens, and rooms with hearths (such as anteroom B6 in the church complex and courtyard B1 in the partially excavated unit in the northern half of mound I). Another piece of information allows us to assume that the abandonment of the site did not occur abruptly but was carefully planned. This was mentioned in the discussion of the archaeological evidence from the church complex, but it applies to the entire area that was the object of investigation at ʿAin el-Gedida. Indeed, no objects of significant value were found in the rooms that were excavated. Overall, very few complete items were retrieved, and most of the material evidence consists of fragmentary objects, such as pieces of textile, bracelets, and ceramics. A few vessels were found in good condition, such as those unearthed in courtyard B1, but they must have not held a considerable value in antiquity, as they included cooking vessels for everyday use, which could be easily replaced once broken. The only finds from ʿAin el-Gedida with a monetary value (in the fourth century) are coins. The specimens gathered during the excavations had not been hidden in hoards; rather, they were found scattered within the stratigraphy of each room. Most likely, the coins collected on the streets leading to the church complex had been accidentally lost by people passing by. At any rate, none of the coins was of a precious metal, and their value individually was very low.
In general, the archaeological record suggests that the buildings of ʿAin el-Gedida, at least those investigated on mound I, had been emptied of any valuable object by their owners. Although it is not impossible, it seems unlikely that they were pillaged after their abandonment. If the whole site had been in fact the target of looters, the latter must have acted in antiquity, before the ancient deposits of wall and vault collapse and the windblown sand filled the rooms. Indeed, even the earliest stratigraphical layers, including the occupational contexts at floor level, did not include precious items, or many items at all.
Possibly, the abandonment of the site was not the outcome of a sudden incident of unknown nature, but was a planned event, taking place in a restricted time span, but not so short that the villagers could not sort their possessions and take with them anything they wanted before leaving.
Similar circumstances apply to other archaeological sites throughout the oasis. Particularly relevant in this context, due to its proximity to ʿAin el-Gedida, is the ancient settlement of Kellis (modern-day Ismant el-Kharab). Its excavators did not find any significant evidence pointing to violent destruction, such as extensive fires, which might have caused the abandonment of the large village toward the end of the fourth century CE.23 It seems, however, that what happened at ʿAin el-Gedida, that is to say, the more-or-less simultaneous abandonment of the entire site by its inhabitants was, in fact, part of a phenomenon involving large parts of the region.24 The scholars working in the area share the same concern about the necessity of shedding light on the possible causes. The discussion has focused on several key issues, such as climate change, economic depression, or political unrest. A phenomenon of increasing “ruralization” was detected, during the fourth century, at several sites of the region, such as Douch in the Kharga Oasis. The archaeologists who worked at this large and once prosperous village recognized traces of this trend in the partition and re-use of earlier buildings as stables. These alterations were combined with the construction of loculi, or low rectangular features at the corner or along the walls of these rooms, likely to feed animals that were employed in agricultural activities.25 Abundant organic material was found that demonstrated the existence of numerous animals inside older houses turned into stables. The archaeological evidence assigns these changes to the latest occupational phase of the site, not long before its abandonment, which was complete by the fifth century. At Kellis, too, some larger buildings were turned, in the site’s later phases, into masses of small rooms and stables.26 This phenomenon is attested also at Trimiths/Amheida, where a school was transformed, around the mid-fourth century, into a stable that became part of the so-called House of Serenos.27
At first glance, one could imagine a similar scenario, at least in part, with regard to ʿAin el-Gedida. Indeed, fairly abundant amounts of organic material, including coprolites and straw, were found in some parts of the site, especially along the passageway to the south of the church complex. Also, two features very closely resembling the loculi of Douch were discovered along the south side of courtyard B13, at the intersection of street B12 with vaulted passageway B11. The fourth-century range of the evidence from Douch also fits the record from ʿAin el-Gedida. However, the relatively limited area that was the object of investigation at the latter site prevents us from drawing conclusions on a generalized phenomenon of ruralization, which would have occurred on all mounds toward the end of their occupational life. Indeed, such a trend, although deserving investigation, is not supported by enough data at the moment. Also, if ʿAin el-Gedida had been, in fact, an agricultural site, it would be difficult to apply the concept of “ruralization” to it and recognize its traces. Animals must have always been a ubiquitous presence at ʿAin el-Gedida, as well as throughout the oasis.
At any rate, notwithstanding economic and social changes possibly taking place at ʿAin el-Gedida during the fourth century, the reasons that led all its inhabitants, as well as those of other sites like Kellis, to abandon their houses and move somewhere else (where, it is not known) have not found a fully satisfying answer thus far.
