1.1 Why Village Institutions in Roman Egypt?

THIS VOLUME IS THE FIRST to survey village institutions in Egypt during the first eight centuries AD, from the beginning of Roman rule to the early Arab period.¹ Village institutions, as defined in the next section, refer here to forms of collective organisation of or in rural communities, whether administrative, religious, social or economic, and collective facilities, practices and behaviour in those spheres.

Traditionally Graeco-Roman civilisation has been viewed, studied and presented as a civilisation centred on the city. However, the general consensus holds that the vast majority of the population lived in the countryside, and it is uncertain, because rarely studied, how far their society, economy and culture were rooted in rural communities rather than in the institutions of the urban centre of their state or region.² General studies of rural life in the Greek and Roman worlds have been mostly concerned with agricultural practices for the simple reason that this is what the literary and archaeological evidence most directly and richly illustrates. In recent years, however, there has been a shift of emphasis which has brought the countryside, broadly defined, more into the centre of scholarly debates on the ancient past, with a focus on the Mediterranean. The most notable attempt at ‘ruralising’ the history of the Mediterranean is to be found in The Corrupting Sea by Horden and Purcell which re-conceptualises the Mediterranean world as a diverse system of micro-ecologies.³ The authors argue that the criteria usually adopted to define towns and villages are ephemeral. They decline to consider cities as a distinct element and instead conceive them as ‘the ultimate microenvironments, taking anthropogene effects on the landscape to their extreme conclusion, while (p.2) remaining as fickle in their forms as any of the ecologies already encountered’.⁴ This hierophantic assertion, however, is not followed by any examination of the workings of communities, rural or other. In some areas of the Graeco-Roman world, as noted below, new archaeological approaches or exceptionally abundant epigraphic evidence are providing islands of information. In this volume we aim to bring the rich papyrological evidence for the villages of Roman to late antique Egypt into the debate.

First, however, we note the general contrast with the early Middle Ages, for which studies of rural communities are far more numerous. This reflects the availability of a wider array of written evidence, including lives of saints, laws and charters and, eventually, tenurial and estate records, in the context of a relative decline of the role of cities and urbanism. In Framing the Early Middle Ages, Wickham dedicated three substantial chapters to villages and peasantries, in which he investigated patterns of change and rupture emerging from a comparative study, focused on institutions, of west and east, in the late Roman and post-Roman worlds. The analysis of a number of case-studies led Wickham to argue for a higher level of institutional organisation in the villages of the eastern Mediterranean, as opposed to more informal activity in the western villages.⁵ Two recent volumes of Antiquité tardive have been devoted to papers on rural worlds in the east and west in late antiquity.⁶ Village studies of specific communities begin to proliferate in the later periods, such as Davies’ study of village society in ninth- and early tenth-century Brittany, based mainly on charters, and Le Roy Ladurie’s micro-history of village life in early fourteenth-century Montaillou, based on the records of a church inquisition.⁷

For the Graeco-Roman world, study of rural communities and their institutions was for long largely limited to the two ‘political’ cases of the demes of classical Attica, attested in many inscriptions, and the pagi and vici in Italy and the Roman west, attested in literary, legal and some epigraphic sources, although the lack of epigraphic evidence for the functioning of pagi means that little progress has been made except to doubt that they represent pre-Roman structures.⁸ Archaeological evidence, however, is being increasingly exploited to reveal the distribution and socio-economic status and life of non-civic communities, as has long been done for Roman Britain, followed by similar approaches for regions of Greece such as Laconia and now too Italy itself.⁹

(p.3) Asia Minor is the region of the Graeco-Roman world with the richest epigraphic evidence from rural communities, and has been a laboratory for the study of villages. One of the earliest examples is Mitchell’s study of Anatolia which, aiming to establish to what extent Roman rule had an impact on the region, dedicates an entire section to rural communities.¹⁰ Here, where the majority of the population lived in the countryside, villages (mainly called kômai or epoikia) constituted the basic unit of social life, while city territories only emerged with the arrival of the Romans. Mitchell argues for a wide diversity of rural settlement types, which displayed various degrees of independence from the nearby city and integration into the regional market, concluding, however, that villages and cities were ‘worlds apart’. A different view of the rural communities of Asia Minor has since been proposed by Schuler, who examines the long-term development of village institutions from the Hellenistic to the Roman period, highlighting elements of continuity and diversity, and concludes that rural communities, which displayed varying levels of complexity, were always modelled on the city.¹¹ The recent volume edited by Thonemann offers new insights into several aspects of rural society and institutions in one of the innermost and most rural areas of Asia Minor, Phrygia, in the Roman period.¹² The long-term perspective provided in the contribution by Thonemann suggests that, despite an increase in the culture of urbanisation under the Romans, settlements in Phrygia remained fundamentally ‘decentralised’. Villages appear to have been, from an institutional and administrative point of view, independent from the cities, with ‘very little sign of the emergence of strong centre-periphery relations either within or between Phrygian poleis’.¹³ Though adopting different approaches and coming to different conclusions, all three works have investigated the changing patterns of rural institutions in a long-term perspective, and have noted distinct impacts from Hellenistic and Roman rule.

