I wish to thank two anonymous reviewers who made useful suggestions towards improving a first version of this article. It was written within the framework of the grammateus project, funded by the Swiss National Science Foundation (# 185205 and 212424). I am also grateful to my colleague Pierre Sánchez, as well as to all members of the grammateus research team for their careful reading, corrections and helpful criticism of a preliminary version: Gianluca Bonagura, Ruey-Lin Chang, Lavinia Ferretti, Susan Fogarty, Elisa Nury. I often took inspiration from their own work within the grammateus project, Bonagura et al. 2019–2026. All remaining mistakes or shortcomings are mine. Passages from Greek documents are quoted in their original spelling; a grammatical correction is provided only in cases where the spelling impedes the understanding of the text. Documents quoted here are dated by the year; month and day are provided only if they have a relevance to the argument.

Introduction

There is a well-known puzzle regarding the distribution of Greek papyri, ostraca and related material found in Egypt: in 30 BCE, the reign of Queen Cleopatra VII ends, and thus the Ptolemaic dynasty; it is replaced by the rule of Octavian, effectively turning Egypt into a Roman province in 27 BCE; and precisely in those years, available documents become very sparse, frustrating scholars of a possibility to observe in detail the shift from the Ptolemaic to the Roman period.

Skeat 1962: 100: “blacked-out landscape”; Habermann 1998: 148–150; Haensch 2008: 82; Monson 2012: 4–5: “The transitional period – the decades before and after 30 BCE – is one of the worst-documented in the whole Greco-Roman period in Egypt.”

The sparseness of sources is even more remarkable because there are several clusters concentrated on some types in definite areas, and the geographical distribution of our documentation is uneven.

Thomas 1982: 179–180.

The aim of the present study is to provide a wide survey of this black hole, as it were, focusing on the available documents from the beginning of the reign of Cleopatra (51 BCE) till the end of the reign of Octavian, a.k.a. Caesar or Augustus (14 CE).

As a matter of simplification, Augustus will be the preferred name in this survey, although he bears this title only from 27 BCE onwards. Gaius Octavius first became Gaius Iulius Caesar Octavianus through his adoption by Julius Caesar in 45 BCE (hence the name Octavian), and he is consistently named Caesar in papyri, but this title will be avoided whenever possible, to prevent a confusion with Octavian’s adoptive father.

The intended perspective is typological, considering the way documents were produced by scribes, mostly in terms of format, layout, and structure. It will be possible to show that the scribes of the reigns of Cleopatra and Augustus followed a set of basic rules when producing their documents, and that they adapted those rules according to each specific type of document.

Clusters are groups of documents belonging to the same batch, where the provenance, the period and the type are all similar. They do not necessarily constitute an archive in the narrow sense of the word, although in some cases they do, and indeed they have been classified as such in the portal Trismegistos.

The methodological aspects surrounding the concept of archive have been described in detail by Jördens 2001: 256–264.

Yet the focus here is placed mainly on the production of the documents, and less on their content, which would constitute an essential element in the study of an archive. The consistency in the mode of production is a first indication that a scribe – or a group of scribes – follows a general pattern corresponding to unstated, but recognizable rules (format, layout, structure).

Hereditary succession among scribes in the Ptolemaic administration may have contributed to this phenomenon; Ricketts 1980: 107–108.

Clusters are useful for identifying patterns that allow better comparison of types; they may allow further comparison with individual documents that do not belong to the cluster but nonetheless display similar features.

The scarcity of available papyri near the end of Cleopatra’s rule also raises the question – to this day not entirely answered – of a possible relation with the kingdom’s political and administrative downfall: did the scribes write less because the country was coming to a near standstill, or is the black hole to be explained through the chance of papyrus discovery? Livia Capponi summarizes the matter:

In the years between 39 and 31 BCE, Egypt became the battlefield for the civil war between Octavian and Antony. (…) Some scholars believe that in the 40s BCE Egypt went through an economic crisis, due to the combination of bad harvests, shortage of water and overseas debts; however, there is no secure evidence to support this, and others have argued that, on the contrary, under Cleopatra Egypt enjoyed a period of economic and cultural renaissance.

Capponi 2005: 7.

She refers to a pair of broad surveys according to which: a) papyri provide evidence for difficulties at the beginning of Cleopatra’s reign, around 50 BCE; b) similar evidence is lacking for the last decade of Cleopatra’s reign.

Thompson 1994: 323: “The early years of Cleopatra’s reign were particularly hard in the countryside as natural disaster combined with political problems.” Maehler 1983: 6–8 argues for a positive view of the situation in Egypt during the reign of Cleopatra; this is also the position held by Ricketts 1992: 275, and esp. 277: “There is nothing of the disaster in Cleopatra’s reign, however, of which Seneca writes and modern authors suggest.”

Recent research has brought new arguments supporting the hypothesis of a possible collapse of the administration in the last decade of Cleopatra’s reign.

McConnell et al. 2020.

The massive eruption of Alaska’s Okmok volcano in 43 BCE had a marked cooling effect on the environment in places as far away as Egypt and may also have influenced the Nile flood in this period. Ancient sources report famine in Egypt.

Fischer-Bovet 2023: “In the years that followed Caesar’s assassination, Cleopatra had to negotiate carefully with the different players in the Roman civil war to maintain Egypt and Cyprus’s independence while at the same time facing bad Nile floods in Egypt (43–42 BCE).” McConnell et al. 2020: 15447 cite notably (for the year 43 BCE) App. Bell. civ.4.8.61: ἡ βασιλὶς δὲ Κασσίῳ μὲν προύφερε λιμὸν ὁμοῦ καὶ λοιμὸν ἐνοχλοῦντα τότε Αἰγύπτῳ “The queen opposed to Cassius the famine as well as the pestilence affecting Egypt at the time.” They also mention an inscription from Thebes, dated around 39 BCE, OGIS 194.10 (= TM 6325): [ἐπιγιγνομένης τῆς] σκληρᾶς σιτοδείας ἐκ τῆς γενομένης ἀνιστορήτου ἀπορίας “when a harsh famine [was caused] by the unparalleled paucity”.

A causal linkage between the eruption and the famine cannot be proven but remains likely.

McConnell at al. 2020: 15447: “It is (…) challenging to establish direct links between Okmok II and the demise of the long-lasting Ptolemaic dynasty in Egypt.”

In addition to the difficult political circumstances, the eruption may have contributed to a near shutdown of the administration in Egypt in the last years of the Ptolemaic dynasty.

Data and methodology

This survey rests on the examination of some 869 testimonies recorded in the Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens (HGV) in the period from 51 BCE till 14 CE. About two thirds (548 items) consist of papyri, the remaining part (321 items) of ostraca. The latter are relevant because, in spite of the difference in writing support, they offer an interesting basis of comparison with similar texts written on papyrus. Moreover, the formulation of receipts written on ostraca illustrates the malleability of models based on types that were widely used on papyrus too.

A distinction should be made between documents that can be precisely dated, and those for which a rough estimate is provided by the editors. This is especially relevant in the period around 30 BCE. Items for which the dating span does not exceed two units (e.g. 34–33 BCE) are recorded under the label ‘precise dating’; when there are more than two units (e.g. 30 BCE – 14 CE, or ca. 16 BCE), ‘imprecise dating’ is applied.

Fig. 1: Distribution of papyri and ostraca, showing precise and imprecise dates.

Because of the existence of several important clusters, the provenance and origin of the papyri and ostraca matter a lot. By provenance, one should understand the place where the document was found, which may differ from the place of origin, where the document was produced.

Salmenkivi 2002: 28.

In the present study, the topographical scale will be that of a nome (not a city, town, or village).

Wilcken 1899: 127 chose a similar scale in his coverage of receipts written on ostraca.

Alexandria is considered at the same level as a nome, although – strictly speaking – it does not belong to Egypt. The provenance of some ostraca is vaguely given by their editors as ‘Upper Egypt’.

Fig. 2: Geographical distribution.

Entries were ordered according to their location along the Nile Valley, upstream from Alexandria to Elephantine: for the period under consideration, with the exception of the Arsinoite nome, where beside papyri a considerable number of ostraca are recorded, in the northern part of Egypt (including the Oxyrhynchite nome) the available material consists of papyri only; conversely, starting around the Coptos area southwards, ostraca predominate massively, with only a handful of papyri on record.

Syene: P.Strasb. 1 79 (16 BCE); Theban area: P.Grenf. 1 41 (2 BCE); SB 20 14948 (3 CE). This general trend was already noted by Wilcken 1899: 12. Private and official tax receipts from the Fayum (Arsinoite nome) were normally written on papyrus, whereas ostraca were in use in Thebes and Syene.

It should also be noted that the Hermopolite nome, which has yielded many papyri from the Roman and Byzantine periods, is virtually absent from the data, with a single item in the period under consideration.

P.Berl. Möller 7 (8 CE).

The distribution of sources is rendered even less balanced by the fact that, among the groups of papyri that will be discussed below, several of them were extracted from the same ensemble of mummy cartonnage found on the site of ancient Busiris, in the Herakleopolite nome (today Abusir al-Malaq, close to Beni Suef, between the Nile Valley and Medinet el-Fayum), which was excavated by Otto Rubensohn in 1903, 1904, and 1908.

Schubart 1913: 35; Preisendanz 1933: 179–180; Sarischouli 2000: 213–214, whose brief survey of the mummy-cartonnage papyri from Abusir al-Malaq remains useful, in spite of the fundamental reinterpretation of P.Bingen 45 offered by van Minnen 2000 (see below); Salmenkivi 2002: 9–27. The correspondence relating to the excavations in Abusir al-Malaq was published by Essler 2025: 165–168.

Every document on record in the period from 51 BCE till 14 CE was given a preliminary examination, with a focus on the provenance and origin, the date, and the text. All images available online were checked, and in most cases, when a print illustration was at hand, it was also examined.

In the course of preparing this study, the British Library’s website was hacked, thus blocking access to all their digital images. Fortunately, a first survey of all relevant images had been made before the hack. Service is being gradually restored but in many cases further verification proved impossible.

Regarding geographical place of origin, the level of the nome seemed to provide an appropriate criterion; likewise, regarding typology, documents were grouped not according to specific labels, but to general types: contract, declaration, letter, list, official proceedings, order, petition, and receipt, and other (i.e. five decrees on papyrus, six dates on ostraca).

Fig. 3: Distribution by types.

If we combine the information about geographical distribution and types of documents, it appears that ostraca in the period under consideration are for the most part receipts (tax, bank payment, delivery etc.) from Upper Egypt (starting around Coptos). They are nonetheless important for the overall picture because they provide separate evidence on the black hole of 30 BCE, to which we must now turn.

General trends

One may first consider the general appearance of the data, considering only documents with a precise dating, papyri and ostraca, from all areas of Egypt, including all types of texts, year by year. In the graph presented below, the vertical axis corresponds to the years of the reigns of Cleopatra and Augustus, starting from the bottom (51 BCE) and moving upwards (till 14 CE). On the horizontal axis, each dot or short line corresponds to the quantity of documents attested for a given year and is positioned to indicate the cumulative amount of material, year after year.

Overall, the scarcer the documents for a given year or period, the steeper the slope of the curve. In the top right portion of the graph, a regular progression of testimonies appears from ca. 19 BCE till 14 CE. This corresponds to what one would expect in an ideal situation where the production of documents at the time was even, and where a large quantity of material compensates for irregularities produced by the chance discovery of papyri and ostraca in various locations in Egypt. The steep curve in the bottom left portion of the graph, however, with a total absence of testimonies in the years 40, 39, 38, 36, 33, and 31 BCE, suggests an anomaly in the distribution of the material in the years preceding the end of the Ptolemaic dynasty.

As mentioned above, two major hypotheses may explain this phenomenon:

The presence of clusters, both in terms of location and of types of documents, may have influenced the data. The availability of papyri and ostraca is determined to a large extent by chance discovery. The scarcity of papyri and ostraca in the last years of the Ptolemaic dynasty reflects a widespread downfall in administrative activity, caused by a collapse of central power.

Fig. 4: Chronological distribution, by year.

The graph shown above combines both ostraca and papyri. Considering the two writing supports separately, the curves are similar, with scarcer testimonies from the late 40s till the early 20s, when the slope becomes regular again. Since ostraca come mainly from Upper Egypt, and papyri mainly from Lower Egypt, the consistency in the data speaks against hypothesis a), according to which the chance of discovery may have distorted the state of our documentation. [[[FIG. 5 and 6 Chronological distribution, separating papyri and ostraca.]]]

Fig. 5 and 6: Chronological distribution, separating papyri and ostraca.

Clusters do not account for all available documents. It may therefore be assumed that scribal activity was impeded in the last decade of Cleopatra’s reign, perhaps due to both economic and political circumstances (see above), and that Augustus somehow restored some order in the new province of Egypt, so that scribal activity became once more regular.

Clusters

Letters from the Athenodoros Archive

Many letters from the Herakleopolite nome are preserved in Berlin. They belong to the large ensemble of papyri extracted from mummy cartonnage found in Abusir al-Malaq (see above). This specific batch is known as the Athenodoros Archive, named after an estate manager.

TM Archive 26; Bagnall and Cribiore 2006: 123–125.

There are 74 documents on record, mostly letters (also one uncertain and one related item), dating from 25 BCE till 1 CE.

In this archive, the scribes who produced the letters followed unstated rules that give a regular consistency to the lot. The format and layout, as well as some variations, make it possible to establish these basic rules. With very few exceptions – some of which will be addressed below – the scribes make use of the so-called pagina format, i.e. a vertical sheet of papyrus cut from a roll, the height of the sheet being about twice its width. The height of the sheet varies from ca. 17 cm to ca. 33 cm, depending on the original height of the roll.

BGU 16 2627: 17.5 cm; BGU 16 2610: 20.2 cm; BGU 16 2643: 33 cm.

The scribe has little control over this parameter, unless he chooses to cut part of the height of the sheet.

This may be the case of BGU 16 2609, a squarish sheet with a height of 17.7 cm, possibly from a taller roll cut in half.

He has more freedom to adapt the width of the sheet to the purpose of the document and to the amount of text he plans to write. The fibres are almost invariably horizontal.