7.2 Monastery or Epoikion? The Monastic Connection and Alternative Readings
Since the Egyptian excavations, which were carried out in the mid-1990s, started uncovering archaeological features in the southern half of mound I, the possibility that ʿAin el-Gedida might have been a monastic settlement was raised.28 Scholars visiting the site at that time noticed a highly peculiar layout of the excavated structures, clustered around a large kitchen with several ovens, and consisting of an intricate network of interconnected spaces, several of which had certainly been in use as storage rooms.29 Indeed, it is not possible to recognize, within the complex spatial configuration of area A, separate buildings showing the distinctive characters of domestic units. The only exception is represented by some rooms located at the southeast end of the hill (rooms A35–A37 and A38–A40). They seem to have been built as separate clusters of at least three rooms each, with the biggest one likely serving as a rectangular courtyard. Unfortunately, this sector was only partially excavated, and the data that are currently available are very limited, not allowing us to draw significant conclusions on the nature of these spaces. Nor, in any case, do the visible layouts resemble in any significant way those of other domestic structures known from this period in the oases.
As previously seen, the survey of the southern part of mound I revealed different construction phases and evidence of restoration and/or re-use of earlier features in later structures. This part of the settlement, more extensively excavated than its northern half, seems to have developed from a central core of buildings, to which other rooms were progressively added, lying against or incorporating the outer walls of the earlier structures. The rooms identified along the south, southwest, and southeast edges of mound I were built with a very poor technique and did not follow any systematic plan. Their hasty construction history suggests, as argued in K. Bayoumi’s report, that these structures were built to satisfy a rapidly increasing population at the site.30 However, no significant information is available to identify who these people were, and the lack of excavation records of area A is particularly unfortunate in this respect.
The existence of a large kitchen, centrally placed and connected with the northern/central half of the mound, and of several magazines, containing clay bins for the storage of cereals or other crops, assigns the whole area mostly utilitarian functions. The absence, once again, of clearly recognizable domestic units and the overall spatial configuration, with most rooms built against and interconnected with each other, points to a social structure based on communal living rather than separate family households.
The discovery of the church (room B5) in 2006 and the investigation of the annexed complex between 2007 and 2008 provided additional data, which shed light on the highly Christianized society living at the site in the fourth century. The complex, capable of hosting a considerable number of people at any given time, was centrally placed on the main hill of the site and, one can assume, easily reachable from all other mounds (if they were in fact discrete units in that period). The excavation of the area to the east and south of the church confirmed the strategic location of the complex along a north–south axis, which connected the two ends of mound I. Furthermore, the network of streets lining the complex allowed for easy access to the latter from area A to the south and, in particular, from its large kitchen through a north–south oriented passageway. An additional bakery (room B15), also bearing evidence of several ovens, was found in 2008 right across the street from the entrance into the complex. The presence of two wide kitchens, undoubtedly not belonging to private, domestic contexts but rather used to answer the needs of a significant number of people; their proximity to the complex and, in particular, to the large gathering hall, capable of seating a considerable number of guests: these elements are, in fact, suggestive of a spatial arrangement entailing the existence, on site, of a large community of people not organized in the manner of a family, whether nuclear or extended. The considerable size of the hall and the existence of mastabas, or benches, that could seat a large number of people at any given time only emphasize the possibility that many individuals could access the church complex together. It does not provide information on who these people were or where they came from. On one hand, evidence of large halls opening directly onto churches is abundant in Egyptian monastic contexts.31 On the other hand, similar spatial arrangements are attested at sites that have not been identified as monastic, like Kellis and its Small East Church. The lack of incontrovertible data on the function carried out by room A46 (which might be identified either as a hall for catechumens, a refectory, or a space for laity) and the shifting, at least to some degree, of its purpose, as pointed to by architectural alterations that occurred before the abandonment of the site, leave the question open. Overall, the presence of the church complex on mound I is not, per se, evidence for a monastic identification of the site, nor is its proximity to large kitchens and the unusual layout of the south half of the hill.
Neither the Egyptian nor the 2006–2008 excavations uncovered any sizable structures at ʿAin el-Gedida that could be identified as large dormitories or houses for a monastic community, nor buildings divided into a series of cells, as were found at monastic sites throughout Egypt. At the same time, no comprehensive data are available on domestic arrangements for family units at ʿAin el-Gedida. A few buildings surveyed on mound I, particularly near the northwest or southeast ends of the hill, and also excavated rooms B1–B3 (plus the two unexcavated rooms to the north of B1) are characterized by a compact, self-contained layout of small rooms opening onto a larger room, possibly functioning as a courtyard, which may suggest their identification as residential units for relatively small groups of people. However, the lack of documentation does not allow us to know beyond doubt if these spaces once belonged to private dwellings or served a different function. The paucity of remains of identifiably domestic architecture found thus far at the site could be explained by the relatively limited area that was the object of archaeological investigation, compared to the overall extent of the ancient settlement. Indeed, it is not to be excluded that most people resided on the other mounds, while the main hill hosted, although not exclusively, buildings of a more communal nature, such as the church complex, or small-scale industrial installations, including the large kitchens/bakeries, the storage rooms, and even a ceramic workshop at the west edge of the hill.