It is at first sight curious that Graeco-Roman Egypt, the region of the Graeco-Roman world with – at least for some areas – the richest resource of written documentary evidence, mostly on papyrus, which would seem to make it the most obvious candidate for investigation of rural settlements in a long perspective, has so far been little touched by this new interest. The place of Egypt in the history of the Mediterranean, from a political and socio-economic as well as geophysical point of view, has long been debated. Bagnall notes that Hellenistic to Roman Egypt, as opposed to medieval Egypt, hardly appears in The Corrupting Sea; within the wider Mediterranean it ‘is part of the fuzzy edges’. This country lacked, Bagnall continues, one of the two requirements for inclusion in the Mediterranean as imagined by Horden and Purcell, namely environmental variation, and most of it also (p.4) lacked the other requirement, which is maritime ‘connectivity’ with the Graeco-Roman world.¹⁴ Traditionally the alleged peculiarity of Egypt is grounded in its special status as a Roman province during the early empire. It was governed by an equestrian prefect, not by a senator, and its cities, including Alexandria, lacked the city council (boulê) and hence local self-government and social organisation typical of the Greek and Roman worlds until AD 200/1.¹⁵ The last few decades, however, have seen the emergence of a new scholarly trend which advocates the typicality of Egypt, or at least its non-exceptionality, within the eastern provinces of the Roman empire.¹⁶ Analysis of the papyrological evidence has revealed a complex administrative and socio-political system, which served well the needs and requirements of the central Roman administration. This can be seen in the imposition of the poll-tax (laographia), the distinguishing mark of Roman rule in the province, and in the establishment of civic statuses for taxation and administrative purposes. Even scholars who have championed a revived view of important continuities from Ptolemaic rule still treat Egypt as a part of the general transformation of the eastern Mediterranean from Hellenistic kingdoms to the Roman empire.¹⁷ Furthermore, the lack of city councils until the third century is no longer regarded as lack of institutions of self-administration, even if debate is continuing over the nature of the groupings and officials who provided it. Roman Egypt is gradually rejoining the Roman empire.

Despite the many studies of society and administration in Graeco-Roman Egypt – from Braunert on internal migration to Kruse on the royal scribe (basilikogrammateus)¹⁸ – which normally include some discussion of village evidence, there are no general studies of village institutions or communities in any one period, let alone in a long-term perspective, or integrated investigation of their relationship to the wider state. Scholars have instead followed the path of studies of one particular village, such as Kerkeosiris in the second century BC, Hawara in the Graeco-Roman period, and Jeme in late antiquity.¹⁹ This volume represents a first response to this gap in the current scholarship. We aim to demonstrate that Egypt is a particularly productive place to develop study of this subject because the rich documentary evidence of the papyri, of which roughly half, at a guess, come from village sites, permits us both to study specific topics in detail by place and time, as the eleven papers of the volume do, and also to make comparisons across a long chronological period. These comparisons across time are beneficial in two ways: they raise questions about changing patterns and perspectives of the (p.5) surviving documents, which may skew interpretation, and they enable us to outline what seem to emerge as recurrent issues, deserving more detailed scholarly attention, in the power-relationships between central and regional authorities and the rural population, as well as some preliminary indications of the trends in those developments across our period. Although one volume cannot offer a comprehensive or systematic study of all the institutions which, at different times and in different places, were found in the villages of Egypt, we think that it does push forward study of the subject by showing what can be done, and also highlights the value and importance of looking at these institutions in a long-period perspective.