Vertical fibres: BGU 16 2605. No kollesis is visible on either side; therefore, it is not possible to determine which is the front or the back.

The text to be written is then placed so as to leave ample margins at the top and the bottom, and a narrower margin on the left. On the right, the scribe aligns his text as much as possible (more on this below). When the sheet is detached from the roll, some stray strokes may stick out of the alignment. Therefore, a narrow margin prevents those strokes from protruding too much into what will become the next sheet on the roll.

Causo 2025. A right margin appears on BGU 16 2617.

This layout may be applied by a well-trained scribe with an elegant hand, as in the case displayed below (BGU 16 2651), or by a scribe with a lower level of competence (e.g. BGU 16 2627).

Fig. 7: BGU 16 2651, image and structural display (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 25183. https://berlpap.smb.museum/05095/).

In the letter displayed here, as in other conspicuous cases, the line-spacing is quite generous; the scribe took special care to center vertically his block of text.

See also 16 2608; 2609; 2610; 2611; 2613; 2637; 2640; in the so-called Tomos of Asklepiades (TM Archive 111, see below), BGU 4 1204 (28 BCE, Bousiris [Herakleopolite nome]). Examples of extra-ample line spacing are to be found in BGU 16 2620, 2638, and 2644, in a layout similar to that of BGU 16 2651.

The line-spacing is in itself unsurprising: in the third century BCE the scribes in the service of Apollonios, Ptolemy II Philadelphos’ finance minister, already produced a neat layout by inserting a generous space between the lines of their letters, as in the well-known case of PSI 5 514 (251 BCE, Philadelphia, found in Philadelphia, written in Alexandria). In this earlier instance, however, the difference lies in the use of a horizontal format and in the narrower margins.

The basic pattern found in BGU 16 2651 appears in other documents of the archive, although the re-use of the papyrus in mummy cartonnage may have caused the loss of a part of the sheet.

E.g. BGU 16 2613.

This is not limited to letters from the Athenodoros archive: on the contrary, in the description of other clusters below, as well as in some other isolated cases, it appears that there is a consistent pattern.

E.g. P.Oslo 2 47 (1 CE, provenance unknown), a neat letter that was prepared following the same rules for layout, with the text centered between generous top and bottom margins, a reduced left margin, virtually no right margin, and ample line spacing. Another example is to be found in P.Tebt. 2 409 (5 CE, Tebtynis [Arsinoite nome]).

Centering the text vertically requires some guesswork on the part of the scribe and it is more easily done with a short text, as in the example illustrated above. In one specific case, presumably because of the extreme shortness of the planned text, the scribe cut a small sheet (H 16 cm × W 8.8 cm) where he placed four brief lines, using a similar layout, but in a miniature sizing.

BGU 16 2638. The same approach was presumably followed in the case of the letter BGU 16 2653 (9 BCE, Herakleopolite nome; H 13.3 cm × W 16.6 cm), and of PSI 16 1638 (73/44/22 BCE, Tebytnis; H 11 cm × W 7.4 cm), a receipt for the payment of rent written in letter form.

If the text turns out to be shorter than the scribe has anticipated, he is left with a more generous bottom margin, thus spoiling the impression of symmetry.

BGU 16 2606; 2608; 2654.

Conversely, when the text turns out to be longer, he may try to accommodate the layout in advance, by reducing either the line spacing, or the width of the left margin.

Line spacing: BGU 16 2629. Left margin: BGU 16 2600; 2607. Outside the Athenodoros Archive, BGU 8 1760 (50 BCE, Herakleopolite nome), a longish letter from a strategos who is forwarding an order from the dioiketes.

He may also encroach on the bottom margin, sometimes squeezing the last lines into the bottom part of the sheet.

BGU 16 2618; 2640; 2650. In BGU 4 1206 (28 BCE, Herakleopolite nome), a private letter, the scribe started with an intended layout that included a sizeable top margin and generous line spacing. The text turned out to be longer than anticipated, thus the scribe reduced the line spacing, made his writing more compact (less space between the characters), and ended up reaching the bottom edge of the sheet, leaving no lower margin. Outside the batch of papyri extracted from mummy cartonnage found in Abusir al-Malaq, SB 14 11294 (2 BCE, Arsinoite nome [?]).

In one extreme case, for lack of space at the bottom, a scribe placed the closing salutation and date in the top margin.

BGU 16 2604.

The ample lower margin may also be used for the addition of a note to the letter.

BGU 16 2611; 2630.

The relative homogeneity of the set of letters found in the Athenodoros Archive raises the question of whether the rules applied by these scribes were specific to a narrow local environment and to the writing of letters stricto sensu, or if they were also in use among scribes following the letter form for various purposes in other parts of Egypt. In the Herakleopolite nome, but outside the frame of the Athenodoros Archive, a document in letter form (no ἔρρωσο at the end) establishing the foundation of a temple (BGU 4 1202, 18 BCE, Bousiris [Herakleopolite nome]) was written in a fashion very similar to the letters from the Athenodoros Archive. Outside the Herakleopolite nome, the scribe of P.Oxy. 78 5166 (ca. 29–20 BCE, Oxyrhynchos) applied to a letter a layout very similar to that of BGU 16 2651 (above Fig. 7): there is an ample top margin, a somewhat narrower left margin, and virtually no margin on the right; the bottom margin is lost but could have been of the same size as the top margin.

The dimensions of the preserved part of the papyrus are H 14.5 × W 10.4 cm; top margin 3 cm, left margin 2 cm; if the lower margin was also 3 cm, the complete sheet would have reached a height of 17.5 cm.

The scribe also made use of a generous line spacing, as is often the case in the Athenodoros Archive. A comparable case is to be found in P.Oxy. 78 5164 (25 BCE, Oxyrhynchos), a receipt for the delivery of oil written in letter form (no ἔρρωσο at the end): although the text is complete, the bottom margin is lost too, but the general appearance of the document is highly reminiscent of BGU 16 2651.

The dimensions of the preserved part of the papyrus are H 15 × W 7.5 cm; top margin 4 cm, left margin 2 cm; if the lower margin was also 4 cm, the complete sheet would have reached a height of 19 cm. The line spacing is also generous.

The same pattern appears in the Arsinoite nome with P.Ryl. 2 73 (32–31 BCE, Euhemeria), a receipt in letter form. This can be contrasted with P.Wash. Univ. 2 106 (18 BCE, Oxyrhynchos), a private letter written by a woman who is clearly not a professional scribe. To quote the editors of this document, this letter is “laboriously drawn in somewhat uneven letters, and rather large (…). The writer’s orthography is impressionistic (…) and her syntax casual, at best.” She nonetheless followed some – if not all – of the conventions that prevailed for letter-writing: margins at the top and bottom, as well as on the left, no right margin, use of the standard initial and final formulae, placement of the greeting ἔρρωσο and of a date. The parallel between P.Oxy. 78 5164 and P.Wash. Univ. 2 106 underlines the obvious, i.e. that a well-trained scribe is apt to follow unstated rules for producing a document such as a letter more closely than an occasional writer. Another notable case of a scribe who does not wholly master the rules for layout is to be found in SB 14 12172 (7 CE, provenance unknown), where the writing style and the spelling betray a clumsy writer. A narrow left margin, as well as a wider but irregular right margin, are also quite atypical when compared to the letters that have been described above.

Another question is whether scribes wrote their letters while the sheet of papyrus was still part of the roll, then cut it off once it was completed, or if they first cut their sheet off the roll before starting to write. BGU 16 2646 (3 BCE, Herakleopolite nome), which displays an unusual format and layout, provides a clue to the matter. The sheet has a horizontal shape (H 24.7 cm × W 30 cm) and carries two columns of writing. The first column holds a letter, complete with opening, main text, salutation and date. Some characters on the right side of the column run across a kollesis in the middle of the sheet; nothing suggests that the two columns were originally two separate sheets pasted together. The second column contains some additional remarks, like an afterthought to the letter itself.

The editor, William Brashear, calls it a “gigantic post scriptum”.

This implies that, when the scribe wrote the letter (col. 1), he still had some free writing space on the right, although he could not foresee that he would need it, or else he would have adapted his format accordingly and would have placed the salutation and date at the end of col. 2. Therefore, the letter was written while the sheet was still part of the roll and the scribe had some free writing space on his right.

The case of BGU 16 2604, mentioned above, where the scribe placed the closing salutation and date in the top margin, suggests the opposite: if he had to cram the ending clauses at the top, this must have been because his sheet was already detached from the roll and he could not start a second column anymore. Therefore, we cannot be certain that there was a standard approach to the matter, and this may have been determined by the level of competence of the scribe. Skilled scribes, when starting to write their letters, would choose the width of their left margin (starting from the left edge of their papyrus roll), then estimate the space that their text would occupy and adapt accordingly the width of the column, as well as the dimensions of the top and bottom margins. With short letters, a scribe could easily balance those parameters; with longer ones, he sometimes ran out of space at the bottom. In the standard practice, his letter should hold a single column of text, which would be cut off the roll when it was completed. In the case of BGU 16 2646, however, an afterthought brought the scribe to keep writing on, producing a second column, before the wide sheet of papyrus was cut off the roll.

This sequence of operations is confirmed by BGU 4 1205 (28 BCE, Bousiris [Herakleopolite nome]), where – at least – two private letters were written on the same roll, the second letter in two columns. The scribe had to complete the third column and place the final greeting ἔρρωσο before cutting the sheet off from the roll. In a later period, P.Sarap. 85-89 (90–133 CE, Alexandria [?]) is a series of five letters written by Heliodoros to his father Sarapion, his mother Selene, and to his brothers Eutychides, Anoubion, and Phibion. Heliodoros started a new column for each letter but did not cut them off the roll. Thus, the multiple letters reached his family in the form of a short roll.

To the basic layout, scribes often add some specific elements. The first letter of the first line may be slightly enlarged, but this is not done systematically.

BGU 16 2600; 2610; 2637; 2640; 2651; 2653.

In one case, the scribe wrote the opening, which contains the names of the sender and of the addressee, in a more formal hand than the rest of the letter.

BGU 16 2640 1–2.

He may also add a paragraphos below the text of the letter.

BGU 16 2600.19; 2613.9; 2611.14; 2637.12; 2640.9. Paragraphoi also appear in other types of documents (see below), and they are not specific to the Herakleopolite nome or to Alexandria in this period: see e.g. P.Oxy. 78 5169.21 (18 BCE, Oxyrhynchos, repayment of a loan); SB 12 10942.36 (4 BCE, Oxyrhynchite nome, land lease contract).

The salutation formula (ἔρρωσο) and/or the date are frequently set off to the right in relation to the left margin.

BGU 16 2600.19; 2606.14–15; 2609.6; 2613.10; 2611.15; 2637.13; 2640.10; 2651.11; 2654.10.

This is a remnant of the old way of writing letters on horizontal sheets transversa charta, which still appear occasionally in the period considered here.

E.g. BGU 20 2846 (50–49 BCE, Herakleopolite nome). BGU 8 1769 (ca. 47 BCE, Herakleopolite nome), a letter, displays the old transversa charta format in use in the earlier Ptolemaic period, but here the format is presumably a fossil, as it were. This letter pertains to the nomination of an official in charge of matters related to the katoikoi. For such documents, this format was maintained until the early third century CE; Ferretti et al. 2020.

An atypical case, BGU 16 2652 (ca. 10–2 BCE), deserves some attention. The papyrus sheet displays a horizontal format (H 9.4 cm × W 16.1 cm), with horizontal fibres. A vertical kollesis on the right part of the sheet shows that it was not placed transversa charta. This relatively unusual format may be explained by the content: it is a payment order, a type of document that starts like a letter – Ἀθηνόδωρος Εὐρυλόχῳ τῷ φιλτάτῳ χαίρ(ειν) – but carries no final salutation. A parallel is to be found in P.Oxy. 78 5165 (24 BCE, Oxyrhynchos), a bank payment order that also starts like a letter but does not carry final greetings.

Bagnall and Bogaert 1975: 99–100.

The sheet is square (H 9.8 × W 9.8 cm) and the writing runs along the fibres.

To come back to BGU 16 2652, it stands in fact among the early cases of a type that is still clearly recognizable in the fifth century CE.

See grammateus, description of ‘Order to pay’, Bonagura et al. 2023. In the fifth century CE, e.g. P.Wisc. 2 63 (410 CE, Oxyrhynchos).

A close parallel, both in form and content, is also to be found in P.Gen. 12 2 (II/III CE, Arsinoite nome [?]). It thus seems that the scribe of BGU 16 2652 followed the general pattern used for letters (horizontal fibres, ample top and left margins, generous line spacing), but that he adapted it to a payment order (horizontal shape, no salutation) and encroached on the bottom margin, even placing some additional information between lines 3 and 4. This case illustrates a certain degree of malleability in the basic rules followed by the scribes of the Athenodoros archive, by which they could adapt a general type (letter) to a specific content (payment order). We shall observe a similar trend below in several instances, notably the so-called Cleopatra ordinance, as well as among receipts written on ostraca. Before we turn to them, however, we must address some other issues, the first being the distinction between professional and occasional scribes.

Professional and occasional scribes: the Tomos of Asklepiades

The parallel between P.Oxy. 78 5164 and P.Wash. Univ. 2 106, mentioned in the preceding section, served to underline the fact that scribes do not all show the same degree of competence while writing a document, and therefore are not all skilled in the same way when it comes to applying the basic rules that have been identified so far. This distinction is confirmed by a small archive that also belongs to the papyri extracted from cartonnage found in Abusir al-Malaq, the so-called Tomos of Asklepiades.

TM Archive 111; Bagnall and Cribiore 2006: 114–122.

It consists of a group of eleven private letters (three of them unpublished) covering the years 29 to 23 BCE.

An update on the location of the documents belonging to this small archive is provided by Monte 2018: 125–126.

Once delivered, a registration docket was added in the top margin of some of the letters, before each was pasted into a roll, i.e. a tomos synkollesimos, an unusual phenomenon in the case of private letters. Normally, this is done for official documents to be archived by an authority. This oddity, together with the registration dockets, suggests that the family of Asklepiades could afford the help of professional secretaries.