If any monks lived at the site in the fourth century, they might have used or re-used forms that were well-established and common in Egyptian domestic architecture, in light of the lack of standardized types for monastic architecture at such an early time. It is possible that early Egyptian ascetics lived in dwellings that would have been indistinguishable from those of the lay inhabitants,32 who would have shared, if Christian, the same church complex as their ascetic fellow villagers. The possibility that these hypothetical early apotactics (“renouncers”) might have adopted standard domestic arrangements, instead of living in buildings with a larger and more complex spatial configuration, might be related, in some instances, to the re-use, by the members of newly formed communities, of older, abandoned structures in villages or cities. According to his Life, the first community founded by Pachomius was in the abandoned village of Tabennesi in the Nile Valley, not deep in the desert.33 As J. Goehring points out, how “deserted” this village was remains unclear.34 At any rate, the other monasteries that came to be part of the Pachomian koinonia seem to have followed a similar pattern, although there is not enough archaeological or textual evidence to allow their precise identification: that is to say, if they were all located within or near villages of the fertile band of the Nile Valley or if any of them lay on the fringes of the desert.35 Surely, some of the reasons for such a proximity to the Nile had to do with easy access to water, markets, and transportation.36 With regard to the oases of the Western Desert, hundreds of miles away from the Nile Valley, communities (of any kind, not necessarily monastic) had to rely upon the availability of water from natural springs or wells dug deep into the ground. The site of ʿAin el-Gedida, with its ease of access to water and its proximity to a considerably bigger town such as Kellis, was an ideal location for any kind of settlement, including a monastery. One must acknowledge, however, that most of these arguments simply leave the possibility open that monks settled at ʿAin el-Gedida, rather than providing evidence for their presence on site at any given time.
The material evidence does not provide significant information in this regard, either. The ostraka that were collected during the excavations are not particularly enlightening on this issue, focused as they are for the most part on economic matters. One of them, though, is worthy of mention here. It was found in 2006, during the excavation of a midden filling room B4, located in the western half of mound I.37 It is a complete ostrakon, made of a ceramic body sherd and of irregular shape, measuring approximately 10.3 by 9.8 cm. Nine lines of Coptic, parallel to the wheel marks, are written with black ink on its convex surface and were dated by Roger Bagnall and Dorota Dzierzbicka to the late fourth century CE, on the basis of palaeographic evidence and context. The ostrakon is a letter written by a certain Apa Alexandros, mentioned at the beginning of the first line, to another man concerning a third man (called Pesente) and his pakton (agreement or rent) and also naming a fourth man, i.e., Apa Kyros. Apart from its utilitarian content, this object is significant as the term apa, often used in a monastic context, might be suggestive of Alexandros and Kyros’ identification as monks. As stated by Bagnall and Dzierzbicka in their analysis of the ostrakon, although the term apa was also used for members of the regular clergy, the late fourth-century dating suggests that a monastic context is more probable.38 At any rate, the significance of this ostrakon in the discussion on the possibly monastic nature of ʿAin el-Gedida is rather limited. The ostrakon was found in a domestic dump, within a deeply unreliable context, and we cannot be sure whether it was written at ʿAin el-Gedida, sent to someone there, or neither. No secure information exists on the identity of either Alexandros or Kyros or ether of the two other men mentioned in the text. Also, it is not possible to know, on the basis of the available data, who among the four men, if any, actually resided at ʿAin el-Gedida. The natural assumption would be that Alexandros lived somewhere else. Therefore, the ostrakon, although a suggestive piece of documentary evidence, cannot be used as proof of the existence of monks at the site in the fourth century. On the other hand, it is worth observing that an ostrakon from Kellis (O. Kell. 121) mentions in an account both an Alexandros and another man described as a monk (monachos, misspelled monochos).39
Other material evidence, more closely associated with the church complex, and this time undoubtedly in situ, consists of the graffiti drawn with black charcoal on the west and north walls of anteroom B6.40 As mentioned in a previous chapter, they include a Greek inscription—a commonplace invocation to God—on the west wall and a Coptic one above several drawings on the north wall. The latter consist of a bird and three boats, motifs that can be found in Christian contexts at several sites in Egypt, including the Dakhla Oasis.41 However, with particular reference to boats, it must be noted that their representation in graffiti is not evidence per se of a Christian environment. In fact, drawings of boats can be seen at several sites throughout the ancient world also in pre-Christian times.42 Nonetheless, at least the Greek inscription on the west wall can be safely assigned to a Christian environment. At the same time, the contexts in which this typical formula is found do not always pertain to monasteries. Therefore, the graffiti of room B6, as well as the Coptic ostrakon from room B4, can hardly be used as evidence for the identification of ʿAin el-Gedida as a monastic site. It is true that the bulk of the Coptic texts from the fourth century found so far come from monastic communities, but Coptic was also in use in non-monastic contexts at Kellis: therefore, language is also not a decisive criterion.