1.2 Villages and Institutions

The title ‘village institutions’ begs two basic questions of definition: what do we mean by villages and what by institutions? The difficulty in answering the first question lies in the variable and slippery nature of the criteria which are adopted to define settlement types, commonly a combination of some of the following: population, size of territory, administrative functions, public buildings, quality of private housing, diversification of non-agricultural activities, cultural habits, level of literacy. It is often asserted or assumed that the village (kômê, also epoikion, katoikia and so on) was the most common type of rural settlement in the ancient eastern Mediterranean.²⁰ In the Hellenised east, villages have traditionally been defined in opposition to cities (poleis): a village was a settlement without civic status – it was not a self-governing community with its own laws, magistrates and coinage, controlling and exploiting its own territory, and ruled by a local landowning elite.²¹ However, it has also been long recognised that the apparently straightforward criterion of official status, and what self-administration a settlement possessed, is subject to anomalies.

Settlements could change status, and we also find villages which look and act like cities and village-like cities. In Syria and Asia Minor, for example, we have evidence of villages being promoted to poleis, sometimes through synoecism, as well as cities being reduced to village status.²² In Syria, again, and also some areas of Asia Minor such as Lydia, we find elaborate types of village organisation which in many ways resemble civic structures, with a high level of administrative independence and also a market economy with regional integration.²³ Thonemann notes that some villages in Phrygia had a popular assembly (ekklêsia), normally (p.6) thought typical of Greek cities, and display other urban practices such as honouring benefactors in inscriptions.²⁴ Indeed he proposes a cultural shift in the mid-first century BC, when civic institutions and ways to express communal identity spread through northern Phrygia and Anatolia, including the epigraphic habit, local coinages, and self-description as a polis.²⁵ Though these settlements were still formally labelled as villages (kômai) they were behaving as cities (poleis). In contrast, as we will see, while a few similar cases are known in Ptolemaic Egypt, in Roman Egypt a sharper distinction was made between villages and the ancient regional ‘capitals’ which became termed mêtropoleis in the Roman period, although they only gained formal civic status in AD 200/1.

The other temptingly straightforward criterion of population size is equally if not more unhelpful, not least because we have little data in the Graeco-Roman world for population size. A recent synthesis for the Roman empire in the mid-second century AD, based on site size multiplied by assumed average population densities, proposed that over 10 million urban dwellers lived in cities with 1,000 to 5,000 inhabitants and over 7 million in cities of more than 5,000 inhabitants. This, however, seems on the high side for large cities compared, for example, with an earlier guesstimate for Roman Italy, generally taken to have been one of the most densely populated areas of the empire, which put over 80 per cent of its urban dwellers in ‘minor’ cities of 1,000 to 5,000 inhabitants.²⁶ Cities and villages in Roman Egypt, as we will see, were larger than the norm; indeed villages such as Karanis and Tebtunis in the Arsinoite nome, with estimated populations of 4,000 people or more, were larger than many small cities in Greece or Italy.²⁷

Historians will never agree a universal definition of ‘village’.²⁸ Pragmatically by ‘villages’ in Roman and later Egypt we mean the rural settlements normally of 500 to 2,000 inhabitants, but with extreme parameters of 100 to 5,000 inhabitants, which were called kômai by the state and their inhabitants. In negative terms, they were not one of the few ‘Greek’ cities such as Alexandria and Naukratis, or a regional capital (mêtropolis) or a dependent hamlet (epoikion). The variations across place and time, such as in size, non-agricultural resources, importance of the local temple, elements of self-administration, socio-economic status, and so on, which made one village different from another, are precisely what we are starting to investigate in this volume.

(p.7) Our definition of institutions is also pragmatic. Although the term is widely used in the social sciences, there is no consensus in its definition.²⁹ We take as our starting point Douglass North’s broad definition of institutions as ‘humanly devised constraints that shape human interaction’, but limit the constraints to those which affected all inhabitants of a village or at least a group within a village wider than a family, and our concern in this volume, unlike that of North, is more with the social than the economic impact of these institutions.³⁰ In other words, we study institutions not to measure or assess economic performance but to identify types and trends of communal organisation and consciousness. More specifically the eleven contributions in this volume investigate collective forms of organisation and behaviour, including village officials and leaders and their roles, policing, monasteries, private associations, record-offices, banks, legal practice and festival culture. The most obvious omissions in this collection are the principal administrative and fiscal officials of the Roman period, such as the village scribe (kômogrammateus), tax collectors (praktores) and granary scribes (sitologoi), each of which could easily fill a monograph of its own. However the next section tries to compensate for this by locating the contributions of the papers in a preliminary survey of developments in village administration focused on the Arsinoite nome.