Four hands are at work, the first of which belongs to a well-trained scribe who wrote BGU 4 1203, 1204, 1207 (and inv. 13152c) on behalf of Isidora, Asklepiades’ sister. This scribe follows the pattern described above, with the text centered vertically, leaving sizeable top and bottom margins, a narrower left margin, a minimal right margin, as well as generous space between the lines. In the case of BGU 4 1207, the top margin was used for a registration docket, and in the bottom margin another scribe, less skilled, added a post scriptum on behalf of Isidora.

Isidora herself seems to have written some of the letters, with a hand that betrays a lower level of competence (BGU 4 1205, 1206, and P.Berl. Monte 7).

Her level of proficiency may be compared to that of the scribe who wrote P.Würzb. 2 39 (16 CE, Oxyrhynchite nome).

She does not master cursive writing (each character is detached from the previous one), and the shape of the letters is irregular. When she started writing BGU 4 1206, she tried to follow the basic rules for layout, leaving a generous margin at the top, a narrower one on the left, and a minimal right margin. The interlinear spacing was also quite generous. She miscalculated, however, the space she would need on the sheet.

Bagnall and Cribiore 2006: 119 state that “the writer visibly grows tired as the letter proceeds so that legibility is difficult toward the end.”

Thus, in the lower part of the document, the writing becomes more compact in letter-spacing, in line-spacing, and in letter-size, and Isidora ends up cramming the final greetings at the bottom of the sheet, leaving no margin. In the case of BGU 4 1205, Isidora opted for a letter in two columns, but the rather irregular writing betrays a hand much less skilled than that of a professional scribe. Yet the most recently published addition to the archive (P.Berl. Monte 7) casts a more favourable light on the skills of Isidora as a writer. Presumably because the text was shorter (six lines including the date), she could apply more consistently the rules that were described above. The sheet of papyrus (H 24.9 cm × W 14.7 cm) compares well with BGU 16 2651 (H 23 cm × W 15.3 cm), displayed above in the section pertaining to the Athenodoros Archive. Isidora centres her block of text, allowing for sizeable top (7.5 cm) and bottom (9.3 cm) margins, leaving a left margin of 3 cm (and virtually no right margin). The line spacing is also quite generous. In other words, when the conditions are right, she masters the basic rules; but when her text gets too long, she is at pains to stick to the same standard.

Asklepiades receives letters written by a third hand, one from his mother Tryphaina (BGU 16 2665) and the other from his brother Tryphon (BGU 4 1209). In accordance with his competence in producing a standard layout, this scribe has a very cursive hand and can be ranked in an intermediate position between the first scribe and Isidora: in BGU 4 1209, he manages the top and left margins well, keeps a generous spacing between the lines, but lacks space at the bottom; although BGU 16 2665 is heavily damaged, it seems to display a similar layout.

Finally, a fourth hand is at work (BGU 4 1208). It belongs to a scribe whose writing is both cursive and very smooth. The letter he writes on behalf of Asklepiades’ brother Tryphon covers three columns, the last of which is narrower than the two first columns. There is a generous top and bottom margin. This scribe has a good command of space management and he calibrates the width of the last column, reducing it, so that in the end the three columns have almost the same height.

The Tomos of Asklepiades thus illustrates the fact that applying the basic rules towards preparing documents in the period under consideration depends in part on the skill of the scribe: the more competent ones master the system in a consistent way, whereas the writing style and the layout betray occasional writers who struggle to follow the rules, with more limited success than the professional scribes.

Petitions: Officials of the Herakleopolite nome

Mummy cartonnage from the Herakleopolite nome has yielded another large set of documents, placed under the general label “Archive of the officials of the Herakleopolites”.

TM Archive 156.

It consists of several parts, of which Group III contains petitions dating from ca. 62 till 47 BCE (34 documents in the reign of Cleopatra, between 51 and 47 BCE).

The bulk of those petitions is to be found in BGU 8 1813-1870; add BGU 20 2845 and 2847; SB 6 9065.

They display a high level of consistency in format, layout, and structure, which suggests that, here too, the scribes were following a set of standard rules, either explicit or implicit. Another petition (BGU 16 2599), belonging also to the same cartonnage ensemble, but dating from after 27–26 BCE, preserves an identical layout.

The sheet of papyrus was cut from a tall roll (ca. 30 cm). The shape corresponded to the pagina format, with horizontal fibres. The name and title of the addressee was regularly shifted slightly to the left (ekthesis), which made this element immediately visible. The text of the petition itself (not including the final greeting) often ended with a paragraphos on the left.

Ferretti 2024: 130–141.

The rules that apply to letter writing are not the same as those for petition writing. In the case of a petition where the plaintiffs’ scribe chose to use the letter form, with its characteristic heading, the format and layout correspond to those of the letters found in the Athenodoros Archive, and not to the petitions found in the Archive of the officials of the Herakleopolites.

BGU 16 2602 (ca. 14–13 BCE, Techtho [Herakleopolite nome]); Mascellari 2021: 48–49, 746, and 1007.

Particular features distinguish the layout of those petitions from that of letters. In the case of letters (see above), the top and bottom margins were preferably of the same size, creating an effect of symmetry. A reason for that may have been that a letter, conceived as a reciprocal means of communication, was normally to be answered by another letter; there was no need to leave a window below for a reply (apostil). In the case of petitions, however, communication was unidirectional: the petitioner did not receive a full document in return, but the official in charge often added an apostil below the main text of the petition.

Ferretti et al. 2023a; Ferretti 2024: 190–191.

The scribes therefore had to provide the space for the apostil, which may explain the asymmetrical dimensions of the top and bottom margin, the latter being often more generous. In several instances, a red stamp was applied to the document, either at the top or – more seldom – at the bottom of the sheet.

Ferretti 2024: 165–166.

This was presumably meant to confirm registration of the petition in the official archives.

Several of the key typological elements that appear in those petitions are already on display in an earlier petition that does not belong to this archive, P.Dion. 11 (= P.Rein. 1.18 = M.Chr. 26 = Sel. Pap. 2.277; ca. 108 BCE, Hermopolite nome). This suggests that the scribes active in the Herakleopolite archive did not innovate: on the contrary, they were following a well-established practice that was already in use at a distance of ca. 150 km as the crow flies, more than half a century earlier.

A similar observation can be made among tax receipts on ostraca, where specific elements that occur in the Augustan period are also to be found at an earlier date. Wilcken 1899: 88–89: “Ich will hier ein für alle Mal einschieben, dass, was uns als neu unter Augustus entgegentritt, vielleicht schon im I. Jahrhundert v. Chr. unter den Ptolemäern Brauch gewesen ist.”

The table below offers a basic comparison between P.Dion. 11 and four well-preserved Herakleopolite petitions.

Table 1 Comparison of typological features of various petitions.

P.Dion. 11

108 BCE

BGU 8 1832

51 BCE

BGU 8 1833

51–50 BCE

BGU 4 1187

ca. 49–48 BCE

BGU 8 1848

ca. 47 BCE

BGU 16 2599

later than 27–26 BCE

pagina, horizontal fibres

H 29.5 cm

W 13.0 cm

H 31.8 cm

W 14.0 cm

H 31.5 cm

W 12.5 cm

H 30.0 cm

W 13.0 cm

H 31.0 cm

W 10.0 cm

H 30 cm

W 12 cm

addressee in ekthesis line 1 lines 1-2 lines 1-2 lines 1-2 lines 1-2 line 1 petition ends with paragraphos line 36 line 18 line 27 line 35 line 29 ? space left at bottom for apostil lines 38–42 line 20 line 29 no apostil no apostil no apostil red stamp upper left corner, round-shaped no stamp upper left corner, losange-shaped top center, losange-shaped no stamp

Fig. 8: P.Dion. 11 (Inv. Sorb. 2027, image Sorbonne Université – Institut de Papyrologie); BGU 8 1833 (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 13806 https://berlpap.smb.museum/03859/); image and structural display..

The general layout of BGU 8 1833 (51–50 BCE, Herakleopolite nome) is also quite similar to that of P.Würzb. 1 5 (31 BCE, Tholthis [Oxyrhynchite nome]), a petition to the strategos; and it appears again in P.Oxy. 49 3468 (first cent. CE, Oxyrhynchos), a petition to the prefect. Therefore, it seems that the same rules prevailed not only through time (from the second century BCE till the first century CE), but also in space (Hermopolite, Herakleopolite, Oxyrhynchite nomes).

On the chronological dimension, Ferretti 2024: 196–200.

Some additional features found in Herakleopolite petitions, however, do not appear in P.Dion. 11.

Ferretti 2024: 150–151 and 160–166.

In the top left corner of some petitions, a cross (in the shape of a St. Andrew’s cross) was added.

BGU 8 1779; 1780; 1834; BGU 20 2847.

This is presumably also a registration mark; it never occurs on documents that bear a red stamp and therefore may have been an alternative means of registration. Incoming petitions were also sometimes registered with a short note placed in the top margin, consisting of a date and abbreviations, the meaning of which is not always clear.

See BGU 20 2845 1–2 note. Other occurrences from 51 BCE onwards: BGU 8 1779; 1834; 1840; 1870; BGU 20 2847.

Similarly to letters described above, scribes could calibrate line spacing according to the length of the intended text. Petitions being often lengthier than letters, a medium spacing was used in most cases. There are a few instances, however, where the scribe adopted a more generous line spacing.

BGU 8 1830; 1832; 1833; 1836; 1839.

BGU 8 1832, an almost fully preserved document, is quite telling on this matter: the text of the petition was relatively short and the scribe therefore increased the line spacing so as to leave a reasonable space below for the apostil, while at the same time keeping the general proportions of margins that were expected for such a document.

Declarations of sheep and goats

The mummy cartonnage from Abusir al-Malaq has yielded a batch of declarations submitted by owners of sheep and goats to the assistants of a supervisor for the pasture-tax in the Herakleopolite nome in 13 BCE.

BGU 16 2578-2587. BGU 16 2586 is an exception: it is addressed to the assistant of the strategos in 5 BCE. The typology of declarations of sheep and goats is covered by Ferretti 2024: 263–287, to whom I owe much information and insight on this matter. Beside Brashear’s useful introduction in BGU 16 (p. 51), Kruse 2001: 213–235 and Langellotti 2012 on the contents of such declarations.

This considerably increased the available material, especially for the early Roman period. Those documents, being declarations, belong to the general type of hypomnema (‘memo, note’), characterized by a heading that starts with the addressee (in the dative), followed by the declarant (with παρά + genitive). In this lot, four items are sufficiently well preserved to allow a comparison of format and layout (table below). Roll height varies from 21.7 cm to 29.5 cm, but the top and bottom margins of the two shorter sheets (BGU 16 2581 and 2586) may not be fully preserved; width is almost identical in all four declarations. The rules followed by the scribes to produce those declarations correspond to the system that was identified above for other types of documents: the scribe aligns his text on the right (allowing for a cut close to the text once he detaches the sheet from the roll), leaves ample space above and below, together with a narrower margin on the left, and he makes the name of the addressees conspicuous by setting the first lines off to the left (ekthesis). He also manages the line spacing according to the amount of information he must fit into the intended space, in order to keep a sizeable bottom margin. The top and bottom margins may be used for secondary purposes, as when another hand adds a registration docket at the top or some additional information in the bottom margin, but this is not done on a consistent basis. A paragraphos closes the declaration in two cases out of four.

Table 2 Typological comparison of declarations of sheep and goats.

BGU 16 H W Wide top and bottom margins, narrow left margin Top margin used for registration docket Bottom margin used for additional notes Declarant in ekthesis εὐτύχει Line spacing

Para-

graphos

2581 21.7 11.7 X generous X 2582 29.5 12 X X X X medium 2583 27.2 12 X X X X medium 2586 21.3 12.5 X X X X X generous X

Fig. 9: BGU 16 2582, image and structural display (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 25293. https://berlpap.smb.museum/05202/).

It thus seems that, in spite of some minor variations, the scribes who produced those declarations followed the same general approach as for other types of documents described above.

Double documents, the management of margins, and innovation

The general pattern used for the layout of letters may be applied to yet other types of documents, and in some cases the generous top and bottom margins serve a specific purpose. P.Ryl. 4 580 (78/49/27 BCE [?], provenance unknown) preserves the left half of a document by which a soldier assigns, in the event of his death, his burial benefit to an unspecified person. Before half of the papyrus was lost, the sheet had a squarish shape (ca. 21.4 × 17 cm) and the text block of the assignment was positioned at mid-height, leaving a sizeable top (8.7 cm) and bottom (7.7 cm) margin.

Plate 2 in P.Ryl. 4.

The left margin was narrower (1.9 cm). Based on the possible restoration of line 8 proposed by the editors, which seems likely, the right margin was very narrow.

P.Ryl. 4 580.8 n. (= Roberts and Turner 1952: 33). The editors suggested in their commentary ὃ ἐστὶν ἀργυρίου (δραχμαὶ) ἑ[κατὸν ἀποδοθησόμενον τῶι τὸ] | σύμβολον ἐπιφέροντ[ι (8–9; last iota not mentioned by the editors, but implicit) “which makes one hundred silver drachmas, to be paid to the person producing this token”. Assuming that the sheet was folded in half and that the break corresponds to the fold, this restoration would neatly fit the width of the lost half of the sheet, leaving virtually no margin on the right.

Below this central block of text, the scribe placed a curved paragraphos protruding on the left.

In the bottom margin, the president and secretary of the soldiers’ association then placed their signatures. In the top margin, a summary of the assignment was added, which turned this into a so-called double document: the top three lines were folded and sealed, thus becoming the scriptura interior.

On double documents and the evolution of the proportion between scriptura interior and scriptura exterior, Ferretti et al. 2023b, grammateus, “Description of Greek Documentary Papyri: Double Document”.

Contracts in the form of a double document have evolved through time, with a shift in the last quarter of the second century BCE: whereas in the early Ptolemaic period, the scriptura interior took roughly the same space as the scriptura exterior, somewhere between 130 and 113 BCE a marked asymmetry between the two appears, with the scriptura interior occupying at most three lines.

On the shift from symmetrical to asymmetrical double documents, Yiftach-Firanko 2008. Later examples are P.Dion. 16 (109 BCE, Akoris [Hermopolite nome]; PSI 10 1098 (51 BCE, Tebtynis [Arsinoite nome]); SB 5 7532 (74 BCE, Nilopolis [Arsinoite nome]), which was revised by Youtie 1973, with a detailed description of the layout of the text.