On the whole, the available archaeological and documentary data do not point to an identification of ʿAin el-Gedida as a monastic settlement, built anew or developed on the remains of an earlier, non-monastic site. But it is fair to add that we lack indisputably monastic sites of the same period to which we could compare it. One could instead suggest, based on the current state of research, that the site was in fact a fourth-century settlement with an economy based mainly on the agricultural exploitation of the surrounding fields.43 Evidence for Late Antique villages is not very abundant from the point of view of their archaeological investigation. Indeed, a considerable amount of data comes from the excavation of sites located mostly in the Fayyum. The village of Karanis is, in particular, a copious source of information about Egyptian villages up to Late Antiquity. Peasant settlements were less the object of investigation in other regions of Egypt, although a renewed interest in domestic architecture is slowly changing this balance.44 On the other hand, documentary evidence abounds on fourth-century villages, shedding light on their economy, society, daily life, and their ties with the rest of the country, especially larger towns and cities.45 R. Bagnall effectively analyzed the many facets of life in Egyptian villages of the fourth century, based on the information provided by written sources such as documents written on ostraka or papyrus, especially the archives of people involved in the management of land.46
Indeed, the picture that emerges from the documentary and, in part, the archaeological evidence is that of a dynamic world, deeply engaged in the economic, social, political affairs of the time, but also involved in religious matters. The link between asceticism and Egyptian villages, mentioned above, is only one example of the not-so-secondary role that the latter held in the overall organization of the country.
Sources, both archaeological and documentary, testify to a large number of villages spread throughout Egypt in the fourth century, with estimates of their numbers ranging between 2,000 and 2,500.47 The size could vary considerably, but most of the rural settlements seem to have been smaller than 80 ha, which is the area calculated for Karanis, the most thoroughly investigated village.48 An average of more than 1,200 people lived in these communities according to Bagnall, who emphasized how the differences could in fact be substantial.49 It is impossible to make an estimate of the inhabitants of ʿAin el-Gedida in the fourth century, due to the lack of documentary or archaeological data on the full size of the settlement and especially because the residential area (or areas) have not yet been identified and excavated.
It is attested that villages had small industrial areas functionally related to agricultural activities, which played a primary role in the economy of Egyptian rural settlements.50 One could usually find among the various buildings in a village granaries, pigeon houses, bakeries, and spaces to manufacture objects of daily use, for example pottery workshops, all located within a usually irregular spatial arrangement.51 With regard to the archaeological evidence for ʿAin el-Gedida, all these features have been identified, although some of them only tentatively, on mound I: i.e., two large rooms with several bread ovens (rooms A6 and B15), a pigeon tower (the large rectangular room at the north end of the main hill), granaries (such as rooms A2–A4 in the southwest part), and a pottery workshop (the complex of rooms B17–B24 along the west edge of mound I).52 Also, the spatial configuration of this area is noticeably irregular, mirroring a seemingly common standard of Egyptian rural settlements.53 Furthermore, most of the ostraka gathered at ʿAin el-Gedida concern matters that reveal a world deeply linked to the agricultural exploitation of the land, a feature that is also distinctive, as just mentioned above, of a village-based economy. On the other hand, the fact that small-scale industrial installations, so typical of village life according to the sources, are present at the site does not necessarily confirm the identification of ʿAin el-Gedida as a standard rural settlement. Indeed, the archaeological evidence is not so abundant to allow comparative analysis to find what a “standard” village might have looked like. Furthermore, spaces destined to activities such as bread-baking or pottery-making were not exclusive features of villages, but could be found associated with other types of settlements, including those of a monastic nature. At any rate, the similarities shared by the archaeological data from ʿAin el-Gedida and the documentary sources about Egyptian villages deserve serious investigation, although it is worth pointing out again a glaring difference, namely that all village sites have readily identifiable houses, which is not the case at ʿAin el-Gedida, at least within the excavated and surveyed areas.