1.3 Village Institutions in Egypt: An Overview

In this section we try to locate the contribution of the papers in this volume (cited by the author’s surname and chapter number) in an overview of the development of village institutions in Roman to Fatimid Egypt. Two main and interconnected issues emerge: the independence of village institutions in relation to the regional cities and the state, and the ways in which, and the extent to which, these institutions reflected or encouraged communal self-identity, in a diachronic perspective with attention to the impact of changes. The reader will note a certain focus on the Arsinoite nome in Middle Egypt (the modern Fayyum). This is simply because it is the nome for which the most papyrological documentation by far has survived from village sites. Although the Arsinoite nome had peculiarities of hydrology, history and administration, some of which will be noted below, all nomes had minor local specifics, and the broad administrative developments in the Arsinoite were not untypical of Middle Egypt.³¹ How different Upper and Lower Egypt were (p.8) from Middle Egypt remains an enigma, but probably the differences were more in detail than substance.³²

When Egypt became a Roman province in 31–27 BC, under an equestrian prefect (governor) appointed by the emperor, it was still essentially a land without cities in the Graeco-Roman sense, with the exceptions of the archaic oddity of Naukratis and the Hellenistic foundations of Alexandria and its pale counterpart in Upper Egypt, Ptolemais Hermiou. Egypt was divided into over forty (eventually around fifty) administrative areas known as nomes, most of millennial antiquity, ruled through central officials in the main town of each nome. Although under the Ptolemies these nome capitals had been given Greek names including the term polis (‘city’), most of them were small towns, often dwarfed by their main temple. In the mid-third century BC, for instance, the chief city of the Arsinoite nome, Krokodilonpolis (‘Crocodile City’), later called Ptolemais Euergetis and then City of the Ptolemaeans, and also known as Arsinoe, had a population of around 4,000.³³ Conversely some villages with large numbers of Greek and other Hellenised settlers had Greek urban institutions such as an orthogonal street plan and a gymnasium.³⁴

The Romans imposed a clearer hierarchy derived from the Graeco-Roman city-territory model.³⁵ On an overarching level Alexandria was the one city and all Egypt its territory (chôra); only the Alexandrians (plus a few other civic communities) were citizens, while all the other inhabitants, of whatever ethnicity, were ‘Egyptians’ or laoi (the people). All laoi, unlike the citizens, were subject to the census and poll-tax (both called laographia). On a second level, however, the chief city of each nome was now termed its mêtropolis (mother-city), and its inhabitants, called mêtropolitai, enjoyed a privileged rate of poll-tax, normally half the rate for villagers of the nome. The mêtropoleis also became the nome centres of Graeco-Roman culture: within a few decades all village gymnasia had disappeared, and the members of the gymnasium of each mêtropolis were officially designated as a cultural elite, with strict qualifications for entry, normally called ‘those from the gymnasium’ (but in the Arsinoite called the katoikoi, ‘settlers’). The mêtropoleis, and indeed Alexandria, were not technically full cities in the Graeco-Roman sense until Septimius Severus gave them all a council (boulê) in AD 200/1, and also autonomous annual magistrates (civic officials). However, already in AD 54 the Arsinoite mêtropolis, calling itself ‘the polis (of the Ptolemaeans)’, sent an embassy to honour Nero on his accession, as they had apparently done in AD 41 (p.9) with Claudius, and to ask for privileges in return.³⁶ In contrast we know of no decrees made by villages, or by an internal group on their behalf, from Roman Egypt. One possible exception, also of the reign of Nero, is a vote of honours to the prefect Ti. Claudius Balbillus (c. AD 55–59) because he had restored the sphinx at Giza, by ‘the men from the village of Busiris in the Letopolite (nome) who live near the pyramids and the district- and village-scribes currently in the (nome)’.³⁷ However, this looks more like an assembly of the main village officials of the nome, so the ‘men’ of Busiris may just mean its officials, and it is anyway a unique case. Indeed the possession of its own officials, especially a ‘village scribe’, is what made a village a kômê, as distinct from an epoikion, a subordinate hamlet (although epoikia which had become kômai kept epoikion in their names). The fundamental point here is that the Augustan settlement of Egypt for the first time established a clear distinction of administrative status between the mêtropolis of each nome and the villages (kômai) with their hamlets (epoikia).