In another double document, P.Merton 1 6 (77 BCE, Nestou Epoikion [Arsinoite nome]), the editors note that “this fact [that the borrower’s name is wrongly given], with the appearance of the scriptura interior, which is in a different hand from the rest of the document, suggests that at this period the scriptura exterior alone was formally written by the principal scribe, and that not till after this had been completed, with the necessary subscriptions, was the scriptura interior added, by a subordinate and often, as here and in P.Würzb. 6, in the most negligent manner.” A similar phenomenon can be observed in the case of SB 5 7532, and also of P.Bon. 10 (45 BCE, provenance unknown). This confirms a pattern by which the main text (scriptura exterior) of contracts produced as double documents was written first, following the same rules as for the letters described above, after which a) the acknowledgement was inserted in the bottom margin, and b) another scribe added the summary (scriptura interior) in the top margin before it was sealed. The general approach to the layout is similar to that of letters, but the purpose is different.

The double documents mentioned above seem to confirm the pattern described so far; but there is also a specific case where, on the contrary, double documents bear testimony to a striking – and short-lived – innovation on the part of the scribes. Village record-offices (γραφεῖα) in the Arsinoite nome produced in a period from ca. 74 BCE till 10 CE a score of contracts that display a peculiar appearance.

Described by Husselman, Boak, and Edgerton 1944: 1–11 (= introduction to P.Mich. 5); also Claytor 2014, with a list of documents on pages 113–115. Four additional notarial contracts from the grapheion of Theadelphia, belonging to the Harthotes Archive, were published by Claytor, Litinas, and Nabney 2016.

They were produced on tall sheets (approximately H 30 cm × W 12 cm), and the scriptura interior, which took ca. 3–4 lines, was placed at the top of the sheet, leaving a sizeable empty space before the scriptura exterior was added in the lower half of the sheet.

Fig. 10 P.Fay. 240, image and structural display (Image: Cairo Museum, Association Internationale de Papyrologues, Center for the Study of Ancient Documents [Oxford], Adam Bülow-Jacobsen.).

The format and layout are very consistent, but the scribes – who worked in at least seven different grapheia – did not follow the pattern which we have found so far: there was hardly any margin at the top and bottom, or on the left and right of the sheet. In some cases, the scribes also resorted to a system of double dating, by the κράτησις of Augustus and by the ordinary regnal year.

Grenfell, Hunt, and Hogarth 1900: 223 (= P.Fay. 89, introduction).

This is an innovation that met with little success and did not last beyond the reign of Augustus. In fact, the whole layout of those contracts is considered as an experimentation that was discontinued after Augustus.

Claytor 2014: 95–98.

Thus, this peculiar type of document is interesting both for its idiosyncrasy (it departs markedly from the established pattern), and for its consistency (this experimentation was conducted in several villages of the Arsinoite nome, presumably with the help of models that were circulated under the authority of the strategos and the royal scribe).

Archive of the Alexandrian scribal office: synchoresis contracts

The paradox regarding papyrological data in the reigns of Cleopatra and Augustus is rendered even more acute in the case of Alexandria: on the one hand, we have only few documents originating from that city; on the other, most of what is available dates from the reign of Augustus, when the overall amount of papyrological testimonies is at its lowest level. For the most part, the Alexandrian material consists of a batch of contracts grouped under the label synchoresis ‘agreement’, because a form of the verb συγχωρέω appears in the text of the contract. This archive, labelled “Archive of the Alexandrian scribal office”, covers a period of twenty years, from ca. 23 till 5 BCE, with a peak around 14/13 BCE; it contains about a hundred documents.

On the dating range, van Minnen 2022: 1006. The papyri were published in BGU vol. 4. See also Schubart 1913: 37 and 47; Wolff 1978: 91–95; Claytor and van Minnen 2021. For an extensive description of the characteristics of synchoreseis, van Minnen 2023: 85–91.

The papyri were retrieved from mummy cartonnage found in Abusir al-Malaq (see above). The attribution to a scribal office in Alexandria is justified by the content of the documents.

Schubart 1913: 45.

Several contracts were drafts, with many corrections; on some sheets, a scribe wrote a first version of contracts unrelated to one another, and in some cases both sides of the sheet were used. After the papyri had served their original purpose, they were presumably sold upstream as scrap paper and recycled to produce cardboard covering for mummies.

Schubart 1913: 42; Preisendanz 1933: 179–180, who notes that the sarcophagi were made in the first century CE, using old documents from the first century BCE.

A detailed study of the content and legal aspects of Alexandrian synchoreseis would be out of place within the present context; instead, we shall focus on the typological aspects of this dossier. The scribes who produced the documents were professionals who – here again – followed some basic rules with remarkable consistency, although there are exceptions to the characteristics listed below. Several documents were produced by the same scribe following a regular pattern.

E.g. BGU 4 1055 (= M.Chr. 104); 4 1058 (= M.Chr. 170 = CPGr 1.4); 4 1102; 4 1103; 4 1115.

Before describing this pattern, however, it is necessary to go back in time to a prior instance of synchoresis, which will provide the institutional context of such documents.

Formally speaking, a synchoresis was shaped as a petition submitted to a court, requesting a judge to confirm the deal. This procedure can be traced back to the mid-second century through P.Merton 2 59 (154 or 143 BCE, Crocodilopolis [Arsinoite nome]), an agreement to terminate a marriage contract, where the court of chrematistai confirms the deal.

On the juristic relation of P.Merton 2 59 to Alexandrian synchoreseis, Wolff 1978: 93. On the typological aspects, van Minnen 2023: 96–97; Ferretti 2024: 181–183 and 300–303.

It is the only synchoresis that is not related to Alexandria in one way or another.

van Minnen 2023: 85: “All other synchoreseis are addressed to the archidikastes in Alexandria and are either exemplars issued to one of the parties and taken by them to Middle Egypt or copies made from them.”

Prior to the publication of P.Merton 2 59, Ulrich Wilcken had already described the mode of preparation of such agreements:

Wilcken 1927: 544 (= UPZ 1 118, introduction).

the scribe wrote the central part of the document, leaving an ample top and bottom margin. In the case of P.Merton 2 59, the text of the agreement was then checked by the eisagogeus (assistant to the chrematistai), before the grammateus (secretary to the chrematistai) signed it for the parties below (note in particular, in lines 29–32, the repeated use of συγχωρῶ / συγχώρησις), and summarized the legal context in the top space. Finally, once the chrematistai had approved the deal, the grammateus added the court decision below, on behalf of the eisagogeus. What matters for our understanding of the production of this agreement is that the scribe deliberately left large blank spaces at the top and bottom of the sheet for the various officials to complete the document. There was an administrative routine in the procedure which is reflected in the layout. Therefore, although the format, layout and structure of P.Merton 2 59 differ from the Alexandrian synchoreseis, the basic method followed by the scribes is quite similar, with a focus on the management of space on the sheet of papyrus, as well as on a display that made the various steps of the procedure evident at first glance.

Fig. 11: P.Merton 2 59 (from the printed edition, pl. VIII).

We may now come back to the production of Alexandrian synchoreseis. The format of the sheet was in most cases vertical, although horizontal sheets are also attested at a later date, in so-called landscape format with several columns.

van Minnen 2023: 92–93, who cites – among others – P.Fam. Tebt. 20 (120–121 CE, Ptolemais Euergetis) and P.Tebt. 2 319 (ca. 248 CE, Tebtynis).

With a height of ca. 37 cm, the rolls used by this office in the first century BCE were comparatively tall. The sheet, once it had been detached from the roll, had a width of about a third of the height, ranging from 9 to 16 cm, on average 13 cm.

A synchoresis started with a heading along the model of a hypomnema with the addressee (dative), followed by the contracting parties (παρά + genitive).

For a more detailed survey of this Alexandrian archive and its relation to the general type of hypomnema, Ferretti 2024: 292–304.

The addressee was highlighted by an insertion in ekthesis (shifted to the left) at the top, sometimes with a space before the main body of the text. Because of the length of the texts, scribes opted for a small script and a medium to dense interlinear spacing. They could not center vertically the main block of text on the sheet (as was possible for short letters described above). A top and left margin of moderate size – and virtually no right margin – provided the main outline of the text to be written, and the scribe kept writing until he had included all the necessary content. In other words, the size of the bottom margin varied according to the space required for the text.

Fig. 12: BGU 4 1118, image and structural display (Image: Cairo Museum, Association Internationale de Papyrologues, Center for the Study of Ancient Documents [Oxford], Adam Bülow-Jacobsen.).

The last line of the main body of text was regularly marked by a paragraphos, in the shape of a curved stroke protruding into the left margin. This feature appears also in letters and petitions from the Herakleopolite nome (see above). The date, placed at the bottom, is detached from the main body of text and shifted to the right. The layout of an Alexandrian synchoresis may also be compared with contracts drafted as double documents (discussed above), with sizeable top and bottom margins, a narrower left margin, virtually no margin on the right, and the main text ending with a paragraphos. The top margin is used for the scriptura interior, and in the bottom margin the contracting party confirms the deal.

E.g. PSI 10 1098 (51 BCE, Tebtynis [Arsinoite nome]).

Those were the main rules followed by the scribes who produced synchoresis contracts. Other features were added on an irregular basis. When the final version of the contract was produced, one copy could be archived and pasted to a roll (τόμος συγκολλήσιμος), with a classification number. Therefore, in some of the drafts the scribe anticipated this classification by adding, in the top margin, κολ (for κόλλημα).

Schubart 1913: 46.

The bottom margin, with its frequently generous dimensions, served another purpose: in BGU 4 1107, a contract that was nearly completed, two women wrote their consent (συνχωρῶ) between the main body of text and the date, displaying a mastery of writing far below that of the professional scribe who had prepared the contract.

A similar phenomenon occurs in another type, namely a declaration on oath (addressed below, BGU 2 543).

Thus, the asymmetrical margins, with a bottom margin taller than the top margin, allowed some space for the last stage of the procedure, when the contracting parties gave their approval to the agreement.

The Cleopatra ordinance and other documents from individuals of high standing

P.Bingen 45 (received Feb. 23, 33 BCE, Alexandria [?]), known to many papyrologists as the ‘Cleopatra papyrus’, provides an interesting point of comparison with synchoresis contracts from Alexandria, precisely because of the potential for confusion, and also because the rules followed by the scribe in writing this document, although not identical, nonetheless bear some resemblance to synchoresis contracts on some important aspects. After examining the main typological features of this document, it will be worth comparing it with a few others, in order to better understand the implicit guidelines followed by the scribes in the service of individuals of high standing in Egypt in the reign of Cleopatra and also in the early Augustan period.

This papyrus, which belongs to the Berlin collection, came from the same mummy cartonnage ensemble found in Abusir al-Malaq that also provided several clusters described above.

van Minnen 2003: 40.

Yet there is no direct relation to any of those archives. The presence of the verb σ̣υνκεχωρήκαμ̣[εν] (3) led to its erroneous identification as a synchoresis contract in the editio princeps. It was soon recognized that this document is of a different nature: it is an ordinance granting fiscal privileges to a Roman of high standing, and it was convincingly argued that such an ordinance could originate only from the highest level of authority in the kingdom, namely Cleopatra VII herself.

van Minnen 2000.

In the ample bottom margin of the document, a single-word approval (γινέσθωι “let it be so”) was added, possibly in the Queen’s own hand, although this is far from certain.

Pro: van Minnen 2000: 29–34; contra: Sarri 2018: 168, n. 593. The superfluous iota at the end of the imperative γινέσθωι is a common mistake that should not be taken as a sign of sloppiness on the part of a ruler who was famous for her linguistic skills (Plut. Ant. 27.4–5; Legras 2021: 35–40; Wasserstein 2014 is more skeptical, but his position does not rest on any convincing argument). In the main body of the text, Cleopatra’s own scribe made a consistent use of iotaadscript but also inserted a superfluous iota at the end of an imperative (15: γραφήτωι). On this phonetic oddity, Clarysse 1976. In SB 5 7337 (= C.Ord. Ptol. 75), a prostagma preserved by an inscription, one finds iota adscript for the dative singular (3: στρατηγῶι), mistaken iota adscript at the end of an imperative 3rd sing. (5: μεταγραφήτωι), and correctly spelled imperative without iota adscript (6: ἐκτεθήτω). The confused spelling with superfluous iota (Gignac 1976: 185) appears at least till the mid-second century CE, e.g. O.Did. 448.6: ἔρρωσωι; SB 16 12334.8: ἔστωι. Beyond the second century, the only attested case is unreliable, P.Gascou 67.10 (368 CE): [ἔστ]ωι.

In the present context, we shall leave aside the – admittedly important – question of the name of the beneficiary of this fiscal privilege.

Whether he was the better-known Publius Canidius Crassus (van Minnen 2000), or a more obscure Quintus Cascel(l)ius (Zimmermann 2002), matters for the historical interpretation of the papyrus but has little bearing on the way the document was prepared by the scribe. A summary is provided by Bagnall and Derow 2004: 109–110.

This document corresponds to the first step in the dissemination of a royal ordinance (πρόσταγμα).

van Minnen 2001: 77; van Minnen 2003: 38–39.

A scribe wrote down the text dictated by the Queen – presumably with the help of her staff –, who then confirmed it in writing, either adding γινέσθωι in her own hand, or ordering the scribe to write it himself. The document was passed on to an official at the highest level, whose identity remains unknown (2: [ ̣ ̣ ̣ ̣]ωι); he would himself write to local officials, appending a copy of the ordinance. A close parallel to this procedure is offered by the text of an inscription where the transmission letter was displayed in Ptolemais, followed by the text of an ordinance of Cleopatra dating from 46 BCE.

I.Asylia 226, the text of which is reproduced – with translation – by van Minnen 2003: 43. This source was published too recently to be included in C.Ord. Ptol.

A similar chain of transmission is also briefly outlined in a letter from a strategos to the superintendent of the harbour of Herakleopolis in 146 BCE: (…) κα̣θ̣ότ̣ι̣ γέγραφεν Δημήτριος τ̣[ῶ]ν πρώτων φίλων καὶ γραμματεὺς δυνάμεων προστεταχέναι τὸν βασιλέα, παρασταθ̣ή̣τωι πλοῖον παραχρῆμα “(…) in accordance with the letter of Demetrios, (of the rank) of first friends, secretary of the (armed) forces, (stating that) the king issued an ordinance, let a boat be prepared immediately.”