A strong female component was identified with regard to ʿAin el-Gedida, largely on the basis of material evidence, for example numerous fragments of bracelets found in almost all excavated contexts. Women were very visible in Egyptian villages of the fourth century, in some instances playing a considerable role in the social and economic life of that period.54 This is another point of contact between what the documentary sources say about rural communities in Late Antique Egypt and what the archaeological evidence from ʿAin el-Gedida allows us to decipher about its society. However, not even a sizable female presence on site can be used to prove beyond doubt that ʿAin el-Gedida was an ordinary village, consisting of family units living in the proximity of a central industrial and religious center that was mound I.
In fact, rural settlements in fourth-century Egypt were not all villages. An alternative type consists of epoikia: that is to say, small rural centers associated with the management of large agricultural estates and often dependent on larger villages nearby.55 The work-force could be employed, sometimes on a seasonal basis, to work the land under the direction of overseers, but tenancy could also be used to help supply the necessary labor. It is possible to suppose that the workers moved to the estate and lived there for the duration of their contract. The spatial arrangement of these epoikia is unknown, because none has ever been identified and excavated. On the basis of documentary evidence (P.Mich. 11.620), D. Rathbone reconstructed the possible appearance of an epoikion, which consisted of a complex of buildings functionally associated with the agricultural activities carried out in the farmstead.56 It seems that Egyptian epoikia were created either as isolated entities, later developing in some cases into regular villages, or they were integrated, since their very beginnings, into pre-existing villages.57 A fully communal life-style should not be necessarily implied for the people involved within the system of the epoikia. In fact, it cannot be ruled out that the wage-workers moved to these rural settlements with their families, occupying houses that may well not have differed substantially from those found in other types of settlements. At ʿAin el-Gedida, the south half of mound I might reflect the spatial arrangement of part of an epoikion, consisting not of its residential area but rather of a sector where the buildings more closely associated with agricultural activities were concentrated, including installations, such as bakeries, built to satisfy the needs of a relatively large community. The ceramic workshop, located along the western edge of the hill, would also fit within the context of a farmstead.58 The existence of a church at the center of mound I, largely consisting of public spaces of a utilitarian nature, is not surprising in association with an epoikion. Indeed, written sources attest to the possibility that churches were associated with this type of rural settlement.59 An ostrakon (inv. 830) found at the site acknowledges the payment of money by someone described as ἀπὸ γεωργ(ίου) Πμουν Βερρι, “from the georgion of Pmoun Berri,” the latter being (we may suppose) the likely name of ʿAin el-Gedida in the third century.60 Here georgion should refer to a farmstead or agricultural settlement and, if indeed it refers to ʿAin el-Gedida, establish that as the basic nature of the place.61
Although not conclusively, the archaeological and documentary evidence points to a likely identification of ʿAin el-Gedida as an epoikion. Quite significantly, small rural hamlets of this type (i.e., depending on large agricultural estates) have been known in Egypt, until now, only from documentary sources. Thus, if ʿAin el-Gedida is to be interpreted as an epoikion, it may offer the first archaeological evidence for this type of settlement, as well as more broadly shedding new light on the study of rural economy in Late Antique Egypt.
In 2006, the discovery of a Greek ostrakon (inv. 9) from mound I opened the discussion about the nature of the site to another possible interpretation. The sherd, found during the excavation of the domestic dump in room B4, is a receipt, dated to the fourth century, for the payment of eight artabas of barley.62 It is significant in this context because it mentions that the amount was paid for the annona of mounted archers, who formed a military corps of the Roman army, on duty at Mothis, the capital of the oasis since the fourth century. Further information came from another Greek ostrakon (inv. 660) found in 2008.63 The sherd, broken into three pieces and incomplete, was found within an ash layer filling room B11, the vaulted passageway running along the south wall of the church. The ostrakon, which consists of three lines written in black ink, is datable to the mid-to-late fourth century, based on handwriting and context. The inscription records a certain Joseph as the signatory of a receipt for two artabas of wheat and includes the mention, in the genitive case, of a praepositus. According to Bagnall and Dzierzbicka, the term praepositus could refer to the governor of a rural district, a subdivision of a nome (the praepositus pagi), or, more likely in this case, the leader of the ala (military unit) stationed in Dakhla.