It has been estimated that at its peak in the mid-second century AD Roman Egypt had a total population of up to 5 million, including some 0.75 million in Alexandria and say 1.25 million in the mêtropoleis.³⁸ Roman Egypt was more densely inhabited and far more urbanised than early Ptolemaic Egypt: Arsinoe, for example, is estimated to have had 25,000 inhabitants in the mid-second century AD as compared to 4,000 in the mid-third century BC, and the villages of the nome too were considerably more populous. As a crude sighting shot, Roman Egypt probably had over 3,000 villages with an average population of around 1,000 inhabitants each. However, there were regional variations. The Oxyrhynchite nome is estimated to have had a territory of about 750 km² with around 100 villages; although quite large as a nome, its villages were mostly modest, as was probably typical. The Arsinoite, in contrast, was unusually large, with a settled area of some 1,200 km² and around 140 villages. Arsinoite villages show more variation in size, and several, such as Karanis, Philadelphia and Tebtunis, were extremely large with big territories and around 4,000 inhabitants each. We may expect this diversity in size to be reflected in diversity in administrative services and autonomy, range of economic activities and also cultural features such as social organisation and education.

By the mid-first century AD the Romans had adapted the pre-existing administrative structures to a quite different system which lasted into the later second (p.10) century and, in some respects, even longer. Gone was the dominance of nomes by powerful families holding secular posts such as stratêgos (the ‘general’, that is district officer of each nome) and priesthoods. The stratêgos of each nome was now an Alexandrian appointed by the governor, typically for three years, along with the lower-ranking but independent basilikogrammateus (royal scribe), also an outsider on a short tenure, supported by some other officials based in the mêtropolis, and there was no intermediate level between them and village-based officials.³⁹ There were few or no large private estates either in this period.⁴⁰ In this environment the village officials of various types flourished in power and, presumably, rewards.

Every village had a principal administrator called a village scribe (kômogrammateus), a police-chief (archephodos, ‘patrolman’) and a collector of cash-taxes, sometimes called poll-tax collector (praktôr argurikôn/laographias); some shadowy ‘place-chiefs’ (toparchai) also appear until the 70s. Larger villages might have one or more of the following: a record-office (grapheion) under a notary (nomographos), a public granary (thêsauros) under grain-scribes (sitologoi), or a centre for the local guards (phulakes) of various types, such as the headquarters of the desert-guards (erêmophulakes) known archaeologically at Tebtunis. Some were also the base for a seconded Roman soldier given the courtesy title of ‘centurion’ who supervised local order and administration as the direct representative, independent of the stratêgos, of the Roman military command of the province.⁴¹

All the village officials were appointed by the stratêgos or basilokogrammateus, but the limited evidence leaves their terms of service uncertain. In the first century AD most were probably voluntary contractors, but the police-chiefs and various guards may have been paid, even if the guards were villagers nominated to serve in ‘liturgic’ fashion (in effect temporary conscription). The reasons for taking Kronion, the notary of Tebtunis, to be a contractor, as discussed by Langellotti (Chapter 6), are emblematic: his length of service, backed by the fees he charged for his services and the payments (for his concession?) he made to the central nomographos of the nome. He is, in fact, only one of a number of Arsinoite village officials of the era with lengthy and even hereditary tenures in their home villages. Kronion was nomographos at Tebtunis from AD 26 to 56 (at least), following his father Apion from AD 7 to 26. Also at Tebtunis, Akousilaos was dioikêtês then toparchos from AD 16 to around 27, and the office of village scribe was dominated from around AD 16 into the late 40s by the family of Lysimachos.⁴² On the other side of the Arsinoite, Nemesion acted as cash-tax collector of Philadelphia from the late 30s AD until at least AD 56.

(p.11) Nemesion, as has been delightfully demonstrated from his letters and accounts, was one of a group of local officials who dominated his large village with the backing, when needed, of their friendly centurion.⁴³ Mascellari (Chapter 2) shows how such dominance by local officials was a regular part of the system of law and order, a situation which finds a contemporary parallel in policing in rural Roman Asia Minor.⁴⁴ Victims of crimes or injustice, as their petitions reveal, typically first sought the help of local officials to prepare the witnesses and evidence for their sufferings, and used their aid in submitting petitions. To arrest suspects for questioning, the stratêgos turned to the local police-chief, one of this same group. The various guards too were locals. Mascellari concludes that the initiative in providing local law and order came from local officials, and the role of the stratêgos and centurion of the area was essentially to support them. Nemesion again illustrates this independence. His horizons were wide: it was he who copied out the some what doctored text on papyrus we have of Claudius’ letter of AD 41 to Alexandria, which contains Claudius’ well-known pronouncements on emperor worship and the riots between Greeks and Jews; Nemesion and his fellow poll-tax collectors from five neighbouring villages had the confidence to circumvent the stratêgos and petition the prefect directly to have their obligations reduced, implicitly threatening that, if not, they would give up their contracts.⁴⁵ This strikingly self-confident independence of village officials seems to have been characteristic of the first century AD. Although in Tebtunis of the early second century, as Langellotti notes, Apollonios alias Lourios served as nomographos for over thirty years, by his day, as we will see, things were beginning to change.