P.Gen. 3 131.6–10, also not included in C.Ord. Ptol. The wording παρασταθήτωι πλοῖον παραχρῆμα is presumably a direct quotation from the King’s ordinance. Note the iota ending the imperative παρασταθήτωι, which recalls Cleopatra’s γινέσθωι.

Within the chain of transmission, Demetrios, who as quartermaster-general was in charge of the logistics of Ptolemy VI’s army, held a position comparable to that of the addressee of P.Bingen 45: he was in direct contact with the King and forwarded in writing the ordinance to a nome strategos, who himself sent his orders accordingly through a letter.

Coming back to P.Bingen 45, the scribe who prepared the document worked in a fashion that displays similarities with documents found in the Archive of the Alexandrian scribal office, and especially among the letters in the Athenodoros Archive. There is nonetheless a difference in the format of the sheet, which is squarish and was cut from a roll with a height of ca. 24 cm. The scribe allowed for a sizeable top margin; the left margin was presumably narrower, but it is possible that the papyrus in its present state does not preserve its whole width. The name of the addressee was apparently placed in ekthesis, on the same pattern as Alexandrian synchoresis contracts. The scribe opted for a medium-sized line spacing. The block of text is not centered vertically: when the task was completed, there remained an ample margin below.

Fig. 13: P.Bingen 45, image and structural display (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 25239. https://berlpap.smb.museum/05150/).

The top and bottom margins served a purpose in the next stages of completion of the document. The Queen gave her formal approval in the bottom margin, placing γινέσθωι exactly where, in letters from the Ptolemaic period, the final greeting ἔρρωσο was normally written.

E.g. P.Cair. Zen. 3 59305.6 (250 BCE, provenance unknown); P.Cair. Zen. 2 59285.5 (250 BCE, Alexandria); P.Col. 4 122 (181 BCE, Arsinoite nome [?]); BGU 4 1204.9 (28 BCE, Herakleopolite nome).

Finally, a receipt docket, with a date, was placed in the top margin, following a practice that is also attested in petitions from the Herakleopolite nome (see above), as well as in letters from the same nome.

Petitions: e.g. BGU 8 1831.1 (prior to 13.01.50 BCE); BGU 8 1834.2 (50 BCE); BGU 20 2847.1 (49 BCE). Letters: e.g. BGU 8 1873.1 (mid-first century BCE); BGU 4 1207.1 (28 BCE).

This document may now be compared with PSI 8 969, an official letter dated July 6, 51 BCE.

On the date and the identification of the sender, Peremans and van’t Dack 1957: 122.

The provenance is recorded as unknown but the content of the document, together with some prosopographical data, provides clues both to its origin and to the place where it was found. Achilleus writes to Seleukos, introducing Archedemos who is labelled as τῶν ἰδίων, i.e. a person closely attached to Achilleus. With the assistance of Diokles, Archedemos is to oversee all matters in an unspecified nome (3–4: πρὸς τῆι προστασίαι τῶν κατὰ [τὸ]ν ν[ο]μὸν πάντων). This must be the Herakleopolite nome, where a strategos named Seleukos is in office precisely at that time.

Schubart and Schäfer 1933: 86 (= BGU VIII). My interpretation differs somewhat from Peremans and van’t Dack 1957: 126, who believe that Diokles was already in office at the time when Archidemos, a friend or relative of Achilleus, was designated for the same position. I agree with Peremans, van’t Dack 1957: 126 in reading ν[ο]μόν ‘nome’, and not ν[ό]μον ‘law’. The words προστασία and προίσταμαι may apply to a variety of objects: an estate (BGU 2 365.1–3; P.Ryl. 2 132, 138, 145 and 148; P.NYU 2 3.3–4), an endowment (SB 20 15150.8–9), revenues (P.Tebt. 1 5.58; BGU 16 2601.6–7), a shrine (P.Tebt. 3.1 790.5), a chapel of sacred ibis (P.Fouad 1 16.4–5), policemen (SB 14 12089.1–2), sheep (BGU 16 2586.4–5; P.Tebt. 1 53.8).

Seleukos apparently remains in office since he is attested as strategos both before and after July 6.

Seleukos is recorded – among other attestations – as strategos on June 20, 51 BCE (BGU 8 1832.1–2) and on February 12, 50 BCE (BGU 8 1761.12 and 17). On October 24, 50 BCE, he has been replaced in office by a strategos named Soteles (SB 5 7611.1–2 and 8–9), who is himself succeeded by Eurylochos on August 9, 47 BCE (BGU 8 1811.1). On the possible overlap between the strategoi Seleukos and Paniskos, Peremans and van’t Dack 1957: 122–123 reach the conclusion that Paniskos was in office from year 21 of Auletes (61–60 BCE) till year 26 (56 BCE, taking into account the reign of Berenike IV).

For a reason that is not provided in PSI 8 969, a higher authority temporarily imposed on the Herakleopolite nome a direct form of control through two envoys.

Here my interpretation diverges from that of Peremans and van’t Dack 1957: 127, who believe that Archedemos and Diokles were appointed, on a regular basis, to oversee a specific aspect of the administration in the nome. Thus, Achilleus is identified as the nome’s eklogistes in Prosp. Ptol. 1.146 and 4.1070. Their whole interpretation does not fit in well with Seleukos’ wording τῶν κατὰ [τὸ]ν ν[ο]μὸν πάντων.

The head of this delegation, Archedemos, is under the close orders of Achilleus, who must be a high official in Alexandria.

Peremans and van’t Dack 1957: 123 already raise this possibility.

Therefore, it seems justified to consider that PSI 8 969 was written in Alexandria, in an office comparable to the one which issued Cleopatra’s ordinance.

Fig. 14: PSI 8 969, image and structural display (Firenze, Biblioteca Medicea Laurenziana. With kind permission by the Ministry of Culture. Any further reproduction, by any means, is prohibited).

Whereas in PSI 8 969 Achilleus, writing a letter to the strategos Seleukos, started with Ἀ̣χ̣ιλλε̣ὺ[ς] Σελεύκωι χαίρειν καὶ ἐρρῶσθαι, in P.Bingen 45 Queen Cleopatra would address her high official only by his name ([ ̣ ̣ ̣ ̣]ωι), precisely because this was not a letter, but an order (πρόσταγμα) from the sovereign. This is the usual form of address in interdepartmental correspondence, to borrow from Peter van Minnen’s wording.

van Minnen 2023: 96.

Likewise, the Queen would not sign her order with final greetings, which would imply a kind of reciprocity: instead, she would simply express her will that her order be accomplished (γινέσθωι), and she wrote this single word – or had it written – in place of the usual ἔρρωσο. Thus, in spite of some differences, the basic rules were quite similar, but the scribes adapted them to the purpose of the document they were preparing, as well as to the position of the person issuing the text.

We shall come back to the strategos Seleukos further down, from another perspective (section on πίστις). At this point, we have already encountered several instances where scribes rely on a basic set of rules, but adapt them to their specific purpose: such was the case with some letters belonging to the Athenodoros Archive, and a comparison between P.Bingen 45 (the Cleopatra ordinance) and PSI 8 969 (Achilleus’ letter to Seleukos) also illustrates the case. This malleability, however, is not limited to documents from the reigns of Cleopatra and Augustus: on a broader scale, it can be observed in the production of tax receipts from the Roman period written on ostraca (more on ostraca below). Wilcken’s careful listing of various formularies for tax receipts in letter form highlights a similar malleability and he notes an increasing tendency to drop the salutation χαίρειν.

Instead of malleability, he writes of a formulary gone wild. Wilcken 1899: 85 “(…) sodass schliesslich der ursprüngliche Briefstil ganz verwildert ist.”

This happens exceptionally in the Ptolemaic period, and he points out only two cases under the rule of Augustus, but in the second century CE the phenomenon becomes more frequent overall, and it occurs also in the Byzantine period. There is a similar tendency to drop the final ἔρρωσο.

Wilcken 1899: 84–85.

Besides a possible desire on the part of the scribes to simplify the formulation, omitting χαίρειν was also understood as a way for the sender to mark his higher social standing. Wilcken notes that in Elephantine and Syene, in the Roman period, the only receipt in letter form where χαίρειν is missing is addressed to a slave.

Wilcken 1899: 119, referring to O.Wilck. 235 (158 CE, Elephantine).

He also recalls an anecdote from Plutarch’s Life of Phokion 17.9–10 that stresses a similar point, from the opposite perspective:

ἰδίᾳ δὲ τὸν Φωκίωνα ποιησάμενος αὑτοῦ φίλον καὶ ξένον, εἰς τοσαύτην ἔθετο τιμὴν ὅσην εἶχον ὀλίγοι τῶν ἀεὶ συνόντων. ὁ γοῦν Δοῦρις εἴρηκεν, ὡς μέγας γενόμενος καὶ Δαρείου κρατήσας ἀφεῖλε τῶν ἐπιστολῶν τὸ χαίρειν, πλὴν ἐν ὅσαις ἔγραφε Φωκίωνι· τοῦτον δὲ μόνον ὥσπερ Ἀντίπατρον μετὰ τοῦ χαίρειν προσηγόρευε.

On a private level, [Alexander the Great] entertained with Phokion a relation of friendship and hospitality, and he honoured him in a way that only a few of his close companions enjoyed. This is confirmed by Douris, who says that, after Alexander had acquired great power and had defeated Darius, he gave up writing chairein in his letters, except for all those he wrote to Phokion. To him only – as to Antipater – he sent greetings with chairein.

Thus, a sovereign may drop χαίρειν while addressing a subject, but conversely a subject may not dispense with χαίρειν when addressing the sovereign. In a possible exception to this rule, SB 6 9065.1–2 (50–49 BCE, Herakleopolite nome [?]), the opening βα̣σιλίσσῃ Κλεοπάτραι καὶ̣ [ ̣ ̣ ̣ ̣ ̣ ̣ ̣ ̣] | Ἡρακλεία Ἀπολ̣λ̣ω̣νίου̣ “To Queen Kelopatra and (…, from) Herakleia daughter of Apollonios” should be corrected to βα̣σιλίσσῃ Κλεοπάτραι χα̣ί[ρειν] | Ἡρακλεία Ἀπολ̣λ̣ω̣νίου̣ “To Queen Kleopatra, greetings, (from) Herakleia daughter of Apollonios”.

Schubert 2025.

It should also be noted that the scribe who prepared this enteuxis followed the general pattern that prevailed for the writing of Cleopatra’s ordinance (P.Bingen 45) and Achilleus’ letter to the strategos Seleukos (PSI 8 969): minimal right margin, sizeable top and left margin; the bottom margin seems less generous, but the bottom of the sheet is damaged. Because the scribe could presumably anticipate the length of the text, he opted for a large format (H 32 cm × W 25 cm) and a fairly dense interlinear spacing. Cleopatra’s title and name are placed in ekthesis to the left, to be compared with the official’s name in Cleopatra’s ordinance and with the name of the addressee at the top of synchoresis contracts – as well as of all hypomnemata in the first century BCE.

Information kindly provided by Lavinia Ferretti.

The final εὐτυχεῖτε is placed where, in a letter, one would expect ἔρρωσο (and where Cleopatra, in her ordinance, placed γινέσθωι).

To come back to P.Bingen 45, it is a unique testimony in the sense that we do not have any exact parallel, in its original form, to the writing down of a prostagma under the direct supervision of the sovereign. This brief typological analysis shows that the scribe – presumably well-trained, being under the direct orders of Cleopatra – adapted some standard scribal methods to the specific purpose of the document he was producing. His handling of the margins is paralleled in letters, petitions, as well as in synchoresis contracts from Alexandria, the opening is an adaptation of a letter heading, Cleopatra’s confirmation is placed where the greetings normally appear in a letter, and the registration docket placed at the top follows the practice of petitions and letters too. The general rules followed in all those types of documents prevail also in the writing of an enteuxis.

A few parallels to the Cleopatra ordinance will take us to the Augustan period, where the rules followed by the scribes in the highest strata of power seem to be identical. BGU 16 2558 (12 BCE, Herakleopolite nome) preserves the text of an edict issued by an individual named Proculus who, according to the revised edition of this papyrus, must have been in a very high position in the hierarchy of the Roman administration, either in Egypt or at the level of the Empire.

Hagedorn and Jördens 2006: 172–174. The right to issue an edict (line 1: λέγει) is limited to the highest levels of command, but Hagedorn and Jördens believe that in this case the Prefect of Egypt must be ruled out. They also note (169) that, contrary to the claim made by the editor of BGU 16 2558, this may not be the earliest attested edict from Roman Egypt, but it certainly ranks among the earliest occurrences.

He was recently identified as C. Calpurnius Proculus and was most probably the Prefect of Egypt at this time.

Capponi 2016: 1716–1718, with a reinterpretation of SB 24 16132 (= ChLA 47.1434; Oxyrhynchos, 13 BCE).

The edict, which is related to illegal requisitions, forbids the inappropriate wearing of a paenula (with the probable exception of soldiers). The text is shorter than that of the Cleopatra ordinance, which explains both the more generous line spacing of BGU 16 2558 and the reduced size of the sheet.

H 15 cm × W 14.7 cm, in contrast with P.Bingen 45: H 23.3 cm × W 20 cm.

The proportions, however, as well as the general layout of the document, are strikingly similar; a comparative table will be provided below. Both texts end with a curved paragraphos protruding from the left of the last line of text. Where Cleopatra’s scribe had placed γινέσθωι, Proculus’ scribe wrote a date; an edict required neither a sovereign’s approval, nor a closing salutation. From a typological perspective, the scribe of BGU 16 2558 produced under the reign of Augustus a miniature version of P.Bingen 45, as it were, borrowing the same codes that prevailed towards the end of the Ptolemaic era. Given the content of both documents, it seems likely that they were produced in a similar scribal environment.

The same applies to BGU 16 2623, a letter of recommendation dated 10 BCE that belongs to the Athenodoros Archive (discussed above). In its revised version, this letter offers a striking example of correspondence at the highest level of society in Egypt.