Archaeological and documentary evidence from Dakhla testifies to a number of military settlements in the oasis during Late Antiquity.64 In particular, archaeological investigations carried out, in recent years, at the site of El-Qasr, to the northwest of ʿAin el-Gedida, added new and significant data in this regard. Researchers found, under the remains of the Islamic settlement and partially incorporated into them, consistent traces of a Roman castrum, or fort, whose excavation began in recent years.65 Written sources attest that besides the castra, built in towns, villages, and other strategically relevant areas, the country was dotted with smaller military outposts depending on the main forts.66 Smaller military units were detached there, allowing military control over large areas.67 But no substantial archaeological evidence exists to support the identification of ʿAin el-Gedida as one of these outposts. The ostrakon mentioning the military praepositus might be originally from a context that is not directly associated with ʿAin el-Gedida or refer only to a visit. No traces of outer walls or any sort of fortifications, which one can assume might have existed to secure a military station, were detected during the 2006– 2008 excavations and survey of the entire area. The rectangular building located in the north half of mound I, measuring roughly 16 m north–south by 12 m east–west, probably was not a military fortification that could have been used for the defense of the outpost and as a vantage point to inspect the surrounding area. Indeed, as discussed above, it seems more likely that it was, in fact, a columbarium or pigeon tower, whose presence in villages of Late Antique Egypt is attested to by substantial evidence, including within the Dakhla Oasis.
The existence of a church would not go, per se, against the identification of ʿAin el-Gedida as a military settlement. Indeed, the Christian affiliation of Roman soldiers is attested to not only in Egypt, but also in other regions of the ancient Mediterranean world, even before the fourth century.68 However, the possibility that ʿAin el-Gedida was associated with installations of a military nature and function seems remote. Indeed, the irregular arrangement of most buildings on mound I strongly differs from the carefully planned layout of military camps. Furthermore, the mentioned lack of evidence of fortifications would be puzzling, as it is hard to believe that they would have completely disappeared.
Although not providing evidence that a military detachment resided permanently at ʿAin el-Gedida (indeed, one of the two ostraka mentioned earlier, inv. 9, states that the mounted archers were stationed at Mothis), the content of the two ostraka is, nonetheless, suggestive of, at least, some form of interaction between the army and individuals at ʿAin el-Gedida, if only in the payment of taxes.
7.3 Epilogue
The aim of the excavation project of ʿAin el-Gedida was, from its early conception, to carry out a scientifically rigorous and comprehensive investigation of the site, which had raised, since the Egyptian excavations of the mid-1990s, stimulating questions pertaining to the singularity of its layout and architectural features. The peculiar nature of this venture, which enjoyed the productive collaboration of Egyptian and foreign members and specialists, allowed a particularly intense and rich process of study and interpretation. Different hypotheses were shared and tested in the field, against all the available evidence, in order to try to provide suitable answers to all the issues at stake.
Indeed, three seasons of archaeological investigation at ʿAin el-Gedida provided a wealth of archaeological evidence that went beyond any expectation, shedding light on several aspects of life at the site in the fourth century. Highly exciting was the discovery of a ceramic workshop that had been built within the remains of an earlier temple, whose identification is fairly secure. Not only did this find generate new data on local ceramic-production processes; it also added significantly to our knowledge of the profound changes that the adoption of Christianity brought to the physical, social, and religious landscape of rural Egypt in Late Antiquity. Quite significantly in this respect, another highlight of the project was the discovery of the church and the complex of rooms associated with it. Not only did the church, along with the gathering hall to the north, prove to be one of the earliest examples of Christian architecture discovered thus far anywhere in Egypt. It also testified to an original layout, which partially borrowed standard forms but, at the same time, elaborated them in an uncharacteristic way, closely paralleled only by the Small East Church at the nearby site of Kellis. The study of its construction process led to the identification of different phases, which involved the reuse of earlier structures into the new building. It also revealed the alterations brought to the church complex after its completion, affecting not only its spatial arrangement but also the function of some rooms.
The discovery of the church complex raised considerable interest among scholars for other reasons, which extended beyond its early date and unusual plan. First, it provided significant additional evidence for the flourishing of Christianity in the Dakhla Oasis since at least the early part of the fourth century CE. Although geographically isolated and far from the Nile Valley, the oasis proved to be the location of vibrant communities, exposed to a variety of new ideas, and associated life-styles, that were dramatically changing the ancient world. The archaeological remains of ʿAin el-Gedida testify to a profoundly Christianized society, whose public life centered around the church complex, built in a key spot on the main hill of the site. The evidence from ʿAin el-Gedida supplements that from other settlements in the oasis and, in particular, from Kellis, with which, due to their proximity and similar chronology, strong economic and administrative links certainly existed. Both sites, which did not experience later phases of occupation than the late fourth century CE, have provided substantial information on the earliest known types adopted by Christian architecture, not only in the region of the Western Desert but, at least until now, throughout Egypt. Furthermore, the evidence from Kellis and ʿAin el-Gedida is significant because it allows the possibility of establishing comparative analysis between Christian archaeology in Egypt and the early evidence from other regions of the Mediterranean world. The value of this investigation lies in the opportunity of reconstructing how architectural forms and types travelled with people and were adopted, sometimes rearranged in peculiar ways, in provinces that were geographically distant from each other, although not so at other levels.
