Jump to ContentJump to Main Navigation
Petitions, Litigation, and Social Control in Roman Egypt$

Benjamin Kelly

Print publication date: 2011

Print ISBN-13: 9780199599615

Published to Oxford Scholarship Online: January 2012

DOI: 10.1093/acprof:oso/9780199599615.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 20 June 2018

Who Used the Justice System?

Who Used the Justice System?

(p.123) 4 Who Used the Justice System?
Petitions, Litigation, and Social Control in Roman Egypt

Benjamin Kelly

Oxford University Press

Abstract and Keywords

This chapter discusses the economic status and ethnicity of petitioners and litigants. A good number of people of high and middling economic status appear as petitioners and litigants, and there is no reason to believe that any of the major ethnic groups of the province avoided the justice system as a rule. Nevertheless, the very poor seem to be underrepresented as petitioners and litigants. The costs of petitioning and litigation perhaps deterred some. But the economic profile of petitioners was also the result of ideas about what sorts of disputes were appropriate to bring before a state official: overwhelmingly, the petitions complain of wrongs that involved property or economic transactions.

Keywords:   economic status, ethnicity, litigation costs, property, economic transactions

In this chapter I examine another potential limitation on the capacity of the legal system in Roman Egypt to contribute to social control, namely the economic and ethnic profiles of petitioners and litigants. Just who petitioners and litigants were in a social sense would have had a very real impact on the capacity of the legal system to achieve ‘legal control’. Since there was no state prosecution service, it fell to members of the community to prosecute wrongs. If people in certain socio‐economic statuses never approached officials to request the adjudication of disputes, then this would have limited the impact of the justice system in these milieux. Furthermore, the society of Roman Egypt consisted of people of a variety of ethnic backgrounds; if people from, say, an indigenous Egyptian background tended not to petition or litigate, this would again have represented a limiting factor on the impact that the legal system could have had. On the other hand, if this system was used widely by people from all groups of society, then this would possibly have increased its overall importance.

These are, of course, difficult questions to answer, since our petitioners and litigants are certainly not attempting to give us the sort of data we need to construct a rigorous sociological profile of them.1 The fact of the matter is that for many—perhaps even most—petitioners and litigants, we have no usable information to illuminate their social background. As a result of this, the safest responses to the questions posed above can perhaps be put in the form of three negative propositions. First, there is no major ethnic group that is conspicuously absent from the ranks of petitioners and litigants. On the contrary, people from each of the main groups appear in reasonable numbers (p.124) in these documents, although the evidence does not permit a satisfactory statistical analysis of the question. Second, there is little evidence to suggest that members of the very poorest strata of society engaged in petitioning and litigation frequently, at least not as plaintiffs. Here I am thinking particularly of the general mass of landless and unskilled wage labourers who would have barely managed to eke out a living. One does occasionally find, for instance, shepherds submitting petitions. But generally petitioners of this kind appear only rarely in the documents. Third, we can say that although a good few petitioners and litigants could meaningfully be said to be members of an economic elite, not all petitioners belonged to this level.

To describe the economic status of petitioners in a somewhat more positive (and perhaps rather more adventurous) way, a significant number of petitioners and litigants seem to have come from the middling strata of village and metropolis society.2 There has been a growing consensus that the villages of Roman Egypt contained a significant number of middling landowners who existed at a level well above subsistence.3 Moreover, it is recognized that the metropoleis of the province were centres for a good deal of craft production, and hence contained a sizeable middle stratum of artisans.4 Many petitioners and litigants seem to have come from the ranks of these sorts of people. More precisely, we can say that a significant number of petitioners had one or several of the following features: they owned some urban real estate or agricultural land, although usually not in large quantities; they engaged in relatively complex transactions, either in relation to land or to other economic activities or interests; or they had a range of other engagements with the state beyond the inevitable need to pay taxes: they are found, for instance, carrying out liturgies, holding municipal offices, or leasing public land.

(p.125) One wishes that more could be said about the social level of petitioners and litigants than this. It would be especially significant for our discussion of social control if meaningful generalizations could be made about the relative social levels or ethnic backgrounds of plaintiffs and defendants. If one could prove, for instance, that wealthy plaintiffs routinely sued poorer defendants, then this would allow the conclusion that the law was often used as a weapon to reinforce existing stratification. On the other hand, if one could find a sizeable number of cases of the poor and downtrodden using the law to take on the rich, then this would be interesting as well: it would suggest that law could be a tool for subverting certain aspects of the social order. A suggestion along these lines has indeed been made by Dennis Kehoe in an examination of the role of the legal system in relationships between landlords and tenants in Roman Egypt. Kehoe has suggested that ‘if tenants could use the court system against powerful individuals from whom they were in some degree economically dependent, then we should imagine that the court system could protect humble people in other areas of the law as well’.5 It would also be interesting to know something about what happened in litigation between people of different ethnic backgrounds. One could justly ask whether there are many cases of, say, people of identifiably Egyptian ethnicity suing, say, Roman citizens, and, if there are, whether Romans were given preferential treatment by the courts.

But there are substantial difficulties involved with determining the relative economic statuses or ethnic backgrounds of plaintiffs and defendants. In most petitions and reports of proceedings, no precise claims are made at all about the ethnic background or economic status of the alleged wrongdoer. When such claims are made, they often bear the hallmarks of being hopelessly contaminated by rhetorical topoi, as we have seen.6 Nor can sound inferences be made from the nature of the dispute at hand. One is tempted to say, for example, that cases of creditors and debtors suing each other are instances of people of different economic levels using the legal system against one another. But it does not follow from the fact that X borrows money from Y that X is of a significantly lower economic level. The loan suggests that X is short of ready cash, but he may have considerable assets in a non‐liquid state. Likewise with landlords and tenants: there (p.126) is evidence that people of quite considerable means who owned land themselves also leased additional land as part of their economic strategies.7 Moreover, Kehoe's thesis, even as it relates to landlords and tenants, suffers from the difficulty that only three out of the 568 petitions in the database clearly arose from a dispute between landlord and tenant over an agricultural tenancy, and all were submitted by landlords.8 In relation to petitions complaining of violence, it has also been suggested that most assailants were of the same social status as their victims. Yet this seems to be based simply on the fact that most violence cases seem to have involved ‘neighbors, family members, business partners and fellow villagers of the same status’.9 But none of these facts really suffices to prove the point: neighbourhoods could have contained people of varying status; some branches of families were no doubt richer than others; and it is not unthinkable that people sometimes went into business with those of a different economic level.

In this chapter, I therefore regretfully put to one side the question of plaintiffs and defendants’ relative economic statuses and ethnic backgrounds. Instead, I focus purely on what can be said about plaintiffs. I exploit two types of evidence. First, there are several petitioners about whom we know a great deal, thanks to the fact that their family archives have survived. An examination of a representative sample of these archives will show that these people, although not exactly members of the elite of the province, were nevertheless moderately prosperous small landowners or businessmen, who engaged in relatively complex economic transactions. Some of them had a range of interactions with officialdom that went beyond those required by the taxation system: quite a few of (p.127) these petitioners were liturgists, for example. Of course, all of this raises the question of the typicality of these petitioners whose archives have survived. I therefore turn to exploiting a second category of evidence, namely the many casual details in petitions which are indicative of ethnicity and ethnic status. These details do not lend themselves to statistical analysis. Nevertheless, the evidence does tend to converge in a particular direction, and suggests that the petitioners whom we meet in archival contexts were not especially unusual. These casual details in the petitions also lend some independent support to one of the negative claims that I have already mentioned: that the very poor are underrepresented as plaintiffs in the petitions and in reports of proceedings. Finally, at the end of the chapter, I offer an explanation for the economic profile of petitioners. Part of the explanation no doubt lies with the expenses involved in litigation. Also important was the fact that people who were propertied and who engaged in complex business dealings were simply more likely to become involved in the sort of disputes that were thought to be appropriate for litigation. The past engagements of such people with officialdom also would have meant that they were likely to have sufficient basic knowledge about the workings of the administration to know how to go about using it in their disputes.

Archival evidence

Satabous, son of Herieus

The archive relating to Satabous and his family, as well as illustrating their various engagements with the legal system, contains a considerable amount of other information. It is clear that Satabous was literate in demotic,10 and his son Stotoētis was probably literate in Greek.11 It also emerges from the petitions and the other documents of the archive that Satabous was a comparatively prosperous man, although by no means a member of the highest levels of provincial society. We learn that the house that Satabous purchased from Chairēmōn, and to which he was accused of annexing ownerless land, was purchased for (p.128) 740 dr.,12 had multiple storeys, and included a tower (pyrgos), a central court (aithrion), and various open grounds (topoi psiloi).13 When Nestnephis proved his allegation that Satabous had annexed more open grounds, these were apparently added to the property. They were, according to Nestnephis, worth 300 dr., although Satabous had to pay 500, a sum which perhaps included a fine.14 Satabous also owned a second house, which was a two‐storeyed mill‐house complete with mortar stones.15 Presumably this structure was purchased so as to exploit its economic potential, either by leasing it or working it directly (perhaps with hired or slave labour). The dispute with Harpagathēs arose out of another type of profit‐driven activity: Satabous’ loan to Harpagathēs of 325 dr. This was a loan at interest; it was also secured in a sophisticated way, namely against certain priestly revenues to which Harpagathēs had rights, and which Harpagathēs was not supposed to touch until the principal and interest were repaid.16 This loan to Harpagathēs was perhaps not Satabous’ only loan transaction: our Satabous is perhaps identical with a party in a demotic loan contract from AD 4 mentioning a Satabous, son of Herieus the younger.17

Satabous was also very heavily involved in the finances of the temple of the god Soknopaios. A series of demotic receipts make it clear that he was the Scribe of the Priests of the Temple of Soknopaios for a probably unbroken stint from 11 BC to 5 BC.18 Various other demotic receipts and agreements from the temple, most of them unpublished, give a good sense of the duties involved in this position. These included the receipt of various types of taxes that the Temple was charged with collecting, and the withdrawal from the state bank of the syntaxis—that is, the subventions provided to the temples by the provincial authorities. Satabous therefore played an important role in the temple economy of Soknopaiou Nēsos, and would have (p.129) interacted with a wide variety of people in playing this role, including government officials.19

This picture of a man of reasonable means, who engaged in relatively complex economic strategies involving urban properties and loans, is supported by what we know of activities of other members of his family. The surviving Greek subscriptions to a demotic contract show that in 8 BC, Satabous’ father, Herieus, purchased a part of a house, acting jointly with another man, T[.]sēs, who, interestingly, was possibly the father of Nestnephis.20 Herieus also seems to have been responsible for the purchase of the mill‐house which is later found in his son's possession, and which Nestnephis raided. We can gather this from the fact that in the Greek translation of a deed of cession of 3 BC, two brothers cede their rights to the mill to Satabous, but in doing so mention that somebody—probably their eldest brother—had originally sold the property to Satabous’ father, Herieus, in 14/13 BC. It is not clear just why an additional deed of cession was needed a decade after the sale, although one possibility is that there was an ongoing dispute over rights to the property which was eventually settled by the heirs of the original buyer and seller.21

Satabous’ sons are also found involved in similar activities. In a contract of AD 51, we find one of Satabous’ sons, Stotoētis, purchasing a one‐ninth share in a house.22 In a bilingual contract of AD 54, Satabous’ other son, Herieus, and his wife sell a two‐storeyed house to a certain Psenamounis. The contract is not a true sale, however: there is an attached contract in which Herieus borrows 260 dr., which tends to suggest that the sale contract simply served as security for the loan.23 He might perhaps have used a second property interest that he owned for a similar purpose: in a document from AD 41–54 he is found ceding a one‐ninth share in a house. In around AD 88, his four children are found selling a one‐ninth share of the same house for 68 dr., so it would seem that the share stayed in the family.24 This is perhaps best explained if the cession in AD 41–54 was actually done to (p.130) secure another loan. To complicate things still further, the same four children are found in the period AD 72–8 registering the sale of a one‐ninth share of a house worth 6[.] dr. to the man to whom they later sold a one‐ninth share of a house in AD 88.25 This too could be explained by their using their interest in the property to secure a loan, just as their father had done. One of Herieus’ daughters, Tesenouphis, is also found in a document of AD 69–79 purchasing a one‐third share in some open land adjacent to the house she inhabited with her husband.26

Finally, a man from Soknopaiou Nēsos called Herieus, son of Satabous, submitted a petition during the reign of Nero regarding a dispute concerning an oil press on an imperial estate.27 It has been suggested on the basis of the identities of the name and domicile that this is the son of our Satabous. The identification is not entirely secure, however, especially given the very restricted range of personal names used in this village.28 But if this is the same man as the son of our Satabous, then this would give further insight into the economic activities of the family. According to the petition, which is dated to sometime after AD 60, the petitioner had leased an oil press since AD 52/3 for 200 dr. and three choes of oil (about ten litres) per year. The complaint is that the managers of the imperial estate on which the press was situated had let it fall into such a state of disrepair that the petitioner was forced to invest his own money in repairs, and even then it interrupted production. There seems to be an element of dishonesty in the request, since in an application of AD 55 to lease the oil press, the same man undertakes to pay for various of these repairs at his own expense.29 Finally, in AD 71, a certain Stotoētis, son of Satabous, son of Herieus, purchased from his mother‐in‐law a two‐storeyed priest's chamber and yard situated in the enclosure of the temple of the god Soknopaios. In view of the age of the purchaser and his names, this is most likely the great‐grandson of our Satabous.30