Partial revision by Hagedorn 2006: 165.

The sender, Phaidros, writes on behalf of a man presented as Λεύκιος ἀνὴρ ἀγαθός “Lucius, a gentleman”, presumably a Roman who intends to sail upriver to the Herakleopolite nome in order to collect the income from his property.

3–4: ἀναπλέων εἰς τὸν Ἡρακλεοπολ(ίτην) ἐπὶ εἴσπραξιν ἰδίων κερματίων. The formula ἀνὴρ ἀγαθός and the corresponding verb ἀνδραγαθέω originally refer to a set of values that define a Greek of noble behaviour, and within the context of Graeco-Roman Egypt it may apply to a man respected by his peers, e.g. P.Mich. 5 245.3–6 (= SB 5 8030; 47 CE, Tebtynis): ἔδοξαν {σαν} ἑαυτῦς κυνῇ γνώμῃ προχιρίσαι τινὰ ἐξ αὑτῶν ἄνδρα ἀγαθόν (…) Ἀπύνχιν Ὀρσεῦτος “they have decided by common consent to elect one of their number, a good man, Apynchis, son of Orseus” [transl. Husselman, Boak, and Edgerton]. Quite often, however, it should be understood simply as “to behave properly, to do things the right way, [hence] to make an effort”. BGU 4 1207.10–12 (28 BCE, Herakleopolite nome, from the Tomos of Asklepiades, discussed above): σὺ οὖν κ̣α̣ὶ̣ [Ἁραμώτης] ἀνδραγαθεῖτε καὶ εἰσάγεσθε τ̣[ιμὴν φ]ακο̣ῦ̣ ὀλυρίω “Therefore, would you and Haramothes please do the right thing, namely forward the value of the lentil (mixed with?) barley.” In the same archive, BGU 4 1204.6; 1205.13–14; 1206.13. Also P.Oxy. 42 3069.13–15 (III/IV CE, Oxyrhynchos): ἀνδραγάθε̣[ι] οὖν καὶ τὰ λοιπὰ ἐπιτέλεσον ὡς ἀνὴ[ρ ἀ]γ̣α̣θός “Therefore, behave in an upright manner and carry out the rest like a gentleman.”

Phaidros flatters Athenodoros by describing him as the person of the highest standing he could presently find to look after Lucius in the Herakleopolite nome.

6–7: οὐχ ὁρῶν οὖν ἐν τῶι παρό̣ντι ἄ̣[λλο]ν̣ [ἀ]νότερόν σου, with ἀνώτερον to be understood as “higher in social standing”, not “residing higher upriver”; Hagedorn 2006: 165.

It seems likely that Phaidros is writing from Alexandria, where Lucius normally resides. The opening salutation (1–2) is unusually elaborate: χαίρειν κα[ὶ] ἐρρωμένωι ἄριστʼ ἐπανάγειν ὡς βούλομαι. In anticipation of the comparison with another document discussed below, it should also be noted that Phaidros politely reminds Athenodoros of their acquaintance, in case Athenodoros had forgotten about it (7–8): εἰ ἄρα ἔτι ἔστιν τίς σοι μνήμη [ἡ]μῶν.

The rare wording ὡς βούλομαι deserves particular attention because, one century later, the use of βούλομαι instead of εὔχομαι will become a marker used by officials of the highest standing (for the most part, the Prefect of Egypt) who underline the fact that they do not merely ‘wish’, they ‘want’.

Mascellari 2024: 93.

It was recently shown that, contrary to a common assumption, the use of βούλομαι by the highest officials in Egypt does not occur throughout the Roman period: a list of attestations indicates that the phenomenon appears in the second century CE, whereas in the first century high officials still follow the common practice of the Ptolemaic era, ending their letters with ἔρρωσο.

Haensch and Kreuzsaler 2020: 205–215; Mascellari 2024: 82–83.

It may thus be inferred that the wording used by Phaidros (ὡς βούλομαι) does not mark him as the Prefect of Egypt or as another high official, but rather that high officials took over a specific expression that may have been in current use in the Augustan period among individuals of very high standing in Egypt.

The identity of the Prefect of Egypt in 10 BCE remains as yet unknown.

Unsurprisingly, the proportions relating to format (size of sheet, margins) and the layout (placing of basic elements) of BGU 16 2623 are similar to that of P.Bingen 45 and BGU 16 2558, and reflect the rules followed by the scribes employed in the highest offices in the country.

Table 3 Dimensions of P.Bingen 45, BGU 16 2558, and BGU 16 2623.

height width Proportion height / width margins top bottom left right P.Bingen 45 23.3 20.2 1.15 4.8 7.4 4.5 ca. 0 BGU 16 2558 15 14.7 1.02 2.2 2.5 1.8 0.5 – 3.4 BGU 16 2623 19.6 16 1.22 2 5.5 3 ca. 0

The wording ὡς βούλομαι deserves one last mention. The only other occurrence in the period under consideration is to be found in a letter belonging also to the Athenodoros Archive, BGU 16 2642 (ca. 21–5 BCE, Herakleopolite nome). In this case, however, it is a letter sent by Athenodoros, with the heading (2–3) πλεῖστα χαίρειν καὶ διὰ παντὸς ἐρρωμένῳ ἄριστα ἐπανάγειν ὡς βούλομ(αι), which closely mirrors Phaidros’ opening salutation in BGU 16 2623. Athenodoros apologizes to his addressee – whose name is lost – for not having written more often because he was burdened with some tasks. The text is badly mutilated but is seems likely that this is Athenodoros’ reply to Phaidros, in which he would be mirroring a wording that earmarks individuals of very high standing (notably ὡς βούλομαι). The main perceptible difference between the two documents lies in the format, layout, and quality of writing: Athenodoros’ letter was written by a fast and irregular hand, and the document displays no structured layout as in the other documents discussed above. There is hardly any doubt that this is a draft used by Athenodoros’ scribe – presumably dictated to him. It would explain why a letter by Athenodoros was preserved in his own papers (the final version was sent off to the addressee). This is also a useful element in the reconstruction of the process followed by those scribes of high standing: they first wrote down a fast, preliminary version of the document that was dictated to them by the official; only then did they make a clean copy for sending. In the case of P.Bingen 45, this clean copy was the one which Cleopatra confirmed with γινέσθωι.

Finally, a pair of documents, SB 16 12713 and 12714 (ca. 5 CE, Philadelphia), that preserve a petition to the Prefect of Egypt, raises the possibility that there may have been not two, but three stages of preparation: a first, fast writing down of the intended text (if such were the case, we do not have it anymore); then a first attempt at producing the petition, in a careful hand and already with the layout typical of scribes of high standing (SB 16 12713), which was heavily corrected; and eventually a clean copy, with a similar layout and no more corrections to the text (SB 16 12714).

A parallel to this sequence of several versions of a same petition is to be found in UPZ 1 18-20 (163 BCE, Memphis) and in P.Lond. 2 354 (p. 163) + CPR 15 15 (7–4 BCE, Soknopaiou Nesos); Jördens 2017: 272–273. In the latter case, Jördens shows that the difference between draft and clean version does not pertain only to the layout or spelling, but that the argumentative strategy is developed from one version to the next.

The ordinance prepared in P.Bingen 45 was then ready to be circulated, as we have seen above. In a similar procedure, BGU 8 1730 (= SB 4.7419 = Sel. Pap. II 209 = C.Ord.Ptol. 73; 79 or 50 BCE, Herakleopolite nome) is presumably a copy, received and registered by a local secretary (τοπογραμματεύς), of an original sent by the chancery in Alexandria to the strategos of the Herakleopolite nome. The first editors indicate that, in the top margin, they can barely read συγγενεῖ καὶ στρατηγῶι Ἡλιοδώρωι in ekthesis. They also identify the traces of a squarish red stamp in the middle of the sheet, applied for registration purposes. Below the text of the ordinance, the τοπογραμματεύς added a note, stating that he had made a public posting of a copy.

Temporary grant of protection against prosecution (πίστις)

The typological observations we can make at the level of scribes working together with the Queen and her close officials also apply to a lower level. We saw above that Seleukos, strategos of the Herakleopolite nome, received a letter from the higher authorities, ordering him to collaborate with Achilleus’ envoys (PSI 8 969). In BGU 8 1810 (51/48 BCE, Herakleopolite nome), the same strategos issued a πίστις, i.e. a temporary grant of protection against prosecution on behalf of an individual under his authority, and his scribe worked in a fashion similar to what has been described previously. Before we can examine BGU 8 1810, a short word of explanation on the concept of pistis is necessary.

A pistis could be issued for the benefit of an individual who, in order to avoid arrest, had sought asylum, for instance in a sanctuary.

Schäfer 1933; Lenger 1980: 208–209 (introduction to C.Ord. Ptol. 74).

A temporary refuge from prosecution offered immediate protection, but it could indirectly cause discomfort to the individual, and also some damage to public interest, for instance if the person was prevented from fulfilling an obligation such as taking part in the harvest. There was therefore a mutual interest in suspending prosecution until the individual had accomplished his duties, and a pistis worked as a safe-conduct. The juridical concept was supported by a royal ordinance; the nome strategos was in charge of issuing the document on behalf of the ruler and of implementing the measure.

Royal ordinance: BGU 8 1812.3–5 (= C.Ord. Ptol. 74; 49–48 BCE) εἶν̣α̣ι̣ δ̣ὲ̣ ἀ̣ν̣επάφους καὶ τοὺς παρ̣[αλαβόντας τὰς] παρʼ ἡμῶν πίστεις μέχρι ἂν ἀπὸ τῆ̣[ς γεωργίας] τῶν ἐδαφῶν γένωνται “(and we have ordered) that those who have received the pledges from us be protected from seizure until they are through with cultivating the plots”. Strategos issuing the pistis: P.Tebt. 1 41.11–12 (105–90 BCE) ὧν ἔχομεν παρὰ Λυσανίου τοῦ συγγενοῦς καὶ στρατηγοῦ πίστεω̣ν “the pledges which we have (received) from Lysanias, of the rank of relatives (to the King), and strategos”.

From the farmers’ perspective, some were molested without justification.

P.Tebt. 3.1 785.22–25 (ca. 138 BC, Tebtynis): διʼ ὃν πεπόηται παραλογισμὸν συμβέβηκεν περισπᾶσθαί με ἀλόγως ἀπὸ τῆς βασιλικῆς γῆς. “Through her false statement it came about that I was unreasonably disturbed from the crown land.” (transl. Hunt and Smyly)

Before a pistis became a distinct type of document, the broader concept of temporary protection was already in informal use since the late third century BCE, if not earlier.

SB 18 13861.13–15 (= P.Mich. 10.601; 210 BCE, Philadelphia [?]): σὺ οὖν καλῶς ποείσεις φροντίσας ὡς οὐ περισπασθησόμεθα. “Therefore, would you please see to it that we remain unmolested.” P.Tebt. 3.1 741.1–5 (187–186 BCE, Tebtynis): τῆς πρὸς Ἀνί[κητον] τὸν παρʼ ἡμῶν ἐπιστολῆς ὑπόκειταί σοι τὸ ἀν[τίγραφον]. καλῶς οὖν ποήσεις καὶ σὺ συντάξας μὴ περ̣[ισπᾶν] τοὺς ̣ ἀ̣ν̣θ̣ρ̣ώ̣π̣ους μέχρι τοῦ ἐπιβα̣[λό]ν̣τ̣[α]ς ̣ ἡ̣μ̣[ᾶς ποιήσασθαι] τὴν ἁρμόζουσαν ἐ̣[πι]στροφὴν [ὑ]πὲρ ὧν δη[λοῦσιν]. “Below is a copy of a letter to my agent Aniketos. Accordingly, you will do well to issue orders yourself that the persons are not to be molested before I arrive and give their statements proper attention.” (transl. Hunt and Smyly)

The pistis as a definite type of document came into existence between the late second and early first century BCE. It bears some loose relation to the apostils found in petitions, which eventually were detached from the petitions and became separate documents, i.e. warrants.

Schubert 2018. A close relationship between pistis and warrant may be established through the example of P.Lond. 2 379 (p. 162; Herakleia [Ars. Nome]), a pistis by content with the format and general phrasing of a warrant. It is undoubtedly an official’s response to a petition. This papyrus was dated to the third century CE by Kenyon, but the first century CE seems more likely; Schubert 2018: 263 and 268–269.

The difference, however, lies in the fact that the apostils to petitions are not addressed to the person receiving protection, but to an official in charge of granting protection.

P.Dion. 11.38–43 (ca. 108 BCE, Hermopolite nome; discussed above in relation with petitions): Βίαντι· ε[ἰ ἔ]σ̣τ̣ι̣ βα̣σιλικὸς γεωργός, [π]ρ̣ονοηθῆναι ὡς ἀπερίσπα̣[στ]ο̣ς ̣ κατασταθήσεται μέχρι ἂν [ἀπὸ] τ̣οῦ σπόρου γένηται. (ἔτους) ι Θῶυθ κδ. “To Bias: if he is a royal farmer, see to it that he remain unmolested until he is through with the sowing. Year 10. Thoth 24.” The following day (Thoth 25), two officials sent a copy of the petition, preceded by a letter where they explicitly stated that the royal farmer should be left unmolested (P.Dion. 12). In P.Tebt. 1 43 (= M.Chr. 46; 117 BCE, Alexandria), a petitioner requests from the King and Queen protection against molesting (38–42), which is granted in an apostil to the petition (44–45); also P.Coles 14 (13 CE, Aueris [?; Arsinoite nome]).

To come back to the strategos Seleukos in BGU 8 1810, in 51 BCE he issued a pistis on behalf of a man called Herakleides and of his daughter, granting them a respite of thirty days. In producing the safe-conduct, his scribe used the same general pattern that is also on display in P.Bingen 45 and PSI 8 969, with ample top and bottom margins (the left margin is not preserved), and a generous line spacing. Here again, the scribe had to adapt the pattern to a specific use: in the first two lines, the opening [Σέλ]ευκος Ἡρακλείδηι Εὐτυχίδου [καὶ τ]ῆι τούτου θυγατρὶ Εἰρήνηι “Seleukos to Herakleides son of Eutychides and to his daughter Eirene” is close to that of a letter, but it does not include the customary χαίρειν, because the strategos does not enter an epistolary relationship with the recipient of the document. Likewise, at the end, only a date is provided, shifted to the right, without the greeting (ἔρρωσο) that is customary for a proper letter. The date is indispensable because the pistis offers a protection of thirty days from the time it is granted.