As noted above, the inhabitants of villages in Roman Egypt never had any political identity or mechanism of communal decision-making. All villages, however, seem to have had a variety of associations (sunodoi), whose role in communal life has been a matter of some debate. Using the Arsinoite villages of Euhemeria and Soknopaiou Nesos as case-studies, Paganini (Chapter 3) surveys the nature and roles of village associations. They were private bodies, but with their own rules and regulations, registered memberships (normally men only) with an annual fee, and one or more annually elected ‘presidents’. They could take group decisions, such as to honour a benefactor. Some were named after professions, such as the associations of weavers or tenant-farmers of imperial land (but were not, it seems, compulsory or exclusive like guilds), some after deities and a few after their founders. Although all of them had a purely social function, often in a religious context, those with links to trades tended also to have an economic and fiscal aspect, and could act as a corporate broker between the state and their members in, for instance, the allocation of certain levies or liturgies.

(p.12) This is most evident in the associations of ‘public farmers’ (dêmosioi georgoi), and their elders (presbuteroi) and presidents (hêgoumenoi), which are discussed by Strassi (Chapter 4) and Kruse (Chapter 5), whose rather different conclusions illustrate how difficult it still is to interpret the scattered evidence. Both agree that associations of public farmers, that is tenants of state land of various types, are of Ptolemaic origin and are mostly attested in the Arsinoite nome, which had far more public land than other nomes. Indeed Kruse calls them ‘state’ farmers to stress their continuity despite the change in name from ‘royal’ farmers under the Ptolemies to ‘public’ farmers in the Roman period. Kruse argues that whereas the elders of the public farmers had a general role in village affairs in the Ptolemaic period, in the Roman period they were restricted to issues in the management of the land, and that the ‘elders of the village’, who emerge in the Roman period, were quite distinct. Strassi, in contrast, argues that the associations of public farmers, which originally just dealt with tenancy issues, had broadened into associations of all farmers by the third century AD, still with presidents and a social function, while their elders had in the first century AD morphed into the ‘elders of the village’ and acquired a much broader remit in village affairs. As Kruse and Strassi agree, the position of village elder became a liturgic post in the second century AD; hence, for instance, the elders had to act in place of the kômogrammateus if that office were temporarily empty. The general picture that emerges is of a rich network of village associations which provided their members with a social identity, practical support and some mediation with the state in fiscal affairs. However not even the associations of farmers ever acted or spoke for the village as a whole, and the state formalisation of their elders into liturgists seems to have been a top-down ploy to channel fiscal demands rather than an empowerment of communal representatives. How typical the Arsinoite was in this respect remains a question, but it fits the general trend in Roman imperial fiscality towards forms of collective liability, culminating in the reforms of Diocletian (see below).

Four of the papers below investigate socio-economic institutions in villages, again with an inevitable focus on the Arsinoite nome. Langellotti (Chapter 6) examines the impact of the record-office (grapheion) at Tebtunis through the extensive mid-first-century archive of documents of its then operator, Kronion. Village record-offices, which are attested in larger villages from the second century BC into the third century AD, provided a general scribal service for documents of all types and also notarial service for the registration of contracts, reporting to the central notarial office in the nome capital. Langellotti estimates that at Tebunis in AD 45/6, for which we have a complete register, over 50 per cent of the village’s families made use of the grapheion, with Egyptian names as prominent as Greek ones. This high use of written agreements, mostly concerning financial transactions, implies a broad penetration of functional literacy and engagement with the money economy. Nowak (Chapter 7) looks at the particular cases of the making of wills, both the Roman-style wills which anyone with Roman citizenship had to (p.13) use, which in villages meant mostly discharged soldiers, and local Hellenic-style wills (diathêkai). While on the one hand it is clear that villagers who wished to make formal written wills were able to do so, and sometimes within their village, most wills of both types were drawn up and opened in the mêtropolis because of procedural requirements: for Roman wills, the presence of the stratêgos (normally in a Caesareum) and seven citizen witnesses; for Hellenic wills, registration in the nome agoranomeion (another record-office). Nowak concludes that many villagers must instead have used other traditional unwritten means of distributing property such as marital agreements or gifts promised on death. A similar picture emerges from Lerouxel’s study (Chapter 8) of private banks in villages. Although he argues against the traditional view that banks were purely urban and used only by craftsmen and traders, village banks are not numerous. Whereas seventeen private banks are attested in second-century Ptolemais Euergetis, only two Arsinoite villages (and three more villages in other nomes) are known to have had private banks. But one of them, Dionysias, has four attested banks, two of them contemporary, perhaps because of its strategic location in the northwest corner of the nome and over 50 km distant from the mêtropolis.