(p.131) Tryphōn, son of Dionysios31

The archive of Tryphōn comes from a few decades after Satabous’ various engagements with the legal process. Tryphōn is known to have submitted five petitions.32 There is also a report of proceedings in the archive concerning a case brought against him and his second wife by a certain Pesouris (also known as Syros). Tryphōn was probably the least wealthy of the petitioners whose affairs are illuminated by an archive, especially in the early part of his adult life.33 He was a weaver, like his grandfather and father before him. His brother Onnōphris was also a weaver, as were his sons.34 It is clear that he was illiterate.35 At one stage, his second wife, Saraeus, was forced to accept work as a wet‐nurse.36 This said, he was in no sense a member of the bottom echelon of society. For one thing, he was one of the metropolites who paid the poll tax at the reduced rate of 12 dr.37

Tryphōn also emerges as an owner of property, owning several houses or shares of houses during the course of his lifetime. In a petition of AD 50, there is a reference to an apparent burglary of a house possessed jointly by Tryphōn and others; the fragmentary state of the text does not make it clear whether they owned or leased the house, however.38 In AD 55 he purchased a half share in a three‐storeyed house from his mother's cousin, possibly taking out a loan of 104 dr. to do this.39 Then, in AD 59, he purchased another house, the purchase made possible by loans from several people, including the vendor, one for 160 dr. and one for 314 dr.40 The scale of Tryphon's (p.132) ownership of real property is therefore smaller than most other petitioners known from archives, but there is still reasonable complexity in his property interests and economic activities.

There is another interesting feature of the Tryphōn archive that illustrates the relationship of the weaver and his family to the legal realm. As well as loans from or to people who do not appear to have been a member of Tryphōn's family,41 the archive contains a number of contracts documenting loans between Tryphōn and members of his family, often loans of quite small amounts of money.42 Thus in one, Tryphōn lends his mother 16 dr.43 These contracts are drawn up according to the usual formulae, and include heavy penalty clauses imposing a 50 per cent additional penalty for late repayment, and giving the creditor the right to execute on the person and property of the debtor if he or she defaults. Nor was Tryphōn the only member of his family to try to crystallize intra‐familiar exchanges into strict legal rights. In a loan contract of 22 May AD 37, Tryphōn's second wife, Saraeus, loans him a total of 72 dr., partly in the form of cash and partly in the form of a tunic (chiton) and earrings. The loan is expressed in the document as being for a period of slightly more than five months, but in the event he did not repay it until AD 43. There is also a clause in which Tryphōn undertakes to pay a certain sum of money to Saraeus for child‐birth expenses if they separate while she is still pregnant.44 There has been extensive discussion in the modern literature about just what the purpose of this loan was. Most plausible is John Whitehorne's suggestion that it was a case of Saraeus using a contract to protect her resources. She had lent her new husband money and valuables, perhaps because he was in financial difficulty. At the same time she was pregnant with their first child. Given that Tryphōn's first marriage had come to a sorry end, his new wife wanted to secure the money that she had lent to her husband and the immediate future of herself and her unborn child.45 Finally, there is another fragmentary and undated contract between Tryphōn and Saraeus in (p.133) which she agrees to suckle his child.46 The document was probably drawn up between Tryphōn and Saraeus after they began cohabiting,47 so this would provide yet another case of legal contracts being used to structure relationships in this particular family.

Ptolemaios, son of Diodōros, also known as Dioskoros

This individual was a resident of Theadelphia whose activities are attested for the period AD 138 to AD 166. No fewer than twelve of Ptolemaios’ petitions are extant.48 They are directed against seven different opponents. It is clear from these documents that he was literate enough to subscribe—and probably also write—his own petitions.49 Ptolemaios’ petitions are preserved in a family archive, and some tax lists make references to individuals who are quite possibly Ptolemaios and various of his relatives.

The picture that emerges from the archive is of a family that was perhaps not in the very highest economic strata of Egyptian society, but which was nevertheless moderately prosperous.50 Various members of the family, including Ptolemaios, seem to have owned land. The same texts also hint that Ptolemaios and his family pursued fairly complicated economic strategies of the sort likely to give rise to disputes. Thus, we learn from a property return from AD 147 that his sister, Sambathion, owned a house in the village of Theadelphia which she had purchased herself, and which Ptolemaios, his wife Achillis, and their son inhabited. Sambathion also owned property in the nearby village of Apias.51 A tax list of AD 166 records that (p.134) a certain Ptolemaios paid taxes on 1 ⅛ ar. of vine land and owned another one aroura, which was tax‐exempt for reasons unspecified. The entry also states that the land had previously belonged to a woman called Achillis, and that Ptolemaios, son of Diodōros, had also paid taxes in relation to two ar. owned by a certain Sambathion.52 Given that our Ptolemaios was married to an Achillis and had a sister called Sambathion, this is surely the same man. The family therefore owned a variety of small plots of agricultural land and some urban real estate. Various other tax lists from mid‐second‐century Theadelphia mention individuals who could be Ptolemaios or members of his family, and the fact that the individuals on the tax lists seem to own or farm small plots of land is consistent with what we know about the family.53 With this said, there is the possibility that these are simply cases of men bearing the same (rather common) Greek names as the men of Ptolemaios’ family.

Ptolemaios himself leased as well as owned land. A petition from the period AD 144–7 mentions that he leased usiac land.54 We learn from two other petitions dating to this period that the land leased consisted of a drymos, an area of marshy land, possibly used for the cultivation of reeds and papyrus plants, or for commercial fishing.55 His claims in the petitions are perhaps also supported by an account in which a Ptolemaios, son of Dioskoros, amongst others is listed as paying rent on usiac land.56

Ptolemaios had several types of engagement with officialdom. Various ‘penthemeros’ reports show that he held a liturgical office with the rather elaborate name of ‘superintendent of pastures and the (p.135) marsh of Theadelphia and Polydeukia and of the priestly contracts accounted for under imperial leases and of the other aquatic revenue devolving thereto and of fishing’. He held this in AD 138 and again in AD 148–9.57 The ‘penthemeros’ reports gave an account of the revenues for a five‐day period, as the name suggests. They were submitted to the office of the stratēgos, usually by one of the members of the board of superintendents. In five of the six ‘penthemeros’ reports mentioning Ptolemaios there are subscriptions stating who on the board actually submitted the report to the office of the stratēgos. In all five cases, Ptolemaios is listed as having carried out the submission—four times alone, and in the fifth case in combination with four of the other members of the board. This is striking, since three of Ptolemaios’ surviving petitions were addressed to the stratēgos,58 and several of his other petitions from the mid‐140s AD ask higher officials to delegate cases to the stratēgos, or order the stratēgos to carry out some other task in relation to the matter.59 In spite of the fact that one petition claims that members of the stratēgos’ office had been behaving corruptly,60 Ptolemaios’ petitions suggest that he was a man willing to involve the stratēgos in his disputes. In view of the fact that, at certain points in his life, the submission of ‘penthemeros’ reports would have brought Ptolemaios into very regular contact with the stratēgos and his staff, it is easy to see how this willingness would have arisen.

A similar point can be made in relation to a petition that he submitted to the assistant of the procurator usiacus in AD 144. As it happens, this is one of only two petitions in the database that were submitted to the office of the procurator usiacus.61 The fact that Ptolemaios is found petitioning this relatively specialized office is explicable if we remember that the liturgical office which he had already held in AD 138 was concerned in part with revenue generated on imperial estates. By AD 144 he is also found leasing a drymos that (p.136) was part of an imperial estate; indeed, it was quite possibly some of the same land he had once overseen as part of his liturgy. It is likely, therefore, that Ptolemaios would have either had prior personal contact with officials from the bureau of the procurator usiacus, or at least been familiar with how this particular part of the provincial administration worked. It is hardly surprising that we find him submitting a petition to this office.

Sokratēs, son of Sarapiōn62

Sokratēs, son of Sarapiōn, was an inhabitant of second‐century Karanis. He submitted only one petition about which we know, a complaint relating to some kind of wrong done in relation to fields that he cultivated.63 He was, however, apparently involved in a second legal matter. In a private letter addressed to Sokratēs, a woman by the name of Artemis states that she has arrived in ‘the city’ and asks that he send one of his people to act as her representative (ekdikos) in a lawsuit. She states that it is not possible for a woman to go to law without such a representative—a statement which, as a matter of law, does not appear to have been true.64 She also hints that Sokratēs’ interests are somehow bound up in the success of the lawsuit, so this could be a second case of his going to law.65 Unfortunately, the published documents of the archive do not make it clear who this woman was, or whether Sokratēs was involved as plaintiff or defendant.

Sokratēs’ archive shows that he was at various points a collector of money taxes and a census official.66 His position as a collector of money taxes is confirmed not only by the survival of several receipts issued by him, but also by several enormous rolls which record the receipt of taxes.67 One of the hands in the rolls is the same hand as his petition, so he apparently wrote (or at least copied) his own (p.137) petition.68 The rolls also show that he earned a very high income from his position as collector of money taxes—something in the region of 10,000 dr. a year. This would have made him one of the most affluent inhabitants of the village.69 In his petition, he refers to other lots of land that he cultivates, apart from the one involved in the dispute about which he petitions.70 This claim is consistent with the fact that his archive contains a receipt for a loan of seed.71 There is also an application addressed to him by an individual wanting to sub‐lease a date palm grove which formed part of some royal land on which Sokratēs held the lease,72 so he evidently was dealing with land in an entrepreneurial way, and not just with a view to cultivating it himself.73

As for the members of Sokratēs’ family, it has been suggested that his wife was a Roman citizen.74 His father might have been a tax official in Ptolemais Nea.75 His brother also may appear in another document as a sitologos in charge of the public granaries of the village.76 One of Sokratēs’ sons went on to become a collector of money taxes, just like his father,77 and another appears as a census official.78

The tax rolls make it clear that Sokratēs was literate. In fact, he was clearly more than literate, and had some literary attainments. The archive contains a copy of Menander's Epitrepontes, a specimen of the so‐called Acta Alexandrinorum, and grammatical papyri. Moreover, one of the tax rolls contains Greek translations of the Egyptian nicknames by which some taxpayers were known.79 As Herbert Youtie has recognized, one of these translations uses an obscure word that appears only in Callimachus.80 Sokratēs, therefore, was (p.138) perhaps erudite enough to know his Callimachus, and use it to amuse himself amid the tedium of compiling these dull documents.

Gemellus Hōriōn

The published documents of the archive of Gemellus Hōriōn and family give information about this man and his family for the period from AD 154 through to the reign of Severus Alexander (AD 222–35).81 The archive contains documents relating to Gemellus’ grandfather, C. Julius Niger, his father, C. Apollinarius Niger, and Gemellus himself. This fact tends to suggest that the archive was originally owned by the family, and then passed down the generations. The archive establishes that members of the family, and particularly Gemellus himself, were repeat players in the petitioning game. Eight petitions in Gemellus’ name have survived (three of which have been preserved in duplicate).82 It is clear from the petitions that he was illiterate.83 Two petitions submitted by C. Julius Niger have also been preserved in the archive,84 along with one submitted by Apollinarius Niger.85 Gemellus’ heavy use of petitioning to deal with his disputes should therefore be seen within the context of his family's prior experience with the judicial process.