As discussed above, several questions about ʿAin el-Gedida, in primis about its nature, have yet to receive a conclusive and unequivocal answer, although the archaeological and documentary evidence seems to point more strongly toward an identification of the site as an epoikion. Also, issues of the origins and, at the opposite end of the chronological spectrum, of the abandonment of the site are still debated. Undoubtedly, the full archaeological investigation of the site, including the unexcavated sectors of mound I and the surrounding hills, would provide much needed information on the original extent of the settlement and its overall spatial configuration, especially on the location of the domestic quarters. Hoping for a full-scale and comprehensive investigation of the site of ʿAin el-Gedida is, particularly during this period of political instability for the region, close to wishful thinking. Nevertheless, the information that was gathered is far from being inconclusive. It testifies to a vibrant rural community that settled at ʿAin el-Gedida and was certainly very active in the fourth century. It had well adapted itself to the local environment, exploiting what the surrounding land had to offer and storing the products on site. The small industrial establishments investigated on mound I shed light on a society whose involvement in the local economy extended beyond working in the fields. People crafted pottery, raised pigeons, and baked bread in large open-air spaces. Most likely, other productive activities were carried out on site, whose evidence lies beneath the sand of the desert and waits to be discovered. Like their fellow neighbors at Kellis, these people were also a profoundly Christianized society. This is testified to, for the most part, by the church complex, strategically built at the center of the main hill and, undoubtedly, a preeminent landmark of the local physical environment. The existence of villagers still practicing paganism at the site cannot be ruled out. However, the fact that the temple of ʿAin el-Gedida, located near the church, had been turned into a ceramic workshop by (at the latest) the early fourth century suggests that, if pagans lived at the site at that time, certainly their numbers must no longer have been substantial. If the inhabitants of ʿAin el-Gedida, both Christians and pagans, were in fact villagers or wage-workers and tenants of an epoikion, and if they included ascetics living as a community in a rural environment, this is yet to be known beyond doubt.
7.4 Notes
Sometimes as a result of the excavation of test trenches throughout the mound.↩︎
See Section 6.7.↩︎
Bagnall 1993: 264. About a possible Ptolemaic date for some similar mud-brick temples in Dakhla, see the discussion in Section 6.7 of this volume, in particular pp. 265-67.↩︎
Which, in fact, reinforces the dating established for the site through the numismatic evidence. Also, the lack of later material in disturbed contexts, for example following the excavation of pits, suggests a relatively short time span for the occurrence of these ancient disturbances.↩︎
Although, as mentioned earlier on, this possibility cannot be ruled out.↩︎
This is significant, since Kellis seems to have been abandoned by about 400, and many contexts at Kellis have their last material from the 360s, suggesting that the last few decades of the century may have been vestigial. Indeed, according to R. Bagnall, only two consular dates in P.Kellis 1 are securely dated after 369, with the possible addition of a third one (personal communication, February 2011).↩︎
Particularly in room A25, whose disturbed floor context had been left partially unexcavated in the mid-1990s.↩︎
No evidence was found of imports from the Mediterranean region; see Dixneuf 2012: 469.↩︎
Although some of the ceramic types chronologically extend into the fifth c. CE.↩︎
See Chapter 10 of this volume: for the third-century ostrakon, cf. cat. no. 9 (pp. 524–26 in print edition).↩︎
See Chapter 11 of this volume.↩︎
For Kellis, see Bowen 2002 and Hope 2003. Concerning Amheida, see Boozer 2007; Boozer 2015 and the excavation reports available on-line at: https://isaw.nyu.edu/research/amheida/reports.↩︎
Bowen 2007: 260. On the numismatic evidence dating the abandonment of Kellis, see Bowen 2001: 63–64.↩︎
However, there is evidence that some sites of the oasis, such as Deir Abu Matta and El-Qasr, continued to be occupied after the end of the fourth century.↩︎
Hope 2002: 173, 186.↩︎