Four years later, a successor of Seleukos, Eurylochos, issues a similar safe-conduct for another individual and his scribe makes use of the same layout.

BGU 8 1811 (47 BCE).

From a typological perspective, it seems appropriate to describe the pistis as a specific sub-category, with its established rules regarding format, layout and structure: in 7 BCE, BGU 16 2609, a perfectly preserved pistis from the Herakleopolite nome displays the same squarish sheet size, ample margins (top, bottom, and left, also a narrower right margin after the sheet was cut from the roll) and generous line spacing, as in BGU 8 1810 (written more than four decades earlier). The only notable difference is the presence of a greeting χαίρειν in the opening.

Fig. 15 BGU 16 2609, image and structural display (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 25251. https://berlpap.smb.museum/05162/).

This is clearly an adaptation from the letter type: BGU 16 2610, 2611 and 2612, which belong to the Athenodoros Archive, are letters, written by the same scribe as BGU 16 2609, and they follow the same broad typological pattern.

Another pistis, SB 5 7636 (70 or 41 BCE, provenance unknown), follows a similar pattern, opening with the name of the strategos followed by the addressee and his father (but no χαίρειν). The main text ends with the date (but no ἔρρωσο). The text seems entirely preserved, although the top margin is missing. In the bottom margin, the strategos instructed his scribe to add a personal remark (10–13): ἔδει δὲ μὴ ἀλόγως ὑμᾶς κεχωρίσθαι. τίς γὰρ ὑμᾶς ἐξέβαλε ἢ τί ἠδικήθητε; “But you should not have been excluded without reason. For who expelled you, or what wrong did you suffer?” The general appearance of the document, the preserved margins, as well as the loose writing style, suggest a less formal production in this case, and the strategos’ personal remark added in the bottom margin strengthens this impression. Nonetheless, the shaping of the text, together with the generous margin below the pistis, confirms the existence of a basic pattern for producing such a document.

The pisteis further illustrate some of the principles followed by scribes in this period. First, the writers can borrow from the rules of a general type, such as a letter, and adapt the formulation to fit the specific purpose of the pistis (e.g. omitting the customary initial χαίρειν in some cases, as well as the final ἔρρωσο). Second, they can apply the layout that prevails for letters by producing a block of text that is placed against the right edge of the sheet (minimal right margin), leaving space above, below, and to the left.

This double pattern was not introduced in the reign of Cleopatra but is attested earlier: for example, in a series of sixteen bank receipts from the Herakleopolite nome, dating all from 82 BCE, a consistent model can be identified.

BGU 14 2401-2416.

To this batch may be added an additional series of twenty-four receipts from the years 87–85 BCE, which indicates that both groups were extracted from mummy cartonnage found in Abusir al-Malaq.

Bagnall and Bogaert 1975 (= SB 14 11309-11328 ; four descripta).

They carry a letter opening (with χαίρειν), and they end with a date but no final ἔρρωσο.

Payment orders from the earlier Ptolemaic period always carry the greeting ἔρρωσο, while in the Roman period there is never a closing formula; Bagnall and Bogaert 1975: 99.

The layout of the text is similar to several types described above, with a block of text placed against the right edge of the sheet (minimal right margin), and ample margins at the top and bottom, as well as a narrower margin on the left. The interlinear spacing is, here too, quite generous. In the following example (BGU 14 2403), the receipt was cancelled by being crossed out.

Fig. 16: BGU 14 2403, image and structural display (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 25198. https://berlpap.smb.museum/05108/).

Declarations on oath

Among the documents recovered from cartonnage found in Abusir al-Malaq, there are several declarations on oath that display a consistent format and layout. Those features are not specific to the Herakleopolite nome: they are also attested in a similar document from the Arsinoite nome.

Table 4 Dimensions of BGU 16 2589, BGU 2 543, BGU 16 2590, 2591, and 2592.

reference date provenance height width margins BGU 16 2589 28 BCE

Techtho

Herakl. nome

27 15.8 Top ca. 15.2 cm, bottom 4.3 cm, left ca. 1.2 cm, right ca. 1.3 cm. BGU 2 543 27 BCE

Aueris

Ars. nome

28 12 Top 8 cm, bottom (from line 18, not taking into account subscription written in margin) 9.5 cm, left 1.2 cm, no right margin. BGU 16 2590 25 BCE

Korphotoi

Herakl. nome

29.5

22

(2 cols.)

Top min. 6.8 cm, bottom 7.3 cm, left ca. 1.9 cm, no margin on right of col. 2. BGU 16 2591 2 BCE

Tertonpetochen

Herakl. nome

25.2 13.5 Top 9.5 cm, bottom 8.5 cm, left 1.5 cm, virtually no right margin. BGU 16 2592 reign of Augustus

Phys

Herakl. nome

26.5 20.5 Top 7 cm, bottom 9 cm, left 2.5 cm, right 7.5 cm.

Two items do not fit the pattern of the documents listed above:

BGU 16 2593 (reign of Augustus, Phebichis [Herakleopolite nome]) is incomplete but preserves a bottom margin of ca. 5.5 cm (the other margins are lost). It was presumably written with a layout comparable to the five cases listed above. P.Amst. 1 28 (3 BCE, Oxyrhynchite nome [?]) displays a horizontal shape (H 10 cm x W 18 cm), with the text running along the fibres; the margins are narrow (top 2.5 cm, bottom 1 cm, left 2.5 cm, right 1.5 cm). Yet it is not clear whether this is a complete sheet, or if there is a part missing. The text makes two mentions of preceding data that do not appear in the document.

1: Φατρῆς ὁ προγεγραμμένος; 3–4: τὴν προκειμέ̣νην ἀναφοράν. The editors suggest the possibility that this document was originally submitted together with another one that contained the data mentioned in P.Amst. 1 28.

Several points emerge from this cursory survey. In the table, the four declarations on oath from the Herakleopolite nome originate from four separate villages, but the scribes all followed a similar pattern for producing their documents, and this pattern corresponds to that of other types described above. In the case of BGU 16 2590, two declarations were made, each occupying a separate column, but the general pattern remains the same as regards the outer margins; the only notable difference lies in the narrow margin between the columns, and in the wider format of the sheet, to accommodate two columns.

BGU 16 2592, a declaration on oath relating to the felling or pruning of trees, also deserves special mention because, contrary to the standard practice, the scribe left a wide margin on the right.

The content of the document is interpreted by Schram 2023: 342–343. The approximate dating rests on a titulature of Augustus in the oath.

On the basis of the dimensions provided by the editor, and of the image available online, it seems that the sheet – which was extracted from mummy cartonnage – was pasted to another sheet on the left where no writing appears.

Fig. 17: BGU 16 2592, image and structural display (Ägyptisches Museum und Papyrussammlung, Staatliche Museen zu Berlin, Berliner Papyrusdatenbank, P 25286. https://berlpap.smb.museum/05195/).

In spite of the unusual right margin, among the types described above, this document would compare well with BGU 16 2651 (see above, a letter from the Athenodoros archive that was extracted from the same cartonnage), with P.Bingen 45 (the Cleopatra ordinance), and with BGU 16 2609 (a pistis, also from the same cartonnage). One may note the sizeable top and bottom margins, as well as a narrower margin on the left. The interlinear spacing is quite generous, and the text ends with a curved paragraphos at the bottom left of the block of text. Those features suggest that, if the hand is not the same as in the rest of the Athenodoros archive, at least this scribe followed similar codes for producing his document.

BGU 2 543, furthermore, a declaration on oath from the Arsinoite nome, shows that the general pattern is not specific to the Herakleopolite nome: by the size of the sheet, its orientation, and the dimensions of the margins on all four sides, it follows a muster comparable to the other documents listed in the table. Thus, to conclude on the declarations on oath, the general impression is one of consistency in the method of production. Those documents are written by skilled scribes who follow a basic set of rules applicable also to other types of documents.

Survival of demotic Egyptian scribal practices

Various documents described above conform to a similar basic pattern; but it was also shown above that village record-offices (γραφεῖα) experimented for some time with a new layout, which was eventually discontinued. In contrast with this experimentation, quite the opposite seems to have happened with contracts that were produced in an environment where demotic Egyptian mixed with Greek, and where scribes retained the old way of writing down a contract.

The Archive of Satabous son of Herieus belongs to the large find of papyri from Soknopaiou Nesos, on the north bank of Lake Moeris, where demotic Egyptian was in current use till the Roman period largely because of the temple of the crocodile-god Sobek.

TM Archive 151.

Among the 40–50 documents that constitute this archive, ranging from 20 BCE till about 75–88 CE, a sub-group of eight papyri may be highlighted for their idiosyncratic appearance.

CPR 15 1-4; SB 1 5231, 5247, 5275; P.Vindob. inv. G 31933 (= Hoogendijk 2006).

They are Greek translations from demotic Egyptian contracts, and in one specific case the same sheet displays both a text in Egyptian and in Greek.

Translation: e.g. SB 1 5231.1 (11 CE, Psinachis [Arsinoite nome]): [ἀν]τί[γ]ρ[αφ]ον Αἰγυπτίας π[ρά]σ̣ε̣ως Ἑ[λ]ληνιστὶ με̣θ̣ηρμηνε[υ]μένης [κα]τὰ [τὸ δ]υνατόν “copy of an Egyptian contract translated into Greek as much as possible”. Demotic Egyptian alongside Greek in P.Dime 3 5 (= P.Lond. 2.262 [p. 176] = M.Chr. 181; 11 CE, Soknopaiou Nesos), plate available in Wessely 1900: Tab. 1a and 1b.

The horizontal format of these contracts is striking, as in CPR 15 2 (11 CE, Soknopaiou Nesos): H 27.6 cm × W 58.2 cm. It continues well beyond the reign of Augustus.

The dimensions of P.Vindob. inv. G 31933 (50 CE) are comparable. Hoogendijk 2006: 207: H 25.2 cm × W 54.7 cm.

Although they were found within the context of other Greek documents from the reigns of Cleopatra and Augustus, the peculiar format of these Greek translations of demotic Egyptian contracts recalls some documents prepared by bilingual agoranomoi in Pathyris between the mid-second and the early first century BCE.

E.g. BGU 3 997 (103 BCE; H 16 cm × W 37.5 cm); P.Köln 1 51 (99 BCE; H 15 cm × W 41 cm); P.Lond. 3 1209 (p. 20; 89 BCE; H 15.2 cm × W 58.4 cm). On the linguistic aspects of the bilingual contracts in Pathyris, Vierros 2012.

In other words, they do not reflect an attempt at experimentation, but on the contrary in this case Egyptian scribes retained the ancient Egyptian way of preparing contracts.

Tax receipts on ostraca

Papyrus and ostracon are not only made with fundamentally different materials: the rules for using one or the other writing support are also not the same. Many ostraca were published in the century and a quarter following Wilcken’s Griechische Ostraka aus Aegypten und Nubien.

Wilcken 1899.

The general observations he made, however, remain by and large valid to this day. Wilcken noted that, in the Roman period, tax collectors were liturgists or contractors who had to purchase their writing material at their own cost, and therefore they used ostraca because this was a cheaper writing support.

Wilcken 1899: 11.

For official communication, only papyrus was allowed. Ostraca are in frequent use in the Ptolemaic as well as in the Roman period but become rarer after the third century CE. Wilcken raised the possibility – without providing a definite answer to the question – that the decrease in the use of ostraca in the Byzantine period may have been caused by the fiscal reforms that took place under Diocletian, with the consequence that those in charge of collecting taxes were not allowed to write their receipts on ostraca anymore.

Wilcken 1899: 13: “Vielleicht ist auch dies eine der zahlreichen Neuerungen der neuen Zeit, dass es dem nunmehr mit der Steuerhebung betreuten Beamtenpersonal (in der Regel) untersagt war, sich der Ostraka zum Quittiren zu bedienen.” Note the ‘Beamtenpersonal’, which applies specifically to civil servants, not liturgists.

Of the ca. 321 Greek ostraca recorded between 51 BCE and 14 CE, 90% come from Upper Egypt, 10% from the Arsinoite nome. From a typological perspective, the rules for producing what are, for the most part, receipts for payment or delivery, are rather patchy. The receipts are much less elaborate than the texts we find on papyrus. This is due to the small size of ostraca, to their irregular shape, their coarse surface, as well as to the limited complexity of the texts.

Wilcken could already identify two basic sub-types of receipts on ostraca, and he provided a fine analysis of the formulary, showing how the scribes followed the general rules that governed those sub-types but introduced small variations corresponding to the place of writing. There were other cases where, on the contrary, the formulary was identical in Elephantine and Syene, and downstream in Thebes.

Wilcken 1899: 118.

In the period considered here, the first sub-type is now best illustrated by the New York University ostraca, published in 2022.

These ostraca were purchased in Cairo in 1932, probably from the well-known dealer Maurice Nahman; Baetens et al. 2022: 2 (= O.NYU).

The 59 items considered here (a few were too badly preserved to retain much attention) date from the years 1 to 12 CE, except for two items with an uncertain date.

O.NYU 7 and 8 (26 or 18 BCE?),

Their provenance is defined broadly as Upper Egypt and they preserve receipts for the tax on embalmers or for burial, poll tax, and money tax. This set of receipts invariably starts with a date, followed by the statement of payment, the name of the individual making the payment, and the item being paid or delivered. The text is arranged on several lines, forming a block with a rectangular shape, not necessarily aligned on the edges of the potsherd: the sides of the ostracon may be aslant, and the text does not take up the whole surface.

E.g. O.NYU 45 (8 CE); O.NYU 49 (9 CE).

Fig. 18: O.NYU 49, image and structural display (Image courtesy of New York University Collection of Papyri, Ostraka, and Cuneiform Tablets).

The only element that stands out at first glance is the oversize epsilon of the initial ἔτους, indicating the regnal year. This is more than an ordinary Greek character: it works as a symbol to signal the presence of a date, and to immediately suggest a receipt issued by an authority or a bank. In contracts, a date placed at the beginning of the document indicates in most cases that the document was written in a notarial office (as opposed to private contracts, where the date is placed at the end).