The members of the family enjoyed various special status categories that marked them out from many of the inhabitants of Karanis and its surrounding villages, the region in which the family was active. Gemellus enjoyed Antinoite citizenship, as did his parents, grandfather, uncle, and sister.86 Gemellus’ grandfather, C. Julius Niger, and his uncle were probably Roman citizens, although Gemellus’ (p.139) father seems not to have been a Roman citizen, perhaps because he was born during C. Julius Niger's period of military service, and therefore would not have been given citizenship at the time of his discharge.87 There is no clear sign that Gemellus himself was a Roman citizen before the constitutio Antoniniana of AD 212.88

Like Ptolemaios, Gemellus and his family were small‐scale landowners. Gemellus himself is known to have owned several plots of land in the vicinity of three different villages. He owned 4 ar. of katoikic land near Kerkesoucha and another 1 ar. of katoikic land near Karanis, both of which were cultivated for grain production.89 A declaration of AD 200/1 mentions that he had possession of 3 ar. of sacred land near Psenarpsenēsis, which he is said to have ‘owned’, in keeping with the tendency in the Roman period to assimilate some plots of sacred land with private land.90 The same declaration has him owning another aroura of land of unspecified type, registered under the name of Gemellus’ grandfather, but almost certainly belonging to him by this stage.91 In another text of AD 214 he registers a total of 3 ½ ar. of grain land near Psenarpsenēsis, which might have included the same land registered in AD 200/1.92 In addition to grain land, he owned a 1 ¾ ar. plot of garden land near (p.140) Karanis, and in one of his petitions he claims to have owned an olive yard near Kerkesoucha of unspecified size,93 a claim that is apparently supported by an unpublished Michigan papyrus.94 Thus, we know that in total he owned at least 10 ¼ ar. of land, and possibly 13 ¼ ar. if the two references to land in Psenarpsenēsis in fact refer to separate plots. Of course, this is a cumulative total: it is possible that he owned only some of these parcels of land at any one time, since most are only attested for a single point in his life. We learn from a census declaration that Gemellus and his sister jointly owned, in addition to agricultural property, three houses, four shares in houses, and eleven courtyards (some attached to the houses, others separate).95

Various other members of Gemellus’ family are also known to have owned small plots of land, and some of Gemellus’ land would have been inherited from family members.96 His grandfather, himself a petitioner, owned an olive grove of 1 ar. and 2 23/24 ar. of katoikic land near Psenarpsenēsis.97 A tax roll from Karanis from the period AD 172–4 has him paying taxes which indicate that he owned several small plots of katoikic land, garden land, and gē en aphesei.98 A collection of receipts from the 180s list him as paying taxes on garden land of somewhat over 2 ar. in Karanis, which may represent an expansion of an original holding, or an additional holding or holdings.99 Gemellus’ grandfather and his father, C. Apollinarius Niger, are found in AD 164 declaring a katoikic allotment that they (p.141) owned jointly near Kerkesoucha.100 C. Apollinarius Niger is himself found paying taxes on 91/12 ar. of katoikic land in Kerkesoucha, as well as a small parcel of five‐eighths of an aroura in an unspecified part of the division of Herakleidēs.101 A very fragmentary passage in a tax receipt may also suggest that a member or members of the family paid taxes on an additional 6 ar. of katoikic land in AD 183.102

The information surviving in the archive relating to the family's property holding makes it clear that Gemellus and his family engaged in a wide range of complex economic transactions. The fact that plots held in a given location seem to vary in size over time could suggest that the family was constantly adding to existing plots. The practice of partible inheritance would have contributed to this, but there is also evidence of purchases. We hear that C. Julius Niger purchased his olive grove from land sequestrated by the government.103 Gemellus’ father also acquired his parcel of five‐eighths of an aroura in AD 176 from someone outside the family, and a fragmentary passage in the text has been plausibly reconstructed to suggest that it too was purchased from land confiscated by the government.104 The number and geographical distribution of the houses and plots of agricultural land raise the suspicion that the family would have exploited these assets in a variety of ways. This receives confirmation from a petition in which Gemellus complains about the activities of an individual who was using black magic on Gemellus’ land to halt the work of someone whom Gemellus calls ho geōrgos mou (‘my cultivator’).105 This suggests that Gemellus was exploiting this land through hired labour, or perhaps a tenancy or share‐cropping arrangement. In a later tax receipt, a Julius, son of Eudas, who was actually the individual against whom Gemellus made complaints of black magic, is said to have paid one artaba of katoikic dues on Gemellus’ behalf. This suggests that the two men had mended their differences, and Julius was working Gemellus’ land as a tenant.106

(p.142) In view of the relatively complicated economic strategies pursued by the family, it is little wonder that most of their petitions relate to transactions pertaining to their land, or offences against property and produce. Two petitions, one submitted by Gemellus and one by his grandfather, C. Julius Niger, complain of thefts of wheat and donkeys respectively.107 C. Julius Niger's other petition deals with some kind of dispute with a scribe of the overseer of sequestrated property over the plot of such property that he had purchased.108 The text is fragmentary, but presumably something about the transaction gave rise to the dispute. Gemellus himself also is known to have submitted five petitions complaining of disputes with officials. Two are directed against a certain Kastōr, who was some sort of official connected with the collection of dues and taxes who allegedly pursued his task in too heavy‐handed a fashion.109 The complex pattern of landholding in the family meant that its members were liable to pay a wide range of taxes, so it is no surprise that such disputes arose. Another petition is submitted by Gemellus and a second man in their capacity as landowners in Kerkesoucha, and on behalf of the other landowners and public farmers of the area. It complains about the lax way in which local officials charged with directing irrigation works were doing their job, and the danger this posed to the agricultural land of the area.110 A fourth petition is directed to the epistratēgos against local presbyteroi who tried to nominate Gemellus to a liturgy, ignoring immunities that he supposedly enjoyed. This text refers to an earlier petition which he had submitted to another official on the same matter.111 This case is a typical instance of wealth and property holding giving rise to liturgical responsibility, and all the disputes that inevitably accompanied that. Finally, there are two petitions complaining about the behaviour of Julius, son of Eudas, and his brother. The origins of the dispute are obscure, and the matter seems to have escalated to something of a feud, but one suspects that land tenure or rights to land of some kind were at the root of the matter, given that (p.143) Gemellus claims that his opponent came repeatedly onto his land and raided or damaged crops.112

Internal evidence: ethnic background

The archival evidence therefore allows us to see in detail the affairs of several petitioners. But it is not enough merely to focus on these few richly documented cases. One may justly suspect that these individuals are not representative of the general run of petitioners. After all, people who owned immovable property, who engaged in complex business transactions, who held liturgical positions, or who were literate were, thanks to these very facts, likely to generate archives. Relying purely on the evidence of these archives therefore puts us at risk of making a circular argument. Thus, in this section and the next, I focus on various categories of information in the petitions themselves which are relevant to the ethnic backgrounds and economic statuses of petitioners. This sort of evidence, I suggest, helps to establish that the petitioners discussed above were not especially unusual by the standards of petitioners. It also helps to establish independently the three negative propositions with which I began this chapter.

First, there is the slippery question of ethnicity. One could reasonably ask whether an individual's ethnicity had an impact on his or her likelihood to submit petitions or engage in litigation.113 The short answer to this question is that there is enough evidence to suggest that no major ethnic group tended on the whole to avoid petitioning and litigation,114 but that no statistical precision is possible on the question of whether some groups were somewhat more inclined to use the justice system than others. Stating the position in this summary way occludes the true complexity of the situation, however. For in drawing conclusions about the impact of ‘ethnicity’ on petitioning and litigation practices, one faces two problems. First, the ethnicities of at least (p.144) some individuals in Roman Egypt would have been ambiguous, both to the community and to themselves. Second, the petitions and reports of proceedings usually do not give us the types of information that we would need to make inferences about the identities of even those individuals whose ethnicities would have been relatively unambiguous.

The matter of ethnicity is one of the most difficult and heavily studied issues in modern scholarship concerning Graeco‐Roman Egypt. Even before the Ptolemaic period, Egypt had been an ethnically diverse region. The Ptolemaic period then saw a large influx of immigrants from around the Greek world. As these new arrivals began to mingle with (and in some cases intermarry with) the indigenous Egyptian population, a complex set of ethnic and cultural interactions ensued. Not every case is problematic, of course. In Ptolemaic Egypt, there were three poleis, Alexandria, Ptolemais, and Naukratis. These were patterned after the cities of old Greece, both politically and culturally. Their citizens would have almost all been unequivocally Greek, both in the eyes of the government (by virtue of their citizenship of the polis), and in the eyes of themselves and their community (by virtue of language, religion, and their participation in civic and political rituals, inter alia). At least, we assume that this would have been the case: these Ptolemaic poleis are very poorly documented in the papyri, although the literary record does help to illustrate the strong Greek culture of Alexandria.115

Outside of the three Greek poleis, ethnicity was a more complex phenomenon during the Ptolemaic period, both for the ancient and modern observer. It would have become more complex as time went on. For some people, ethnic labels would have depended on who was doing the labelling, and the context in which the labelling was being done. One source of ethnic labelling was the Ptolemaic government, and, given the nature of our papyrological documentation, such bureaucratic efforts to classify the population are fairly well attested. Thus, for example, several official rulings during the third century BC required certain people to state their ethnic designations along with their names and other pieces of identifying information in particular types of Greek legal and administrative documents.116 The change of (p.145) ethnic designation was controlled by the authorities.117 In the third century BC, the government also gave an exemption from the obol tax and from compulsory labour on the dikes to certain men who were designated ‘Hellenes’, as well as to those labelled as ‘Persians’ and ‘Arabs’.118 Such ethnic labels used for official purposes sometimes concealed a much more complex reality, however. Certain individuals who enjoyed the status of ‘Hellene’ for taxation purposes were actually Jewish; others came from families of Egyptian and Thracian origin.119 And official designations do not, in isolation, tell us terribly much about how individuals’ ethnicities were really perceived by themselves and their community, and how they expressed their ethnicity in cultural practices.

To probe more deeply the question of an individual's ethnicity in Ptolemaic Egypt, we would have to have either specific statements from that individual or members of the community on the question, or at least detailed information about matters such as language, nomenclature, religion, education, leisure activities, and culinary habits. Unfortunately, such information is lacking for most people who appear in the documentary record. Very occasionally, individuals will explicitly state in documents that they are Hellenes—for instance, Ptolemaic petitioners will juxtapose their identities with those of misbehaving Egyptians.120 More commonly, we at least have the names of individuals mentioned in the documents, and names in Egypt generally had clear ethnic origins. But the use of nomenclature as a sole criterion for ethnic identification is rightly viewed with a certain amount of ambivalence by papyrologists, especially when it comes to the later stages of the Ptolemaic era.121 Part of the reason for this is that some people, especially those involved in the administration and the army, adopted dual names—one Greek and one Egyptian—which they would use in different contexts.122 Then there is the fact that an indeterminate number of people of Egyptian origin would have adopted Greek cultural practices (or vice versa). Such people would therefore have been in an unclear category as far as their own (p.146) and others’ perception of their ethnic identity was concerned. None of this is to say that we should conclude that there were no significant ethnic distinctions in Ptolemaic Egypt: many people would have unequivocally seen themselves (and been seen by others) as Egyptians, Greeks, or Jews. The problems come when one tries to gauge the number of people who would have fallen into a more ambiguous category, or to determine with certainty the ethnic identities of specific individuals.

The already complex problem of ethnicity in Egypt became still more complex with the advent of Roman rule. In studying the Roman period, many of the evidentiary problems that one encounters in the Ptolemaic period remain when it comes to telling how individuals’ ethnic identities were perceived subjectively by themselves and those around them. It is true that the Romans instituted a rigid system for classifying the population, with Roman citizens at the apex of the hierarchy, citizens of the three (later four) Greek‐style poleis next, and ‘Egyptians’ at the base.123 These categories suggest ethnic division, but the reality was a good deal murkier. Most citizens of the poleis of Roman Egypt, like their Ptolemaic forebears, were probably Greek in an unambiguous sense.124 On the other hand, many, perhaps most, Roman citizens in the first two and a half centuries of Roman rule were military veterans who had settled in Egypt following discharge, or their descendants. They had Roman status as far as the government was concerned, and very many would have identified them as such. But the recruitment patterns of the Roman army meant that their ethnic origins and cultural practices very often would not have been Roman or Italian. After AD 212, when nearly the whole free population of the empire was given Roman citizenship, this observation applies a fortiori. The category of ‘Egyptians’ is also a misleading one, since it included the inhabitants of the nome metropoleis, many of whom would have considered themselves to be Greeks and been considered as such by others in their communities.125 The fact that such people were classified as ‘Egyptians’ by the Roman authorities suggests that, by the end of the Ptolemaic period, the question of (p.147) ethnicity had become such a complicated one from an outsider's perspective that the Romans were forced to adopt this rough‐and‐ready alternative.126

What, then, can be said about the ethnicity of petitioners and litigants in Roman Egypt? One could approach the question from the perspective of the official status categories imposed on the population by the Roman authorities—with the important caveat that these only indicate ‘ethnicity’ in a narrow, bureaucratic sense. The finding here is that, while the majority of petitioners make no specific or implicit claim about the category to which they belong, one finds a good number of petitions submitted by members of each of the main status groups. Thus, there are twenty‐seven petitions from before the universal grant of citizenship in which petitioners claim explicitly or implicitly that they are Roman citizens.127 A good number of petitions contain claims of citizenship of one of the Greek poleis of Egypt; the overwhelming majority are from citizens of Antinoopolis,128 with a few from Alexandrians,129 and none from citizens of Ptolemais or Naukratis. This imbalance is simply the result of patterns of documentary survival: citizens of Antinoopolis lived in a region from which large numbers of papyri have been recovered; citizens of the other three poleis did not. There are also a good few petitions from people who would have been officially ‘Egyptians’ in the eyes of the Roman administration. People who are said to be metropolites are an easily identifiable group of petitioners in this (p.148) category.130 There are also very many people who give their original domicile (idia) as a village, and who would therefore have been ‘Egyptian’ as far as the official classification was concerned.131 It is tempting to do a statistical comparison of the relative numbers of Roman citizens, citizens of Greek poleis, and ‘Egyptians’ appearing in petitions. But this would be entirely invalid. The survival patterns of papyri mean that certain groups would be underrepresented not because they eschewed petitioning, but because of their geographical concentration (for instance, citizens of Naukratis, Ptolemais, and Alexandria). In any case, we do not know the overall proportions of Romans, ‘Greeks’, and ‘Egyptians’ in the population, so it would be difficult to say that a particular group petitioned with a frequency disproportionate to its overall size.