Bayoumi 1998: 57–62. The excavations also extended, although partially, to the central sector of the hill, leading to the discovery of the gathering hall (room A46); its connection with a larger complex, however, remained unnoticed at that time. See Brooks Hedstrom 2017 on the archaeology of Late Antique monasteries in Egypt. On the phenomenon of early Egyptian monasticism, see, among others, Wipszycka 2009.↩︎
Such as rooms A2–A4 in the southwest corner of mound I. On room A2, see Section 2.1.1 in this volume, particularly pp. 42–45, pls. 2.5–7 (Plate 2.5, Plate 2.6, Plate 2.7).↩︎
See the discussion in Section 5.1 of this volume, particularly pp. 209–10. On the issue of food consumption in a monastic environment, based on written sources, see Harlow and Smith 2001 and Layton 2002.↩︎
Making it particularly difficult to recognize the nature of any of these buildings as “monastic”.↩︎
Possibly it was only partially abandoned and the monks settled in the buildings that had been left empty by their previous owners: see Goehring 1999: 97. On the overall economic situation of Egyptian villages and the reasons for their possible abandonment, see Bagnall 1993: 144.↩︎
Goehring 1999: 108 believes the first possibility to be more plausible.↩︎
Goehring 1999: 95, on the bearing of efficient communication means among the communities of the Pachomian koinonia.↩︎
See Section 10.3 below (cat. no. 1).↩︎
Idem 512.↩︎
See Section 3.1.3 above and Sections 10.4-5 (Section 10.4, Section 10.5) in this volume.↩︎
Winlock 1936: 17–18, pls. IX–X.↩︎
See Section 10.5 below, in particular pp. 533–37.↩︎
On Egyptian villages of the Byzantine period, see Keenan 2007, which considers a large amount of relevant documentary evidence.↩︎
For information about archaeological work on Graeco-Roman Egypt in recent decades, see Bagnall 2001 and Bagnall and Davoli 2011. Cf. also p. 75, footnote 63 in this volume.↩︎
See Bagnall 1993: 110–47.↩︎
Idem.↩︎
Bagnall 1985: 291–96.↩︎
On agriculture in Egypt since the Pharaonic period, see Bowman and Rogan 1999, in particular pp. 139–216 on the Roman and Byzantine periods.↩︎
Bagnall 1993: 113. Cf. also the evidence from the accounts of the Apiones’ pronoetai (overseers) in Mazza 2001: 87–88.↩︎
For a recent bibliography on pottery production centers in Late Antiquity, see Putzeys 2007: 63–65. With regard specifically to Egypt, see Marchand 2014.↩︎
Bagnall 1993: 111, 114 on the irregular spatial articulation of Egyptian villages.↩︎
On women and their role in the economic and social life of Late Antique Egyptian villages, see Bagnall 1993: 130–33. See also Wilfong 1999: 117–49, and Wilfong 2007: 318 ff. On documentary evidence pertaining to women, see Bagnall and Cribiore 2006.↩︎
Konstantinidis 2015; Hickey 2012: 25–26; Banaji 2007: 11–12; Mazza 2001: 79; Bagnall 1993: 151; Lewuillon-Blume 1979. On landholding and its role within the economy of Late Antique Egypt, see Bagnall 1992; Banaji 1999 and Banaji 2007 (especially chapters 5 and 7); Hickey 2007.↩︎
Rathbone 1991: 22–43. There seem to be, however, significant differences between his reconstruction and the layout of the buildings on mound I at ʿAin el-Gedida. Similarities have been established between ancient epoikia and modern Egyptian ezab (sing. ezbeh): cf. in particular Lewuillon-Blume 1979: 185 and Grenfell and Hunt 1907: 356. On the ezbeh, see, among others, Lozach and Hug 1930: 156–60; Binder 1978: 106–8, 216–55; Springborg 1982: 90; Alleaume 1999: 331–35.↩︎
Idem 31.↩︎
Installations for the production of pottery, and associated with agricultural centers, were found at other Egyptian sites, such as at Marea (Bagnall and Rathbone 2004: 76). On the Polish excavations at Marea, see Szymańska and Babraj 2006 and Szymańska and Babraj 2007. On centers of ceramic production in ancient Egypt, see the recent survey by Marchand 2014.↩︎
See Section 10.3 below (cat. no. 9).↩︎
On the translation of the term and parallels, cf. in particular p. 526 below.↩︎
See Section 10.3 below (cat. no. 5).↩︎
See Section 10.3 below (cat. no. 8).↩︎
See Rémondon 1955 and Rémondon 1965; Schwartz et al. 1969: 1–26; Carrié 1974; Carrié 1977; Carrié 1986; Bagnall 1993: 174–75. Palme 2007 deals with the Roman government and army in Egypt during Late Antiquity. On Egyptian geography based on the Notitia Dignitatum, see Worp 1994.↩︎
On the presence of the army in Egyptian villages, see Aubert 1995.↩︎
One example is the Early Christian building recently found at Megiddo, in present-day Israel. There is solid evidence attesting that the domus ecclesiae was used by members of the Roman army since the second half of the third century CE: cf. Tepper 2006; Tepper and Di Segni 2006; Adams 2008.↩︎