Lastly, the nature and role of public festivals in villages, particularly at Soknopaiou Nesos, an outlying temple-village on the north side of Lake Moeris (the Birket Qarun), is examined by Jördens (Chapter 9). She first stresses the antiquity, number and length, and exuberance of the major festivals organised by temples, and their enormous popularity in drawing crowds of men and women from near and far. However, although temple festivals appear to have been flourishing through the first and second centuries AD, there were significant changes. First, reflecting the municipalisation of the nome capitals, lay associations and individuals became more prominent in their running, and changed their nature from the traditional Egyptian religious focus to one more of civic-style entertainment, as was also occurring in contemporary villages in Asia Minor. Second, the festivals seem to have been becoming more local, an expression of communal identity reinforced by the patronage of associations, and less attractive to outsiders. Hence, perhaps, the third-century abandonment of Soknopaiou Nesos whose isolated temple and village economy depended on outside visitors. One general pattern which seems to emerge is that the larger villages on the edge of the Fayyum and more distant from Ptolemais Euergetis tended to have more independent institutions; caution, however, is advisable because most Arsinoite villages lay within a 30-km radius of the mêtropolis, and we do not have papyri from the large villages closer to it.

An issue that recurs in several of these contributions is the hollowing out of village institutions which seems to have been part of the gradual ‘municipalisation’ of the metropoleis and hence their nomes, and perhaps also the growth of large private estates. Beginning in the early second century, but developing, it seems, unevenly, important village posts such as village scribe and cash-tax collector were turned (p.14) into liturgic appointments, that is, a service imposed by the state on men of a certain wealth.⁴⁶ As Strassi (Chapter 4) and Kruse (Chapter 5) agree, the originally private position of presbuteros (elder) of associations of farmers was also absorbed into a state liturgy with a broad, perhaps deliberately vague, responsibility for all village dues on land and for trying to resolve local disputes in general. Although elders and grain scribes could still be locals, village scribes and cash-tax collectors now had to come from another village, and seem to have served for a few years only, breaking the dominance of local families in favour of central control. Langellotti notes that the running of village record-offices became far more subordinated to control by officials in the mêtropolis; Reiter has elsewhere proposed a similar spread of liturgic inspectors (epitêrêtai) to oversee the collectors of indirect taxes.⁴⁷ Cultural municipalisation is also detected by Jördens (Chapter 9) in the changes in religious festivals, and the turning-in on themselves of most village festivals. The trend was accelerated by the creation of town councils in AD 200/1.⁴⁸ In 219/20 in the Arsinoite nome village scribes were replaced by municipal amphodokômo-grammateis (village scribes from urban districts) and around AD 240 local grain scribes gave way to dekaprôtoi (ten leading men, standard too in other provinces as decemprimi) who had a general responsibility for taxes on land. By the late 240s the amphodokômogrammateus had been replaced by two kômarchai (village leaders) in each village, who were once again locals, but now under stricter central control, and in the early fourth century a revived scribe of the village (grammateus kômês) was added. As part of Diocletian’s reforms, which tried to systemise administration throughout the empire, starting around 297/8 taxation was assessed on and levied from village inhabitants as a koinon (group) with collective responsibility. In 307/8 villages were aggregated into districts (pagi) for fiscal and general administrative purposes, replacing the previous shadowy toparchies, under the supervision of praepositi (supervisors, attested from 303) drawn from the town council, itself now more or less under the direct orders of central officials. Silently, too, the once vigorous private associations disappeared, presumably as part of the cultural changes which included the third-century decline of temple-based religion and the fourth-century spread of Christianity.