As for the reports of proceedings, these are less informative since their format lacks the sort of crisp statements of the parties’ names and domiciles with which petitions open. One can identify a few Roman citizens,132 citizens of Greek poleis,133 and metropolites in them.134 Unfortunately, though, one cannot identify the status of enough litigants to decide whether lawsuits initiated by people of higher status reached judgment more often, or whether judges gave preferential treatment to any particular groups.

(p.149) The evidentiary situation is still worse if we move away from the technical ‘ethnic’ categories imposed by the Roman administration, and try to find other criteria for determining subjective perceptions of ethnicity. One generally cannot tell which language(s) petitioners and litigants spoke. Greek was the language of both petitions and trials. But bilingual scribes would have been available, so the fact that a person's complaint was written in Greek proves nothing about the languages that he or she could write or speak. In court, interpreters could be used, and one finds reference to their involvement in a handful of cases.135 But the fact that interpreters are not mentioned in the majority of other reports of proceedings does not prove that most litigants spoke Greek: the compressed nature of the reports of proceedings and the fact that many of the surviving versions are abstracts from longer texts mean that references to interpreters could have been left out in some cases.

One could look at onomastic criteria, and it would certainly be possible to produce impressive lists of petitioners with Greek names and others with Egyptian names. But, given the unquantifiable margin for error involved with deducing ethnicity from nomenclature, a rigorous statistical approach would be potentially misleading. At most, onomastics confirm what we already know: that the major ethnic groups were well represented amongst petitioners and litigants. Finally, religious practices are another potential indicator of ethnicity. A reasonable number of petitioners specify that they are priests. In the case of people who, like Satabous, are said to be priests of Egyptian gods or priests in temples devoted to such gods,136 a strong inference of Egyptian ethnicity can be made. Something similar can be said for people with Egyptian names resident in villages who are described as ‘priest’ or ‘priestess’.137 One can also obviously make inferences about the religious practices and ethnic identities of petitioners who are described as Jews, or have clearly Jewish names, of (p.150) which there are a few.138 However, members of the significant Jewish population in Alexandria are largely absent from the petitions, just like the rest of the population of Alexandria, thanks to the survival patterns of papyri.

Finally, an individual's actual place of residence could be used to support an inference about ethnicity. (This is an issue slightly separate from whether they were technically a metropolite or not: metropolites sometimes were resident in villages, and not everyone resident in a metropolis would have enjoyed metropolitan status, since it was hereditary.) As a very general proposition, Greek culture was stronger in the metropoleis than in the villages. But it is certainly possible to find Hellenized individuals living in villages and villages with strong elements of Hellenic culture. Sokratēs, son of Sarapiōn, with his obvious engagement with Greek literature, was not an aberration: Greek literary papyri have been found in reasonable numbers in a variety of Fayum villages.139

Internal evidence: economic status

Ethnicity is by no means the only category that we can try to use to examine the identities of petitioners and litigants. ‘Economic status’ offers another category of analysis, and one that deserves a separate consideration. Whilst there was often a rough correlation between a person's ethnicity and his or her economic situation, ‘economic status’ has its own set of indicators and evidentiary challenges. One of the most striking categories of evidence that bears (albeit indirectly) on the economic status of petitioners relates to literacy. When a petition was submitted to the relevant official, the practice was for the document to be subscribed to authenticate that the person named in the petition really was the person submitting it. These subscriptions in some senses served to confirm identity in the same way that signatures do in the modern world. A subscription could take a number of forms. If the petitioner was literate, he or she would typically write a subscription in the form: ‘I NN. have submitted it.’ Generally the perfect form of the verb epididōmi is used. For example, (p.151) a petition complaining of the theft of a calf submitted in AD 66 by a certain Ammonios, son of Eirenaios, has the following words written at its base in a second hand: ‘I, Ammonios, son of Eirenaios, have submitted the above petition.’140 This, however, was a society with widespread illiteracy. So if the petitioner happened to be incapable of subscribing his own petition, one of two authentication devices was used. A petitioner might have someone subscribe the ‘epidedōka formula’ on his or her behalf, in which case this amanuensis would mention his own name in a so‐called ‘illiteracy formula’. For instance, a petition from AD 190 has the subscription: ‘I, Nechthenibis, have submitted it. I, Diogenēs, wrote it for him, since he does not know letters.’141 Alternatively, a physical description of the illiterate person submitting a petition was given,142 again presumably as a sort of authentication device if the petitioner subsequently attempted to disown the petition.

Of the 568 petitions in the database, 117 have one of these three formulae at their bases.143 This rather low total is simply the result of the fact that most surviving petitions are either copies or drafts, or are damaged at the base.144 Of the 117, 59 have simply the ‘epidedōka’ subscription that one expects from petitioners literate enough to write their own names, 20 have a subscription with an ‘illiteracy formula’ attached, and 38 have a physical description. Thus, the natural inference is that around half of these 117 petitions were submitted by petitioners with some degree of literacy, and about half were submitted by illiterate petitioners. The 132 petitions containing subscriptions or physical descriptions really are likely to be a random sample: that is, it does not seem any more or less likely that petitions with (p.152) autograph subscriptions would survive than petitions with ‘illiteracy formulae’ or physical descriptions.

The significance of these subscriptions and physical descriptions becomes apparent when we compare them to those in another class of document that was routinely submitted to the authorities: census declarations. In these documents the practice was also for declarants to subscribe the document using an ‘epidedōka formula’. In their fundamental study of the census declarations, Hombert and Préaux found that two thirds of census declarations had an ‘epidedōka formula’ with an illiteracy formula, and only one third had such a formula with no indication of illiteracy.145 Bagnall and Frier's more recent study of the census declarations confirms that these proportions have been maintained in the declarations published since Homert and Préaux wrote.146 The inference is, therefore, that only a third of census declarants were literate in any sense. So the conclusion to all of this is that petitioners as a group tended to have somewhat higher literacy than census declarants.147

The question then is what conclusions can be drawn from this. The first thing to be said here is that illiteracy in Roman Egypt did not appear to carry the same stigma that it does in our own world.148 Many prosperous farmers, artisans, and businessmen managed to conduct their affairs perfectly successfully without being literate. There are also some known cases of minor government officials being illiterate.149 With all of this said, it is generally agreed that there was a rough correlation between literacy and social status: people (especially men) in the highest levels of the provincial and (p.153) metropolitan elites would mostly have been literate; amongst the very poor, literacy would have been quite rare. Thus, in one of his fundamental studies of literacy in Graeco‐Roman Egypt, Youtie observed that most of the people who appear as illiterate in the documents were members of what he calls the ‘lower middle class’, being ‘in general tradesmen, craftsmen, rivermen, village officials, and soldiers’.150 On the other hand, members of the gymnasial class were expected to be literate (even if this was not always entirely the case), and there might have been a similar expectation for bouleutai.151 The other thing to be said about literacy rates and social standing is that census declarants, although perhaps more representative of the general population than most individuals appearing in Greek papyri,152 still are not perfectly representative. The census would have probably still missed very poor people such as transients. A disproportionate number of metropolitan census declarations have survived.153 There is also the fact that declarants tend to be the heads of households, and hence are mostly adult males.

Thus, we can draw a number of conclusions. First, the mere fact that over 50 per cent of petitions with subscriptions were submitted by apparently literate petitioners is not consistent with a high rate of petitioning in the very bottom ranks of society. Secondly, the fact that there is apparently a higher rate of literacy amongst petitioners than amongst census declarants points in the same direction, especially since there was perhaps a slight social bias against the very poor even amongst the latter. Thirdly, the literacy rates that we can deduce from subscriptions to petitions suggest that the fact that petitioners known from archives are often literate did not make them terribly unusual. Of course, we cannot assume that all our petitioners spent their spare time making wry references to Callimachus: literacy is not an absolute skill, and many of the petitioners who subscribed their own petitions might have been capable of little more than writing their own names. Nevertheless, even this attainment allows some cautious inferences to be made about social position.

Another usable category of information relates to petitioners’ occupations. It is quite common for petitioners to mention their (p.154) primary occupation in the address section of the petition. Sometimes when such an explicit statement is missing, a reasonable inference about occupation can nevertheless be made from the narrative in the petition. Table 4.1 gives the main occupational categories for petitioners, and the number of petitions from people of each category.154 A number of features of the table deserve comment. First, 40 per cent of the petitions for which petitioners’ occupations are mentioned were submitted by farmers. This is surely lower than one would expect, given the central importance of agriculture as an occupation in this society,155 and the fact that around 70 per cent of the petitions come from villages. It seems highly unlikely that in Roman Egypt the number of priests, officials, estate administrators, and present or former soldiers combined would have been greater than the total number of farmers. No pre‐industrial, agricultural society can have supported so many unproductive members. Yet Table 4.1 has about as many farmers as it has people described as priests, officials, estate administrators, or present or former soldiers. The shortage of farmers could theoretically be explained by assuming that petitioners said to be engaged in unproductive occupations actually moonlighted as agricultural workers. For example, a man labelled as a priest in a petition might have engaged in agriculture when not carrying out his religious duties.156 This is possible; but we can nevertheless still say that the table below gives no positive support to any notion that the very poor petitioned just as frequently as those at other levels of society.

Secondly, Table 4.1 again supports my negative statement that petitioning was not confined to the highest ranks of the elite. (p.155)

4.1 Petitioners' occupations


Number of petitions









Estate administrators


State officials/public employees








Cameleers, artisans, and most farmers would most certainly not have fallen into this category. Thirdly, the spread of occupations again suggests that the petitioners whose archives I have examined above were not entirely unusual: there are many parallels for petitions being submitted by priests like Satabous, by farmers like Gemellus and Ptolemaios, by liturgists like Sokratēs, and by artisans like Tryphōn (indeed, eight of the thirteen artisans in the table are weavers).

There are two other categories of ‘occupation’ which, since they were not necessarily sources of income, have been left out of the table above, namely liturgical service and service as a municipal magistrate or member of a town council. Thirty petitions are submitted by people who held or had held liturgical offices. Furthermore, forty‐nine petitions contain complaints connected with liturgies, some of them submitted by the people who held or had previously held liturgical offices. These are mostly cases of people complaining that they were wrongfully nominated to liturgies. The mere fact that the responsible officers saw these individuals as worthy targets of nomination suggests at least that they were perceived in their local communities as being people of some substance.

Also prominent in petitions from certain times and places are men who hold or have held municipal offices and their families. One of the conventions of the opening formula of a petition is that the tenure of such offices by the petitioner or a husband or father is mentioned. Two examples illustrate this practice:

(p.156) To Aurelius Philōn strat(ēgos) of the divisions of The(mistos) and Pol‐(emōn) of the Arsinoite (nome) from Aurelius Sarapiōn, also known as Charmrōn, former agoranomos and bouleutēs of Arsinoē.157

To Aurelius P[rō]tarch[os], also known as Hērōn, stra(tēgos) of the Oxy‐(rhynchite) (nome) from Aurelia Sarapias, also known as Dionysarion, daughter of Apollophanēs, also known as Sarapammōn, former exēgētēs of Antinoopolis, acting without a guardian according to the ius liberorum.158

There are thirty petitions in the database in which people holding offices within their metropolis or polis (or their relatives) petition. These people are said to hold or have held a rich array of positions, and a number of them have held or hold two or more. The thirty petitions mention the following: thirteen bouleutai;159 eleven gymnasiarchs;160 six exēgētai;161 four (possibly five) agoronomoi;162 four kosmētai;163 four eutheniarchs;164 two epi tōn stemmatōn;165 two archiereis (civic priests) of Oxyrhynchus;166 one (p.157) prytanis;167 and one agōnothetēs.168 Petitions from such people and their families are perhaps the clearest cases of genuinely elite people using the justice system.

There is, for obvious reasons, a strong geographic pattern in these petitions: whenever we can discern the origin of the document, it is metropolitan. There is also a strong chronological pattern: none of these texts dates to the first century, only six date to the second, and the remaining twenty‐four date to the third century. Again, this is unremarkable, and is unlikely to have been the consequence of a genuine change in the social level of petitioners and litigants. Rather, it is simply a reflection of the fact that evidence of municipal offices is scarce to non‐existent for the first century or so of Roman rule, depending on the office, and only starts to increase in the second century. This scarcity of evidence could be read to suggest that many of these offices did not exist as municipal offices until some stage late in the first century, at least outside Alexandria.169 In the case of town councils, it is known that in the nome metropoleis and in Alexandria there were no town councils until the early third century.

Another indication of the social status of the petitioners is their tenure of immovable property. In many petitions, the object of dispute is either land or a house, or such property is relevant in some other way to the dispute. There are thirty cases in which the petition makes claims in relation to the ownership of a house or part thereof.170 As far as agricultural land is concerned, sometimes the mere fact of ownership of such property is mentioned, without any attempt to quantify the size of the land owned. There are twenty‐nine cases that fall into this category.171 Moreover, in another twenty‐six petitions, there is internal evidence for the precise amount of land (p.158)

4.2 Land areas in petitions

Land area (ar.)

Number of petitions

0 to 4


5 to 19


20 to 99


100 +


owned or claimed by the petitioner (or the family member that he or she is representing). These references to precise land areas are indicated in Table 4.2 above. (This table excludes the petitions submitted by the various petitioners for whom we have archival evidence, since their landholdings have been discussed above.)

One wishes that we could put this data in context by comparing land areas in our petitions with what we know about general patterns of landholding in Roman Egypt.176 Such an exercise is hampered considerably by the fact that, for all we know, our petitioners might have owned other plots of land that are not mentioned in their petitions. There is also the fact that, with the exception of data from Theadelphia in AD 164/5 and Philadelphia in AD 216, the cases in which we have such knowledge of the distribution of land in a particular area are strictly outside our period.177 The only thing that can perhaps be said is this: there is no evidence in the above table (p.159) that landowning petitioners owned land in radically different patterns from landowners in Egypt. If land areas in the petitions were all in the hundreds of arouras, or were all below five, this would be striking. But there is instead a mix of comparatively large holdings and fairly small ones, and a few in between. More positively, we can perhaps say that this data helps to put the petitioners whose affairs are known from archival evidence into some kind of context. They show that the amounts of agricultural land owned by Gemellus and Ptolemaios and their families were not dissimilar to those owned by many other petitioners. Furthermore, the ownership by Satabous, Tryphōn, and Ptolemaios or their families of urban real estate is a feature shared by many other petitioners as well.

How should we explain the social profile of petitioners?

My suggestion that people of the very lowest social strata tended not to petition raises the question of what factors limited the participation of such people. After all, the rich have no monopoly on disputes and conflicts. A number of hypotheses could be raised here. It could, for instance, have been that people of middling and higher levels of society had a stronger ideological commitment to the Roman state and its justice system, and thus were more confident in calling on it for assistance in their disputes. Such people were, after all, ‘stakeholders’ in the Roman order, whose property rights, contracts, and privileged tax statuses were ultimately guaranteed by the Roman justice system.

This scenario sounds plausible, but it seems impossible to prove. Evidence simply does not exist to show that the very poor were intrinsically more suspicious of state apparatus than those of a higher socio‐economic position. We certainly cannot infer mentality from the mere fact of participation (or non‐participation) in the justice system. In fact, two nuggets of information from archives of petitioners caution against such an inference. First, apart from having multiple petitions in his archive, Satabous also had a copy of the (p.160) Demotic prophetic text ‘The Lamb of Bocchoris’, written in his own hand.178 This text contains fantasies about a dramatic end to an epoch whose evils include foreign domination, and about the restoration of law and order under indigenous rule. These are hardly bedtime reading for a wholehearted supporter of the Roman order.179 There is something similar in the archive of Sokratēs. The specimen of the Acta Alexandrinorum that it contains tells a story of cruelty by Roman soldiers and the Roman prefect.180 As Peter van Minnen has aptly commented, this fact ‘illustrates the ambivalent position of the local elite in a Greco‐Roman town rather well’.181

Another possible restraint on the very poor using the justice system was perhaps the cost of doing so. Having petitions drafted by a scribe often cost money, and if the matter went to the stage of a hearing, a plaintiff potentially faced the costs associated with travel if the hearing was away from his or her domicile.182 A hint of the inconvenience of travel prompted by lawsuits is found in a private letter of the late second century AD:

Taōs to Agathōpous his dearest, greetings. I wish you to know that Taharmiusis summonsed us, me and Seias and Faustiōn and the wool‐dealer and his wife. They said that I in particular attacked her together with you and an assistant (hypēretēs) when she was feasting, and that we have taken away golden things from her—so she says. You will do well, therefore, if you take care of me, just as you have done of the others—not only because the matter is your own, but also because I trust in you that you will not trouble me (to go) to Alexandria.183

Several petitioners also complain about the inconvenience of having to wait in a particular location away from their home, even though their opponent has failed to attend the court.184 Being absent at critical points during the cycle of the agricultural year is mentioned as a particular burden, and this would have been one that the very poor were least able to bear.

(p.161) But we should not overstate the financial burden involved in going to law. For one thing, the cost of having petitions drafted would not appear to have been enormous, to judge from evidence preserved in the archive of Kroniōn, a head of the grapheion of Tebtynis in the first century AD. Two registers from this archive record the price charged for documents drawn up in the grapheion during AD 45–7; these documents seem to have included a number of petitions (although the relevant words on the register can also refer to other types of documents). Mostly, the prices were in the range of a few obols, although some seem to have been drawn up for free.185 As for the costs of travel, these were only relevant if an official was approached who was likely to hear the case elsewhere. Yet, as we have seen, a variety of more local officials such as soldiers and stratēgoi regularly exercised what essentially amounted to adjudicative powers (even if they did not strictly have these, unless they were delegated by the prefect). It was also possible to submit petitions through a representative,186 and to be represented in court in one's absence.187 And it should be remembered that not all petitions resulted in a court hearing: some asked the addressee to do something that did not require a hearing; and some petitions, as we shall see, were used in a strategic (or even vexatious) way, and the petitioner might never have intended to take the matter all the way to a hearing.

Apart from the cost of having petitions written and of travelling, there is little clear evidence of litigation costs. For the Roman empire of late Antiquity, imperial legislation preserved in the law codes and in a recently discovered inscription from Caesarea Maritima gives clear evidence that the government permitted the charging of fees at various steps of a lawsuit. Some of these fees were quite high, especially for the poor.188 But there is little evidence for the systematic levying of such charges in an earlier period. This does not necessarily mean that the practice did not occur: the charges that the government of the later Roman empire sought to regulate possibly began as an informal practice of sportulae (‘tips’) given to court (p.162) staff.189 We would not therefore expect them to leave a trail in the documentary record.

For Roman Egypt during the Principate, there is certainly some evidence of charges being levied on litigants in a few isolated contexts. It would appear that in Roman Egypt, as elsewhere during the Principate, litigants were sometimes required to deposit sums by way of sureties. But most of the attested cases relate to defendants being required to guarantee their appearance with a surety,190 or appellants the good faith of their appeal.191 There is little evidence for sureties being required of plaintiffs. One third‐century petition complains that the petitioner has been summonsed by a certain Petronius, son of Deios, to appear before the dioikētēs in Alexandria. As well as requesting a delay for this hearing, the petitioner complains that Petronius ‘is neither a public prosecutor nor has he given security to the treasury for the fine for vexatious accusations’.192 The precise nature of the accusation is not specified, but it is generally assumed that it must be fiscal.193 Also relevant is a text from which we learn that, as part of the standard procedure for execution on a defaulting debtor, a 12 dr. fee was charged for the registration of the loan contract.194 But these few scraps of evidence are certainly not enough to allow us to conclude that in the general run of cases, plaintiffs in (p.163)

4.3 Factual claims made in petitions and reports of proceedings

Type of Dispute

Number in Petitions

Number in Reports of Proceedings

Debt dispute



Illegal pasturing



Liturgical dispute



Misconduct by an official






Property damage



Real property



Succession dispute












Egypt in the first three centuries AD faced the same extensive collection of court charges as their late antique counterparts. The conclusion must be, therefore, that although there were undoubtedly costs involved in litigation, our evidence does not quite allow us to conclude that the expenses of litigation were the sole reason for the limited social diversity of petitioners.

Moreover, there seems to have been another factor that contributes at least part of the explanation for the social profile of petitioners. There are signs that there was a cultural preference towards using the legal system mainly to deal with disputes with an economic element. The vast majority of petitions and reports of proceedings deal, at least on their faces, with disputes regarding damage to, or the ownership, transfer, or theft of property. To illustrate this, I have in Appendices I and III assigned each petition and report of proceedings to one or several categories of dispute. Table 4.3 records the number of times that each offence type occurs in these Appendices, with the uncertain cases excluded. As will be obvious, there are more occurrences of each dispute category than there are petitions. This is because some petitions tell tales of disputes that have elements in several categories, so some petitions effectively appear multiple times in Table 4.3. The crudity of this exercise should be stressed: even if one assigns petitions to multiple categories of dispute, one still risks simplifying the narrative of many petitions. It should also be said that these categories are modern ones: they represent a typology of factual claims, a typology that is useful for sociological analysis and does not necessarily reflect attempts by the authors of the documents to present (p.164) complaints in terms of neat and easily identifiable juridical categories. (Indeed, the language of the petitions is often so vague about the legal basis of claims that in many cases it would be impossible to categorize them by reference to the juridical category in which their authors attempted to place them.) Nevertheless, categorizing the petitions by modern types gives an overall idea of the factual claims made by petitions. The other thing to emphasize is that this table really only represents what the narratives in petitions say the dispute was about. It in no way represents what the disputes were ‘really’ about. In many of these cases there could well have been all manner of complex elements to the dispute which were considered inappropriate for official attention, and hence not mentioned in the petitions. This is not, however, enormously important for present purposes, since the question is what kinds of issues were thought appropriate for such attention.

As is immediately apparent from this table, a very high proportion of the complaints in petitions relate to property. Many petitions present the dispute as arising from transactions relating to the transfer of property (debt and succession), interferences with rights to property (theft and real property disputes), or damage to property (illegal pasturing and property damage). Liturgical disputes are often essentially disputes over property, given that liturgical service could be financially draining. The category of ‘other’ includes a good number of marital disputes that are essentially cases in which a couple in the midst of a divorce disagree over common property or dowries. In the ‘other’ category, especially in the reports of proceedings, disputes over taxation and revenues due to priests are also quite common. For petitions, the second largest category of wrongs is violence,195 (p.165) but it should be noted that over half (seventy‐seven) of the allegations about violence also contain an underlying economic dispute, or a dispute involving property. For example, some claim that the petitioner was assaulted in the course of being robbed, or was assaulted in the course of a dispute over a debt or an inheritance.

It is instructive to reflect on the types of disputes that are not in this table, but which were commonly brought before the courts in other premodern societies. Disputes relating to sexual conduct are noticeably absent. There are no petitions or reports of proceedings from our period relating to rape or sexual assault.196 Nor are there any signs of complaints to the authorities about adultery: petitions and litigation relating to marital disputes tend to focus on matters of property. There is also a near total absence of petitions relating to magic and witchcraft accusations: there is only one petition in the database with what appears to be an accusation of magic;197 and there are no clear cases of trials relating to magic. As we have seen, there is also a shortage of complaints relating to agricultural tenancies, and disputes between employee and employer were rarely brought to the attention of the authorities.198

Our evidence does not allow us to see in detail the cultural attitudes that kept these disputes from official attention. One may hypothesize that a concern for privacy kept sexual matters out of petitions and the courtroom. Likewise, it could be that the relations of social dependence and patronage that surrounded tenancy and employment meant that disputes that arose from these relationships were dealt (p.166) with using avenues more subtle and private than the legal system. Perhaps magic was mostly kept out of the legal realm thanks to a concern about the capacity of the state to provide an effective defence against the dark arts. The mentalities of administrators and judges might also have played their part: it could be that it was widely recognized that some matters would not be taken seriously by the courts. This suggestion could be supported by the fact that cases of assault, illegal pasturing, and petty theft, although common in petitions, appear in reports of proceedings fairly rarely. (Although it could be that there was less reason to preserve reports of hearings regarding more petty ‘criminal’ matters like these, in which title to substantial amounts of property was not involved.) But the truth is, these sorts of explanations can only be hypothetical. Whatever the case, it seems clear enough that there were cultural aversions against official involvement in many of the types of dispute in which all members of society, including the very poor, were likely to be involved. The matters that were thought fit for petitioning and litigation tended to be precisely the sorts of disputes in which the really poor would be less likely to be involved, by the very fact of their poverty.


Whilst our information regarding the ethnic identities and socio‐economic status of petitioners is far from ideal, some conclusions are nevertheless possible. When we have archives to give us information about the affairs of the petitioners, these give us a picture of moderately prosperous people owning property, working agricultural land, engaging in small‐time entrepreneurial activities, holding minor official or liturgical positions, and, sometimes, being literate. These people are not all of an identical level: one would place Sokratēs at a higher social level than Tryphōn, for instance. But none is a member of the very highest strata of the elite of the province.