Berkes (Chapter 10) offers a broad overview of the nature of village administration in the period from Diocletian through the Arab conquest, that is the fourth to eighth centuries AD, illustrated and expanded by the papers of Schenke (Chapter 11) and Papaconstantinou (Chapter 12) which focus on the rural role of the church. For this period the provenance of our documentation changes to Oxyrhynchus and other towns and religious sites down the Nile to Thebes. The basis of rural administration remained the koinon, the fiscal community of the village with its collective (p.15) responsibility. It has been debated whether the koinon was in essence a top-down imposition to exploit the rural population or an expression of local identity and solidarity with which the authorities had to negotiate. Berkes shows that while villages typically had one or two headmen who served for a few years, there was an extraordinary variety across time and place in the titulature of village leaders. This he interprets as reflecting their independence from central control (hence no top-down system of offices) and their socio-economic dominance of their localities. He is, incidentally, sceptical that direct control of villages by large estates was common, although this is slightly nuanced by Schenke’s study of the monastery of Apa Apollos in the seventh to eighth century, which shows that monasteries could provide an alternative community, for shorter or longer stays, to lay village society. The lands and number of residents of Apa Apollos made it equivalent to a middling large village; it employed agricultural labourers, tenants, craftsmen and so on, and it contained a wide variety of communal and private buildings. More broadly, these three papers agree that the main social role of the church in the countryside of late antique Egypt was to provide help and relief to villagers oppressed by their local elites, implicitly in the absence of central government protection. Papaconstantinou, again focusing on the seventh to eighth centuries, argues that the traditional view that the church provided leadership in Egyptian villages from the fourth century on is the product of ecclesiastical writers of the post-Ummayad period justifying by back-projection their new role. She argues that the domination of villages by amorphous groups of local notables was rooted in their leasing of land and lending money to poorer villagers; Berkes would stress more their power as collectors of the collective fiscal dues, and suggests that the Abbasid reversion after AD 750 to individual taxation broke their power.

This collection of papers is inevitably exploratory. In aiming at a long chronological coverage we have had to skate over or omit many topics, and the third to fourth centuries have been little discussed.⁴⁹ We have made no attempt to eliminate the occasional differences of view between contributors because they illustrate the provisionality of conclusions drawn when dealing with evidence which, though rich in comparison with what is available for other areas of the Roman and late antique world, is tantalisingly localised and incomplete. The broad lines, however, of developments over the period seem fairly robust. The first century AD stands out as a period of extraordinary independence and self-confidence in village communities and their local leaders. The power of these local leaders, whose families dominated office-holding, seems to have derived from their offices, rather than from any exceptional wealth, and was regulated by the willingness of the state, mainly the nome stratêgos, to hear appeals from villagers. Thus villages, and in particular large villages, were able to prosper in general, with high levels of (p.16) functional literacy and market engagement, and a vibrant religious culture. While the prosperity largely continued in the second century, increasing state intervention, in particular the conversion of local roles into liturgies assigned to outsiders, began to limit, presumably deliberately, the administrative independence of villages. This was intensified by municipalisation in the third century, especially following the creation in AD 200/1 of town councils in the mêtropolis of each nome, through the tendency to replace local officials with council liturgists. This trend culminated in Diocletian’s reforms, through which councils themselves became in effect subordinate agents of the central state. The economic and cultural independence of villages also seems have declined over the second to third centuries as urban centres boomed, sucking in economic and cultural activity, as large private landholding spread, and as the state increasingly targeted its fiscal demands at collectivities of taxpayers rather than individual villagers. Thus in the fourth century – and on beyond the Arab conquest – the typical village, if not an estate or a monastic community, came to be a fiscal collective (koinon) dominated by a new village elite (the ‘great men’) whose power was rooted in economic dominance and the lack of interest of the state in how they acted as long as they delivered the collective dues. In social and cultural terms the koinon seems to have been an empty shell. In this void, the church gradually emerged as the main intercessor and source of relief, and perhaps identity, for the rural poor, on the basis of which, in the context of later fiscal reforms and decline of the great men, it would be able to claim a supposedly antique leadership of village communities. As to how typical these developments were of the ancient world as a whole, the normality of Egypt has never been doubted for late antiquity; Berkes, for instance, can happily assert that the fourth to eighth centuries were the age of the koinon through the eastern Mediterranean. For the first to third centuries the question remains more open, largely because it has not seriously been studied and also because Roman historians tend to compare Italy and the west. For now we note that in some respects, such as law and order (Mascellari) and the nature of temple festivals (Jördens), broad similarities can be discerned with developments in the villages of Asia Minor, for which study from epigraphic evidence has recently boomed. We hope this volume prompts more historians to research village institutions in Egypt and Asia Minor and other areas in parallel.