The internal evidence of petitions in general suggests that the petitioners whom we encounter in archives were not unusual in respect of their ethnic backgrounds or economic statuses. There is no sign that people of any particular ethnic background tended mostly to avoid petitioning and litigation. Casual references in the petitions also give independent support to my suggestion that the very (p.167) poor tended not to participate as much in petitioning: the level of literacy amongst petitioners in general seems to have been higher than what we would expect from the population at large; farmers seem to be underrepresented in occupational descriptions; and owners of property, municipal officials, and members of town councils are quite well attested. But at the same time, these casual references show that petitioning and litigation were not the exclusive preserve of the astonishingly rich.

All of this adds nuance to our discussion of the social impact of the justice system. It is important to know that some categories of dispute were virtually never brought to official attention. This means that there were some facets of social relations on which the institutions of state adjudication would have had a negligible impact. Furthermore, the fact that petitioners and litigants were a somewhat socially exclusive group means that any conclusions that we draw are qualified by the fact that they mainly apply to certain sections of the population. At the same time, what we know about the social position of petitioners and litigants means that the justice system cannot simply be dismissed as a realm where very privileged people fought battles over enormous areas of land or huge sums of money. This system had meaning for a much broader segment of the population than that.


(1) For a discussion of the economic and ethnic backgrounds of Ptolemaic petitioners, see Hengstl 1997: 280–1.

(2) In this sense, I give systematic support to John Crook's impression that the people who appeared in court tended to be members of ‘the Greco‐Egyptian middle class (to use that phrase in a general and perfectly intelligible sense). They have a little property and are concerned about its disposition and the taxes on it and about loans and dykes and local violence’ (1995: 63). See also Kehoe 1995, who has a slightly different emphasis, although one that is perhaps not in a strict sense contradictory, stating, for example, that: ‘[t]he numerous petitions preserved on papyri make clear that people from a wide range of backgrounds looked on petitioning governmental authorities as a means to solve legal problems’ (237).

(3) Bagnall 1992; 2005: 198; Sharp 1999: 171.

(4) Bagnall 2005: 198–9; van Minnen 1986; 1987.

(5) Kehoe 1995: 262.

(6) Above, 60–2.

(7) Bagnall 1993b: 131–2; Kehoe 1995: 236; see too Rowlandson 1996: 97–101 on the social status of tenants of public land.

(8) Chrest.Wilck. 376; P.Bingen 58; P.Giss. 46; cf. P.Lond. III 895 (pp. 130), BL I 281, II.2 83; SB I 5672. For disputes arising from leases of property other than agricultural land: BGU XI 2066; Chrest.Wilck.176; P.Mich. V 226. In P.Ryl. II 293 descr. the nature of the lease is unclear. There are also several petitions and reports of proceedings involving disputes between tenants: P.Oxy. IV 707; P.Sakaon 37; cf. P.Oxy. XIV 1630. Kehoe also discusses petitions in which a tenant asks an official to notify his landlord that he is withdrawing from a lease (above, 122 n. 193 for references and further discussion). In these cases, there was not necessarily a dispute afoot, and the tenants were not asking for the positive enforcement of rights. Rather, they were creating an evidentiary shield which they could use if sued by their landlords.

(9) Bagnall 1989: 211–12, p. 211 for the quotation. See too Drexhage 1988b: 314–16.

(10) Schentuleit 2007: 109–12.

(11) SB XII 10882.36–7; see too Schentuleit 2007: 114–15.

(12) P.Vind.Sal. 4.6–7; cf. CPR XV 5.4 (= SB I 5234).

(13) Schentuleit 2001: part C, line 1; part D, line 4 (= P.Lond. II 262 [p. 177] = Chrest.Mitt. 181); CPR XV 2.3, 10–11, 14; 3.3; 4.8; SB I 5231.3; 5232.5–6, BL II.2 118; SB I 5240.i.11–13, BL II.2 119, IX 243; SB I 5275.3–4

(14) SB I 5237.5; 5240.i.17–18, ii.24; CPR XV 6.8; P.Vind.Sal. 3; see above 4 n. 16 for literature.

(15) CPR XV 1; SB I 5235.9–10; 5238.4–6.

(16) CPR XV 8; 10a.

(17) P.Wein D inv. 6823 = Migahid 2003: 328 with Schentuleit 2007: 109–10 and n. 52 and P.Dime II, p. 13 n. 120.

(18) P.Dime II 56–64, discussed by Schentuleit (2007: 109).

(19) For the role and duties of the Scribe of the Priests in the Temple of Soknopaios, see Schentuleit 2007: 109. For the financial affairs of the Temple in general, see Lippert and Schentuleit 2005.

(20) P.Harrauer 32.

(21) CPR XV 1.10 with commentary to line 10.

(22) P.Vind.Tand. 25.

(23) P.Zauzich 39. See too commentary at pp. 339–40.

(24) P.Stras. IV 265; 208; with P.Zauzich, pp. 367–8.

(25) BGU I 184, with P.Zauzich, p. 368.

(26) P.Lond. II 140 (p. 180), BL I 259–60; see too P.Zauzich, p. 361 n. 26.

(27) Chrest.Wilck. 176; cf. Chrest.Wilck. 312.

(28) P.Vind.Tand., p. 173 for the identification. For doubts, Schentuleit 2007: 116. For onomastic practices in Soknopaiou Nēsos: Jördens 2005: 45.

(29) Chrest.Wilck. 312.9–11.

(30) P.Ryl. II 161; Schentuleit 2007: 116–17.

(31) On the archive, see Brewster 1927; 1931; Biscottini 1966; Whitehorne 1984; Gagos, Koenen, and McNellen 1992: 189–92; Rowlandson 1998: 112–17; Piccolo 2003.

(32) P.Oxy. I 38; II 282; SB X 10239; 10244; 10245.

(33) See Whitehorne 1984: 1273–4 for Tryphōn's financial position in the mid‐30s AD.

(34) Tryphōn: P.Oxy. II 288; SB X 10243.i.7, ii.7. Tryphōn's grandfather, Tryphōn: SB X 10220.2. Tryphōn's father, Didymos: SB X 10220.4. Tryphōn's uncle, Thoōnis: SB X 10220.7. Tryphōn's brother, Onnōphris: SB X 10236 = P.Oxy. II 322 descr.

(35) P.Oxy. II 267 = Chrest.Mitt. 281.26–7; SB X 10244.10–12.

(36) P.Oxy. I 37 = Chrest.Mitt. 79.

(37) SB X 10223.2–3.

(38) SB X 10245.

(39) P.Oxy. I 99. For the loan and its date, see SB X 10246 = P.Oxy. II 304 descr., with Biscottini 1966: 253, 258.

(40) The purchase is mentioned at SB X 10249.13–16 (cf. BL VII 218) = P.Oxy. II 318 descr. and P.Cair.Preis. 43.24–8 = P.Oxy. II 306 descr. The loans are documented by SB X 10248 = P.Oxy. II 320, recto, descr.

(41) For the loans that Tryphōn apparently took out to buy real estate, see above, nn. 39 and 40. He is also found loaning money to a person who does not appear to have been a family member in P.Oxy. II 269 = Sel.Pap. I 69.

(42) Note too SB X 10246 = P.Oxy. II 304 descr., in which the creditor might have been the cousin of Tryphōn (P.Oxy. II, p. 288; Biscottini 1966: 255).

(43) SB X 10238, BL VII 217 = P.Oxy. II 319 descr.

(44) P.Oxy. II 267.20–1, reconstructed by Whitehorne 1984: 1271 n. 14.

(45) Whitehorne 1984. See too Gagos, Koenen, and McNellen 1992: 189–91.

(46) SB X 10235 = XIV 11415 = C.Pap.Gr. I 16 = P.Oxy. II 321 descr.

(47) Whitehorne 1984: 1268–9; Gagos, Koenen, and McNellen 1992: 189 n. 36.

(48) P.Mich. III 174; XI 617; PSI VII 737; 806 (cf. BL V 124); PSI XIII 1323; P.Wisc. I 33; 34 (dupl. = 35); SB XIV 12087; XX 14401. P.Wisc. I 33 contains two separate petitions in the same document; SB XIV 12087 contains three.

(49) P.Mich. III 174.23; PSI VII 737.17; P.Wisc. I 33.26; SB XIV 12087.A.16–17. Whitehorne (1991: 251) suggests on the basis of a comparison with Ptolemaios’ subscription in P.Mich. III 174 that P.Mich. XI 617 and SB XX 14401 were drafted by Ptolemaios himself.

(50) On the subject of Ptolemaios’ family, it should be mentioned that the suggestion that his father was a veteran (P.Wisc. I, p. 122) has been invalidated by the revised reconstructions of P.Wisc. I 33.2–3 proposed by D. Hagedorn (reported at BL VI 73) and P. Sänger (Korr.Tyche 598). Bryen and Wypustek (2009: 541 with n. 12) appear to support the reconstruction originally proposed by the editor of P.Wisc.

(51) P.Wisc. I 36.

(52) BGU IX 1896.271–4.

(53) See Sijpesteijn, P.Wisc. I, pp. 119–21 for references and discussion. Youtie (1976b: 135) treats these texts with rather more scepticism, suggesting that the individuals in the tax lists could be homonymous individuals with no relation to Ptolemaios or his family.

(54) P.Mich. III 174.12–13.

(55) P.Wisc. I 34.4 (dupl. = I 35.5–6); P.Mich. XI 617.2–3. For the nature and uses of the drymos, see P.Wisc. I, pp. 125–6.

(56) P.Mil. II 65.12 = SB VI 9014. Sijpesteijn's suggestion (P.Wisc. I, p. 121) that this Ptolemaios, son of Dioskoros, is actually a nephew of our Ptolemaios seems precarious. A Ptolemaios, son of Dioskoros, appears in PSI VII 775.3 (AD 138) as a ‘member of the board of supervisors of pastures and marshland’ at Theadelphia. This must be Ptolemaios, son of Diodōros, also known as Dioskoros, since he is known to have held the same post in the same village at the same time: P.Oslo III 89.28; 90.17–18. Thus P.Mil. II 65.12 could simply be another case in which his father's alias appears alone in the filiation.

(57) P.Leit. 14 (AD 148); PSI VII 735 (AD 138); P.Wisc. I 37 (AD 148); P.Oslo III 89 (AD 138); 90 (dupl. = PSI III 160) (AD 138); 91 (AD 149).

(58) P.Mich. XI 617 (AD 145/6); PSI VII 806 (AD 158); SB XIV 12087 A.1–4, B.6–7 (AD 162).

(59) P.Mich. III 174.17–19 (AD 144–7); P.Wisc. I 33.6–7, 22–3 (AD 147); SB XIV 12087.A.15–16 (AD 162); XX 14401.23–31 (AD 147).

(60) SB XX 14401.14–16.

(61) P.Oxy. XLIII 3089 was probably also submitted to this official: see commentary at P.Oxy. XLIII, pp. 5–6.

(62) For the archive, see Strassi Zaccaria 1991; van Minnen 1994: 237–49; Harker 2008: 113–14; Clarysse 2009.

(63) SB XVIII 13306.

(64) Anagnostou‐Cañas 1984: 358.

(65) P.Mich. VIII 507.

(66) For the latter position, see: SB VI 9554 (2), (3), and (5); 9555 (a), with van Minnen 1994: 242.

(67) Receipts: BGU II 391; XV 2534; SB VI 9428; XVI 12798; cf. P.Mich. VI 383.iv.33. Rolls: P.Cair.Mich. 359; P.Mich. IV 223–5; 357–63.

(68) P.Mich. IV 223, with Van Minnen 1994: 244–5.

(69) Schuman 1975; van Minnen 1994: 246.

(70) SB XVIII 13306.5–7.

(71) P.Kar.Goodsp. 78.

(72) P.Mich. IX 564, BL VII 113.

(73) Van Minnen 1994: 242.

(74) P.Mich. III 169 = FIRA III 4 = CPL 162, with Strassi Zaccaria 1991: 258–9; van Minnen 1994: 241.

(75) P.Mich. VI 419, with BL V 70; Korr.Tyche 584.

(76) P.Mich. VI 392; 393, with van Minnen 1994: 241.

(77) BGU III 819, BL XI 20; P.Aberd. 35, with van Minnen 1994: 242.

(78) BGU I 97; II 577, with van Minnen 1994: 242.

(79) Van Minnen 1994: 245–6.

(80) Youtie 1970a.

(81) The earliest document mentioning a family member is P.Mich. VI 428; however, a title deed of a house owned by C. Julius Niger dates from AD 93: P.Mich. IX 554. The latest published document is P.Sijp. 12f.

(82) P.Mich VI 422 (dupl. = SB XXII 15774); 423 (dupl. = P.Mich. VI 424); 425 (dupl.? = P.Mich.inv. 2926a, see P.Mich. VI, p. 127 ); 426; P.Sijp. 12f; SB XIV 11478. P.Mich. VI 425 preserves two petitions, one to the prefect and one to the epistratēgos. P.Mich. VI 426 preserves two petitions, one to the epistratēgos, and the start of another to a second official, perhaps the a censibus (BL VI 82).

(83) P.Mich. VI 425.25–6; cf. 422.39–40; 423.27.

(84) SB XXII 15781; XXIV 16252.

(85) SB V 7558.

(86) Gemellus: P.Mich. VI 422.4 (dupl. = SB XXII 15774.2–3); P.Mich. VI 423.2; 425.2; 426.15–16; 370.13. Mother (Tasoucharion): P.Mich. VI 370.10. Father (C. Apolinarius Niger): P.Mich. VI 364.4. Grandfather (C. Julius Niger): SB XXIV 16252.2; SB XII 11103.1 = P.Mich.Mchl. 1. Sister (Gaia Apolinaria): P.Mich. VI 370.13. Uncle (C. Julius Longinus): SB IV 7427; XII 11103 = P.Mich.Mchl. 1.

(87) Bieżuńska‐Małowist 1957–8: 158–9, 163–4.

(88) Note, however, SB IV 7360.2 (AD 214), in which a Gaius Gemellus Horigenes makes a registration of taxable land, and gives details about himself which make it virtually certain that this is our Gemellus. It is not clear why, if he was only enfranchised by the constitutio Antoniniana, he did not take the gentilicium (family name) of Aurelius. Various theories have been proposed to explain this, each of which is open to objections: Bieżuńska‐Małowist 1957–8: 162–3; Alston 1995: 130–1; Bryen and Wypustek 2009: 540 n. 10.

(89) Kerkesoucha: P.Mich. VI 397 (AD 204); Karanis: P.Mich. VI 398 (AD 207). These texts record the payment of 4 art. and 1 art. of grain respectively by way of tax. We can extrapolate the size of the two plots on the assumption that the standard tax of one art. per ar. was charged on this land (for this standard charge, see Rowlandson 1996: 29, 54).

(90) See Rowlandson 1996: 30, 61–2.

(91) Gemellus’ father and uncle were both dead by c. AD 189 (P.Mich. VI 422.8–16). His grandfather, C. Julius Niger, is last attested as alive in AD 188 (P.Mich. VI 387), and probably died soon after: in AD 189, a house that once belonged to him is found in the possession of Gemellus and his sister (P.Mich. VI 370.12–15; cf. Bieżuńska‐Małowist 1957–8: 158). The fact that in AD 200/1 C. Julius Niger would have been in his late nineties (cf. P.Mich. VI 428.2) also makes it unlikely he was still alive to own property.

(92) SB IV 7360.

(93) P.Mich. VI 423.9–10 (AD 197).

(94) P. Mich. inv. 2914, cited at P. Mich. VI, p. 119.

(95) P.Mich. VI 370 (AD 189). P.Mich. VI 428 (AD 154) makes it clear that one of the three houses was inherited from Gemellus’ paternal grandfather, C. Julius Niger. We also learn from P.Mich. VI 370.11 that Gemellus’ mother, Tasoucharion, owned a house, a courtyard, and a one‐third share in another house in AD 189. Presumably when she died, these were inherited by Gemellus and his sister (assuming she was still alive then).

(96) See SB IV 7360.4 for a specific mention of his inheritance of land.

(97) Olive grove: SB XXIV 16252.11 (AD 163); katoikic land: P.Mich. VI 395.15–17 (AD 183).

(98) P.Mich. IV 225.2997–3001, with P.Cair.Mich. II, pp. 30–1, 41–2 for the relevant tax rates. The characteristics of land en aphesei are obscure, but it has been suggested that in Roman Egypt the label sometimes was attached to land that had been confiscated and reallocated by the state in the Ptolemaic or early Roman periods: Rowlandson 1996: 44 n. 48.

(99) P.Mich. VI 385, with analysis at p. 75.

(100) SB XII 11109.11–12.

(101) 9 1/12 ar.: P.Mich. VI 395.9–11 (AD 183). Five‐eighths of an aroura: P.Mich. VI 364 (AD 179).

(102) P.Mich. VI 395.ii.4–7.

(103) SB XXIV 16252.10–11.

(104) P.Mich. VI 364.5–7.

(105) P.Mich. VI 423.13.

(106) P.Mich. VI 398.14–15; cf. P. Mich. VI, p. 118, and below, 320.

(107) P.Sijp. 12f; SB XXII 15781.

(108) SB XXIV 16252.

(109) P.Mich. VI 425. The text simply calls Kastōr, the offending official, the χειριϲτήϲ of Karanis (line 10). The text goes on to claim that Gemellus and his mother owe nothing to the treasury, and that Kastōr had provided him with no receipt (lines 17–19), so clearly he was a tax collection official of some kind.

(110) SB XIV 11478.

(111) P.Mich. VI 426, BL VI 82.

(112) P.Mich. VI 422; 423; see below, 318–21 for further discussion of the case.

(113) I thank the Press's anonymous reader for raising this stimulating question.

(114) Note Hengstl 1997: 281 on the ethnicity of petitioners in the Ptolemaic period: the evidence does not seem to suggest a conspicuous avoidance of the justice system by Egyptians in that period.

(115) Thompson 2001: 303–4 for a brief survey and references.

(116) BGU XIV 2367; P.Hamb. II 168; P.Rev. vii.3; discussed by Bagnall 1988: 22; Delia 1991: 22–3; Thompson 2001: 304–6.

(117) BGU VI 1213.3; 1250.11–14, discussed by Clarysse and Thompson 2006: 2.146.

(118) Thompson 2001: 306–11; Clarysse and Thompson 2006: 2.138–47.

(119) Clarysse and Thompson 2006: 2.144–5.

(120) E.g. P.Enteux. 79.9–10; UPZ I 7.21–2; 8.14; 15.16–17.

(121) Martin 1956: 89; Clarysse 1985; Goudriaan 1988: 60–7, 70–87, 91.

(122) Martin 1956; Clarysse 1985.

(123) Above, 27–30. Note, however, that even before the constitutio Antoniniana, a few people were simultaneously Roman citizens and citizens of Greek poleis.

(124) Although the right enjoyed by citizens of Antinoopolis to intermarry with ‘Egyptians’ perhaps complicated the issue on occasion. For this right of intermarriage, see Delia 1991: 19–20, 33–4 and literature there.

(125) Bagnall 1988: 21–2; Delia 1991: 15.

(126) Bagnall 1988: 21–2.

(127) Roman: BGU I 195; VII 1574; 1575; XI 2012; 2013; XV 2458; Chrest.Mitt. 60; 89; 121; 227; 239; Chrest.Wilck. 29; 376; 396; 461; P.Fay. 106.6–25; P.Fouad 28, BL IX 88; P.Mich. IX 526; PSI XIII 1328; P.Warr. 1; SB I 4415; IV 7367; VI 9105; XVI 12678; 12747; XXII 15781; XXIV 16252.

(128) Antinoites: BGU I 275; P.Col. X 266; P.Fam.Tebt. 29; 37; 38; 43; P.Lips. II 145, recto; 146; P.Tebt. II 334; P.Würzb. 9; cf. P.Tebt. II 335, BL I 427, XI 278. In a number of petitions the petitioners are both Antinoites and Romans: BGU VII 1575; Chrest.Wilck. 29; P.Bingen 107; P.Mich. IX 526; P.Oxy. VIII 1119; XVII 2130; P.Turner 41; P.Vind.Tand. 2; 3; P.Warr. 1; P.Wisc. II 86; and SB XVI 12678. Gemellus’ petitions also explicitly state that he is an Antinoite: P.Mich. VI 422; 423; 425; 426.

(129) Alexandrians: BGU XIII 2243; P.Erl. 27; P.Oxy. IV 712; L 3561; cf. BGU XV 2459; CPJ II 151 (with comm. at pp. 29–30 and Delia 1991: 26); P.Giss. 34; P.Tebt. II 335, BL I 427, XI 278. Several petitions involve petitioners who are both Roman and Alexandrian citizens: P.Flor. III 382; PSI X 1102; XIII 1328; P.Turner 34; SB XXIV 16251 (partly reconstructed).

(130) Metropolites: BGU I 181; 340; III 908; Chrest.Mitt. 111; 112; 116; Chrest.Wilck. 268; CPR I 20.ii; P.Alex.Giss. 31; P.Bodl. I 40; P.Fay. 108; P.Flor. I 6; P.Fouad 27; P.Gen. I2 4; II 107; P.Giss. 9; P.Heid. IV 325, recto; P.Köln III 143; P.Mich. XVIII 787; P.Münch. III 73; P.Oxy. II 285; VII 1032; XVII 2131; XXXVI 2758; XLI 2997; XLVII 3364, BL IX 202 (restored); XLIX 3466; 3468; LIX 3978; LXV 4481; LXVII 4582; PSI X 1103; P.Ryl. II 113; 116; P.Tebt. II 435 descr.; SB IV 7469; V 7870; VI 9238; XII 11008; XIV 11902; XX 14229; XXVI 16526. Tryphōn's petitions also designate him as a metropolite: SB X 10239; 10245.

(131) See, for example, BGU XI 2068; P.Bour. 21; P.Brem. 35; P.Col. VIII 209; P.Gen. I2 6; P.Grenf. I 47; P.Iand. III 27; P.Kron. 6; P.Oxy. III 484; 488; XLIII 3089; P.Ryl. II 133; P.Tebt. II 330; 331; 332; SB XXII 15831; Stud.Pal. XXII 54.

(132) BGU XI 2070; Chrest.Mitt. 87; 372.i.5–13; P.Hamb. I 29.20–6, BL III 75; P.Lips. II 147; P.Stras. I 22.25–33; SB XVI 12555; cf. BGU III 705; Chrest.Mitt. 372.vi.

(133) Antinoites: P.Mich. VI 365; SB V 7558.12–31; 7601. The plaintiffs in P.Lips. II 147 are both Roman citizens and Antinoites. In SB VI 9016, the boulē of Ptolemais is involved in various bouts of litigation. In P.Giss. 34.10–16 an Alexandrian citizen litigates on behalf of his granddaughter, who herself could possibly be an Alexandrian. In Chrest.Mitt. 372.iv.16–v.26 the plaintiff is both an Alexandrian citizen and (probably) a Roman citizen.

(134) P.Mil.Vogl. I 25.ii.4 P.Oxy. I 37; VIII 1102; P.Ryl. II 77; SB I 5676.11–19; XXIV 15940.

(135) BGU VII 1567.B.15; Chrest.Mitt. 93.36; P.Ant. II 87.12; P.Oxy. II 237.vii.37–8; XLII 3074.7 (?); P.Sakaon 32.ii.23, 33; PSI XIII 1326.4; P.Vind.Tand. 8.2–4; SB XIV 11391.6; XVIII 13156.7; cf. PSI V 450, recto, ii.34.

(136) BGU I 321; XIII 2239; CPR XV 17; P.Bacch. 19; 21; 22; P.Louvre I 1; 2; 3; P.Lund. IV 1; P.Mich. III 175; V 226; PSI XV 1529; PSI Congr.XI 10; P.Tebt. II 302; 303; 304; SB VI 9458; X 10564; XVI 12685; 12833; Stud.Pal. XXII 55.

(137) BGU I 36; XIII 2242; Chrest.Mitt. 118; 125; P.Amh. II 77; P.Harrauer 34; P.Mich. XI 618; SB VI 8979; XX 14098; 14313; XXIV 16257.i.

(138) CPJ II 151; III 455.

(139) Van Minnen 1998b.

(140) P.Mich. IX 523.18–19.

(141) P.Oxy. I 69.21–3.

(142) Several petitions state explicitly that a physical description was used to cope with the petitioner's illiteracy: SB IV 7469.11–13; Stud.Pal. XXII 55.21–3, BL VIII 482; Chrest.Wilck. 268.27–8, BL III 13. In several other documents that are not petitions, but which required subscriptions, an express connection is also drawn between physical descriptions and illiteracy: Chrest.Wilck. 31.23–4; BGU I 17.24–6 = C.Pap.Gr. II 38.

(143) Three petitions submitted by groups have a mixture of formulae to reflect the fact that one or more petitioners are literate and one or more illiterate. The figure of 117 also excludes petitions from the archives discussed above, since the literacy of these petitioners has already been covered.

(144) See above, 70–3.

(145) Hombert and Préaux 1952: 124–5.

(146) Bagnall and Frier 1994: 25.

(147) To all of this, it could be objected that sometimes amanuenses probably neglected to insert an illiteracy formula in a petition where they ought to have done. This would mean that some of the petitioners who would appear to be literate were in fact not. However, there is no reason to believe that amanuenses would have been any more or less negligent when subscribing petitions than they were when they subscribed census declarations. Thus, even if we decrease the proportion of literate submitters of petitions and census declarations by x to represent the unknowable proportion of people whose literacy was misrepresented, petitioners still emerge as relatively more literate than those making census declarations. See Youtie 1971b: 255–6 for a few cases from other categories of document in which the fact that a party was a ‘slow writer’ or was illiterate is not properly indicated in the subscription.

(148) Youtie 1971a: 170–1; 1971b: 261.

(149) Calderini 1950: 24–7; Youtie 1971a: 165–9, 171–3; 1971b: 260; Kraus 2000: 334–8.

(150) Youtie 1971a: 173.

(151) Youtie 1971b: 174–5. See too Kraus 2000: 333–4.

(152) For discussion of the social bias in Greek papyri in general: Bagnall 2007: 188.

(153) Bagnall and Frier 1994: 47–50.

(154) The table does not reflect the fact that sometimes several people of the same occupation petition jointly. Since some of these groups are quite large, doing this would distort the statistics. The table does include cases in which we know the petitioners’ occupations from other documents; however, the petitions from the major archives discussed above have been excluded, since these have already been discussed. The cases in which an individual had multiple occupations, or in which several people with different occupations petitioned jointly have been counted once for each occupation.

(155) A point of comparison is provided by a third‐century BC tax register, which shows that at least around 50 per cent of the civilian populations of various Arsinoite villages engaged in agricultural occupations. The true figure would, however, have been much higher, since occupational designations are missing for a significant proportion of persons registered: Clarysse and Thompson 2006: 2.188–201.

(156) For a similar problem in relation to occupational designations in Ptolemaic tax registers, see Clarysse and Thompson 2006: 184, 194; cf. Crawford 1971: 123, 176–81.

(157) PSI III 249.1–6.

(158) P.Turner 41.1–6.

(159) P.Bingen 107; P.Flor. I 6; P.Giss. 34; P.Oxy. I 70; XVII 2130; XLIII 3110; XLVI 3289 (two bouleutai petition); LIX 3978; P.Tebt. II 335; PSI III 249; XIII 1328; SB XXIV 16251. All are third century AD, except for P.Tebt. II 335 (with BL III 242). See also PSI IV 303, an epistula submitted by a bouleutēs and exēgētēs of Alexandria.

(160) Chrest.Mitt. 229; P.Dubl. 18; P.Flor. I 6; P.Köln III 143; P.Oxy. II 237; XII 1418; XVII 2130; P.Oxy.Hels. 23; P.Ryl. II 116 (with BL IX 227); P.Vind.Tand. 2. Three texts date to the second century: Chrest.Mitt. 229 (AD 139); P.Oxy. II 237 (AD 186); and P.Ryl. II 116 (AD 194). The other six texts date to the third century. There is a petition dating to c. 15 BC–AD 14 with a gymnasiarch as a petitioner: BGU IV 1188. But there are now serious doubts about whether gymnasiarchs were really municipal officials rather than private ‘chairmen of village gymnasia’ at this early stage: Hagedorn 2007: 201.

(161) P.Bingen 107; P.Oxy. XLVI 3289 (two exēgētai petition); P.Turner 41; P.Vind.Tand. 2; PSI IV 281, recto, ii.27–38. All of these texts date to the third century, with the exception of PSI IV 281, recto, ii.27–38 (Oxyrhynchus; c. AD 141). They therefore escape suspicion of being state officials rather than municipal officials: cf. Hagedorn 2007: 198–201.

(162) P.Heid. IV 325; P.Oxy. I 70; PSI III 249; SB XX 14229. All are third century. In P.Oxy. IV 718.3 (AD 179–81) the petitioner is described as [ ± 12 ϲ]αντοϲ καὶ ἀρχιερ[ατεύϲαντοϲ]. Grenfell and Hunt (3n.) suggest that this could perhaps be reconstructed ἀγορανομήϲ]αντοϲ in view of the fact that such offices are usually arranged in ascending order, but the reconstruction is by no means secure.

(163) BGU XV 2459; CPR I 20; P.Flor. I 6; P.Ryl. II 116 (with BL IX 227). The latter text dates to AD 194, the others to the third century.

(164) BGU XV 2459; P.Leit. 7; P.Oxy. XII 1418; PSI XIII 1328. All are third century AD.

(165) P.Oxy. XVII 2130; P.Vind.Tand. 2. Both are third century AD.

(166) P.Oxy. IV 718 (AD 179–81); XLVI 3289 (AD 258/9).

(167) P.Oxy. XVII 2130 (AD 267).

(168) SB VI 9421 (third century AD).

(169) Hagedorn 2007.

(170) BGU III 823; 983; VII 1575; XIII 2239; Chrest.Mitt. 241; 242; P.Flor. I 55; P.Giss. 9; P.Hamb. I 10; P.Heid. IV 297; P.Louvre I 2; P.Meyer 8; P.Mich. V 230; IX 525; P.Oxy. VII 1027; XXXIV 2708; XLIX 3468; P.Ryl. II 125; 129; 151; PSI V 463; P.Stras. IV 241; P.Tebt. II 331; P.Warr. 1; SB VI 8979; XVI 12240; XXII 15782; 15831; Stud.Pal. XXII 54, BL VII 266; Stud.Pal. XXII 131.

(171) BGU I 45, BL I 11; BGU I 275; III 823; 983; Chrest.Mitt. 112; 242; Chrest.Wilck. 29; 360; 461; P.Bingen 58; P.Flor. III 382; P.Gen I2 31; P.Hamb. I 10; P.Meyer 8; P.Mich. III 175; IX 526; P.Oxy. III 486; XXXIV 2708, BL VII 152; P.Oxy. XXXVIII 2852; P.Ross.Georg. II 39; P.Ryl. II 147; 293 descr.; PSI IV 292; X 1102; P.Stras. I 5; P.Turner 34; P.Warr. 1; SB XIV 11904; 12179.

(172) P.Oxy. VII 1032 (7/16 ar.); Chrest.Wilck. 364 (½ ar.); P.Ross.Georg. II 21 (1 5/16 ar.); P.Coll.Youtie I 24 (1 ½ ar.); CPR XXIII 17 (1 ½ + ?); SB I 4415 (1 ¾ ar.); BGU XIII 2243 (2 ar.); BGU VII 1574 (3 ar.); BGU VII 1575 (1 ar.); P.Oxy. XLVI 3288 (4 ar. [1 + 3]). Note also P.Grenf. I 47, in which the petitioner claims to manage 4 ar. belonging to his nephews, who are minors.

(173) P.Oxy. III 488 (5 ¼ ar.); Chrest.Wilck. 376 (6 ar.); P.Oxy. XLIX 3464 (6 ar.); Stud.Pal. XXII 9 (9 ar.); Chrest.Mitt. 247 (16 ¼ ar.).

(174) BGU XVI 2599 (21 ar. [7 + 7 + 7]); P.Flor. I 58 (20 ar.); Chrest.Mitt. 241 (24 ar.); PSI XIII 1328 (39.92361 [13 3/16 + 16 ⅛ + 10 + 11/18]); P.Berl.Möller 2 (46 ar.); P.Oxy. IV 718 (53 ar.); P.Ryl. II 119 (83 ¼ ar.).

(175) P.Mert. I 11 (100 ar.); P.Wisc. II 86 (120.15625 ar. [8 ½ + 4 ½ + 7 + 10 + 7 + 6 ½ + 9 ⅝ + 6 +15 ½ + 11 ½ + 10 17/32 + 23 ½ ]); Chrest.Wilck. 361 (168 ½ ar.); P.Tebt. II 302 (500 ¼ ar.).

(176) For these cases, see Bowman 1985; Bagnall 1992; 1995: 75–80.

(177) See Sharp 1999: 190 for the situation in Theadelphia in AD 164/5, which is elucidated by BGU IX 1896: this document seems to evidence private arable land being owned by fifty‐seven separate people; all but three holdings fall below 10 ar., and the largest is c.43–6 ar. For private landholding in Philadelphia in AD 216, which is elucidated by P.Yale III 137, see Bowman 1985: 151 and Bagnall 1992: 130–2. There, the holdings ranged from 1 ar. to 87 ar., and were owned by 186 individuals, who had a wide variety of social stations.

(178) P.Rain.Cent. 3. For the hand and the place of the text in the archive of Satabous and family, see Schentuleit 2007: 110–11.

(179) For the role of apocalyptic literature and oracles in intellectual resistance to Roman rule, see MacMullen 1966: 128–62; Ripat 2006; Harker 2008: 120–4.

(180) Musurillo 1961: no. 22 = Musurillo 1957.

(181) Van Minnen 1994: 244.

(182) Cf. Chalon 1964: 187.

(183) P.Mert. II 83.1–10. See too Bagnall and Cribiore 2006: 317–18 for commentary.

(184) Above, 98–103.

(185) P.Mich. II 123; V 240, discussed by Hobson 1993: 198.

(186) See above, 79.

(187) See Kramer 1990: 306 for discussion and further references. SB XX 14071 provides an example of a document appointing someone to this role.

(188) See Di Segni, Patrich, and Holum 2003 for the Caesarea Maritima inscription. For the evidence in the law codes, see Jones 1964: 1.496–9, 2.591; MacMullen 1988: 156–7.

(189) Jones 1964: 1.496; Burton 1975: 101; MacMullen 1988: 134, 167; Di Segni, Patrich, and Holum 2003: 281.

(190) P.Oslo II 17.17; SB XX 14401.29–30; P.Lips. I 32.15 = Chrest.Mitt. 93; P.Wisc. I 33.7–8.

(191) For examples of the payment of a surety by an appellant, see: P.Oxy. XLVI 3296, 15 (AD 291); P.Stras. IV 227.23–5; SB XII 10929.iii.17–22; CPR V 5 = P.Rain.Cent. 68; cf. P.Oxy. XLIX 3507.32–9. P.Stras. VIII 709.14 also refers to the payment of a παραβόλιον during a bout of litigation; the fragmentary state of the document makes the precise stage of the litigation unclear, but the editors plausibly assume that here also the surety relates to an appeal. For the debate over whether SB XII 10929.iii.17–22 refers to appeals to the prefect or to the emperor, see Lewis 1995: 157–77 and literature at p. 169 n. 1. Similar sureties are attested in litigation from elsewhere in the Roman empire, again in cases involving appeals: An.Ép. (1974), 629.17–8 = IGR IV 1044; Oliver 1970: 3–9, no. 1, lines 48–52. See also Oliver 1979: 543–58.

(192) P.Flor. I 6.6, BL I 134.

(193) Wilcken 1908: 427; P.Flor. I, p. 22. The statement that Petronius is not a ‘public prosecutor’ (δημόϲιοϲ κατήγοροϲ) is reminiscent of the edict of Ti. Julius Alexander. In that text the reference is made to τω̑ν ἐν ἰδίωι λόγωι κατηγόρων (line 41), making it clear that a class of official existed in the office of the idios logos whose tasks included the bringing of prosecutions. The statement of the petition that Petronius is not a public prosecutor only really makes sense if this was the sort of case in which he could be one of these prosecutors—namely a fiscal case of relevance to the idios logos.

(194) P.Mich. XI 614.33–6.

(195) For these purposes, I have used an etic conceptualization of ‘violence’: that is, I use ‘violence’ as a modern heuristic category. In view of the terminological diversity of the petitions, it is futile to take an emic approach. Sometimes petitions clearly state that an action belongs to a conceptual category such as ὕβριϲ. But some petitions, thanks either to scribal choice or damage to the document, offer a purely factual description of the act without applying any kind of abstract conceptual label (e.g. P.Fouad 29; PSI XIV 1435). We have no way of knowing whether the writers of petitions of the latter kind would, if asked, have considered conventional categories like ὕβριϲ to be applicable to the act described, and so it would be quite wrong to exclude such texts. Thus, ‘violence’ is used in Appendix I to refer to the causing of bodily injury to a person, or threatening to do this in an immediate, plausible way. I have considered it reasonable to label as ‘violent’ forcible entries into houses whose inhabitants are present, and the robbery of property which is in the actual physical possession of a person: both of these actions generally produce an immediate apprehension of bodily harm.

(196) Note, however, P.Mich.Mchl. 24 = SB XII 11126, BL VII 226, VIII 367, which is a private letter that possibly refers to a petition being submitted by a woman who had been raped or sexually assaulted. P.Oxy. XXXVI 2758 involves a drunken man who exposed a woman to a crowd of bystanders. To our eyes, this had sexual elements, but it is not entirely clear from the petition that the act was understood in this way by those involved. SB VI 9458 involves an official who stripped the petitioner's maidservants in the street when their owner failed to pay certain dues, but this was probably a case in which the clothing was seized for its economic worth to satisfy the outstanding debt. SB XXIV 15913 is an early fourth‐century AD petition, in which the complaint is that a man seized (ἥρπαξεν) a young girl when she was out and about. Given that it is the father of her fiancé who submitted the petition, and that he mentions that she was still a virgin at the time she was seized, there is a strong inference that this was a case of sexual assault: Łukaszewicz 1997.

(197) P.Mich. VI 423, discussed by Frankfurter 2006.

(198) See, however, P.Mich. V 228; P.NYU II 3; P.Oxy.Hels 23; P.Ryl. II 128; 145.