Wolff 1975: 349; Wolff 1978: 87 and 107.

Such a basic layout appears also in the Arsinoite nome, e.g. O.Deiss. 74.1–5 (6 BCE), with a text in one rectangular block starting with an oversize epsilon for the date. In O.Mich. 3 994 and 998 (1 BCE), the initial epsilon is replaced with the symbol ∟ (= ἔτους). The layout described here is limited neither to the Arsinoite nome nor to ostraca: it appears also on a bank receipt for dike-tax from Oxyrhynchus written on papyrus, P.Oxy. 78 5172 (7 CE, H 12.2 × W 8.3 cm). The simultaneous use of demotic Egyptian and Greek in receipts on ostraca is common in Upper Egypt, but not downriver.

E.g. Elephantine: O.Cair. GPW 100 (18 BCE); Apollonopolis: SB 20 14438 (2 BCE); Thebes: P.Lugd. Bat. 19 25 (46 BCE); Koptos: O.Strasb. 1 313 (14 BCE).

The second sub-type consists of receipts (and a handful of payment orders) written in letter form.

Payment orders: O.Fay. 11 (25 BCE) and 12 (6 BCE); O.Mich. 1 17 (4 BCE), all three from the Arsinoite nome.

They start with the typical opening of a letter (X to Y χαίρειν, the latter often abbreviated), followed by the statement of receipt or the order of payment, and finally a date. They are found in Upper Egypt as well as in the Arsinoite nome.

E.g. BGU 6 1459 (69 or 40 BCE, Elephantine); O.Camb. 51 (8 CE, Hermonthis [?]); O.Wilck. 1009 (19 [?] BCE, Thebes); O.Petr. Mus. 114 (6 CE, Myos Hormos); O.Mich. 1 17 (4 BCE, Arsinoite nome).

Tax receipts written on ostraca undergo a process of evolution. The letter form is attested in the Ptolemaic and Roman periods, but in the latter, subscriptions start appearing at the bottom of receipts, with the tax collector acknowledging the payment in his own hand.

Wilcken 1899: 82.

In bilingual receipts, the main text may be written in demotic Egyptian, while the Greek text is limited to the subscription to the receipt.

Wilcken 1899: 112, referring to O.Wilck. 766 (3 BCE, Thebes).

There are minor additions, such as when a scribe writes οὐθέν σοι ἐνκαλοῦμε ̀ν ́ “we make no (further) claim”.

O.Bodl. 2 1168.6–8 (15 BCE, Thebes); BGU 6 1459.9 (69 or 40 BCE, Elephantine).

This clause indicates that the receipt was not issued for a partial payment, but on the contrary for the final payment.

Wilcken 1899: 62.

In the Augustan period, the initial date found in the first sub-type disappears from receipts, to be replaced with the letter form (second sub-type) exclusively, with a date at the end.

Wilcken 1899: 111.

This shift applies also to receipts for payments in kind, where the letter form becomes standard in the Roman period.

Wilcken 1899: 97 and 103.

Wilcken’s detailed listing of the variations in the formulary shows that the scribes, relying on the general model of a letter, adapted it to the specific purpose of receipts. The most salient changes are the frequent loss of the initial χαίρειν and the nearly total absence of the final ἔρρωσο.

Conclusion

Given the distribution of the available papyri and ostraca in the period under consideration, the ‘black hole’ of the 30s cannot be explained away through the unevenness of finds. Not only does it appear in both papyri (in the north) and in ostraca (in the south), but distinct types of documents bear witness to a gap shortly before the end of Cleopatra’s reign. It seems therefore preferable to assume that the activity of scribes was impeded for a few years, before it took off again after Egypt had been turned into a Roman province.

The scribes, however, did not reinvent their trade during this time of political change. It was shown above that the format and layout of petitions found in the Herakleopolite nome in the mid-first century BCE correspond closely to what was attested in the Hermopolite nome half a century earlier (P.Dion. 11), and that the patterns in use in the reign of Cleopatra do not evolve significantly under the rule of Augustus. Conservatism can also be observed in the practice of Egyptian scribes who translate Egyptian contracts into Greek, retaining the traditional Egyptian format and layout in Soknopaiou Nesos. In the absence of any known handbooks for the writing of documents, we may assume that scribes learned by imitating the work of other scribes. In the Roman period, models were issued and circulated, and it seems likely that this was already the case in the late Ptolemaic period. This would explain a consistent pattern of experimentation in the grapheia in the Arsinoite nome.

A comparison of several clusters, as well as the addition of external documents, shows that scribes working in offices at a high level of the administrative structure tended to follow a simple set of implicit rules to produce their documents. These rules are attested both under Cleopatra and under Augustus. The Athenodoros Archive provided evidence indicating that scribes could prepare a document in several stages: first a fast copy under dictation, then a version that would include the appropriate layout but still required some corrections, and eventually the final version to be circulated. Contrary to the earlier Ptolemaic practice, where both vertical and horizontal formats were in common use, the scribes of the later first century opted mostly for vertical sheets. Letters from the Athenodoros Archive show that, if writers could make a prior assessment of the length of the text to be written, they placed their block of text against a virtual right margin corresponding to the place where the sheet would be detached from the roll after the writing was completed. The intended layout was provided with wide margins at the top and bottom, and a narrower margin on the left. In the best of cases, especially when the text was short, a balanced appearance could be achieved by increasing the line spacing. This resulted in a vertical symmetry, enhanced by the generous top and bottom margins. This pattern, however, is not an innovation from the reign of Cleopatra: it was shown that a set of sixteen bank receipts from the Herakleopolite nome dating all from 82 BCE already displays the same pattern.

The top and bottom margins offered space that could be used according to the purpose of documents: a registration docket might be placed at the top, or an additional note at the bottom. In the case of Alexandrian synchoreseis, a taller bottom margin affected the symmetry but made it possible to add the agreement by one of the contracting parties.

The longer the texts, however, the more difficult it became to follow this basic pattern. Line spacing could be made more compact, the margins could also be reduced, but in the end the result might become asymmetrical: the block of text would take up most of the space saved for a margin, or on the contrary leave an excessively ample bottom margin.

It seems likely that the scribal offices of the high administration in Alexandria established an implicit standard. The Cleopatra papyrus, together with a handful of other conspicuous examples, show that the pattern described above was applied in a very competent manner by scribes who also knew how to adapt it to some specific needs, as when the sovereign issued an ordinance, or when a high official wrote to a strategos. Whereas the general pattern is most visible among documents from the high administration, it was also followed by scribes with a more rudimentary training, who were not quite as skilled at mastering the implicit rules. This is illustrated by the letters that constitute the Tomos of Asklepiades. On the geographical dimension, the Herakleopolite nome is overrepresented in the pattern identified among documents from Cleopatra’s and Augustus’ reign, but parallels from the Arsinoite and Oxyrhynchite nomes show that this phenomenon was not restricted to a specific area.

Letters, contracts, petitions and ordinances are not the only types where the basic pattern identified above is represented: declarations for small livestock (sheep and goats) are prepared according to the same general rules, and so are temporary grants of protection, or declarations on oath.

Receipts written on ostraca did not require the same skill from scribes who produced a single short block of text. The only conspicuous feature was, in some of the receipts, the enlarged epsilon that marked the place of regnal year in the top left corner. Ostraca are nonetheless useful in showing how malleable the standard formulas could become. Receipts in the form of a letter would sometimes display the standard X to Y χαίρειν, but often the scribes rearranged the traditional elements, or removed one of them. Coming back to papyri, this is also what the scribes of the high administration did when they used a general pattern corresponding to a letter but adapted it for the Queen to issue an ordinance. Conversely, the idiosyncratic heading – derived from that of a letter – used in enteuxeis remained unaltered till the end of the Ptolemaic period.

The typological pattern of petitions to officials of the Herakleopolite nome rests on models that go back to the second century BCE, and it is still attested in the first century CE. This raises the more general question of whether the shift from the Ptolemaic period (death of Cleopatra) to the Roman period (Augustus’ seizure of Egypt) also meant a momentous change in the scribal practices in Egypt. In other words, do documents on papyrus become recognizably Roman early in the reign of Augustus? In spite of the black hole in the 30s, there is sufficient evidence to show that the scribes, once they had resumed their activity, did not fundamentally modify the patterns on which they were working.

Ferretti 2024: 196.

The edict circulated by Proculus in 12 BCE (BGU 16 2558) does not indicate a change in practice among the scribes of the high administration. Likewise, the format and layout of pistis documents remain by and large unaltered for half a century, with a specimen (BGU 16 2609) in the year 7 BCE displaying no notable difference. It is also no coincidence that, as noted above, high officials drop the Ptolemaic greeting ἔρρωσο only in the late first century CE, replacing it with a more elaborate ἐρρῶσθαι βούλομαι. If a shift took place at all, this is more likely to have happened when the Roman administration asserted a stronger position around the mid-first century CE. Concrete signs of this change are to be found, for example, in the creation of the registries of property, the δημοσία βιβλιοθήκη and the βιβλιοθήκη ἐγκτήσεων, or in the abolition of gymnasia in the villages, together with the establishment of examination of civic status (ἐπίκρισις).

Creation of the registries of property: Schubert 2019: 287–289. Gymnasia and epikrisis: Broux 2013.

But this is another story.

Bibliography Baetens, G., Bagnall, R.S., Caputo, C., Mazy, É., and D.M. Ratzan (2022) Ostraka in the collection of New York University. New York. [= O.NYU]. Bagnall, R.S. and R. Bogaert (1975) “Orders for payment from a banker’s archive: Papyri in the collection of Florida State University,” Anc. Soc. 6: 79–108 [= SB 14 11309-11328; repr. in Bogaert, R. (1994), Trapezitica Aegyptiaca. Pap. Flor. XXV, Firenze: 219–244]. Bagnall, R.S. and R. Cribiore (2006) Women’s letters from ancient Egypt 300 BC – AD 800. Ann Arbor. Bagnall, R.S. and P. Derow (2004) The Hellenistic period. Historical sources in translation. Revised ed., Malden MA - Oxford - Victoria. Bonagura, G., Chang, R.-L., Ferretti, L., Fogarty, S., Nury, E. and P. Schubert (2019-2026) grammateus [dataset as of 02.2023] https://grammateus.unige.ch. Bonagura, G., Chang, R.-L., Ferretti, L., Fogarty, S., Nury, E. and P. Schubert (2023) Description of Greek Documentary Papyri: Order to Pay (grammateus project) [Accessed: May 31, 2024] https://grammateus.unige.ch/descriptions/letter_order. Broux, Y. (2013) “Creating a new local elite: The establishment of the metropolitan orders of Roman Egypt,” APF 59: 143–153. Capponi, L. (2005) Augustan Egypt. The creation of a Roman province. New York - Milton Park Abingdon. Capponi, L. (2016) “C. Calpurnius Proculus and an example of Greek stenography under Augustus,” in Derda, T., ׅŁajtar, A., and J. Urbanik (eds.), Proceedings of the 27th international congress of papyrology. Warszawa: 1709–1723. Causo, S. (2025) “Medium, procedures and conventions. A framework for analysing the materiality of documents on papyrus,” in Bentein, K. (ed.), Everyday communication in Antiquity: Frames and framing. Ghent. Clarysse, W. (1976) “Notes on the use of the iota adscript in the third century B.C.,” CÉ 51: 150–166. Claytor, W.G. (2014) “Rogue notaries? Two unusual double documents from the late Ptolemaic Fayum,” JJP 44: 93–115. Claytor, W.G., Litinas, N., and E. Nabney (2016) “Labor contracts from the Harthotes archive,” BASP 53: 79–119. Claytor, W.G. and P. van Minnen (2021) “An antichretic loan from early Roman Alexandria,” ZPE 217: 158–164. Essler, H. (2025) “Die Korrespondenz zu den Papyrus-Grabungen und Survey in Abusir el-Meleq und Darb Gerz (Philadelphia),” AnalPap 39: 163–217. Ferretti, L. (2024) τῷ δεῖνι παρὰ τοῦ δεῖνος. A history of the hypomnema in Greek papyri from Egypt (300 BCE – 100 CE). Unpublished PhD thesis, University of Geneva. Ferretti, L., Fogarty, S., Nury, E., and P. Schubert (2020) “Cession of cleruchic land: from procedure to format,” ZPE 215: 201–210. Ferretti, L., Fogarty, S., Nury, E., and P. Schubert (2023a) grammateus, Classification of Greek documentary papyri: Transmission of information [Accessed: December 12, 2024] https://grammateus.unige.ch/classification/ti https://doi.org/10/gs9qn2. Ferretti, L., Fogarty, S., Nury, E., and P. Schubert (2023b) grammateus, Description of Greek Documentary Papyri: Double Document [Accessed: September 23]. Fischer-Bovet, C. (2023), “Cleopatra VII, 69–30 BCE,” Oxford Classical Dictionary online https://doi.org/10.1093/acrefore/9780199381135.013.1672. Gignac. F.T. (1976) A grammar of the Greek papyri of the Roman and Byzantine periods. Vol. I, Phonology. Milano. Grenfell, B.P., Hunt, A.H., and D.G. Hogarth (1900) Fayûm towns and their papyri. London. [= P.Fay.] Habermann, W. (1998) “Zur chronologischen Verteilung der papyrologischen Zeugnisse,” ZPE 122: 144–160. Haensch, R. (2008) “The romanity of Roman Egypt: a growing consensus?,” in Piso, I. (ed.), Die römischen Provinzen. Begriff und Gründung. Cluj-Napoca: 81–105. Haensch, R. and C. Kreuzsaler (2020) “ Drei Kandidaten, bitte! Die Rolle des praefectus Aegypti bei der Ersatznominierung öffentlicher Funktionsträger zu Beginn des 2. Jahrhunderts,” Chiron 50: 189–215. Hagedorn, D. (2006) “Bemerkungen zu Urkunden,” ZPE 156: 165–168. Hagedorn, D. and A. Jördens (2006) “Ein unbekannter Proculus und das Verbot der paenula: Eine Neuedition von BGU XVI 2558,” ZPE 156: 169–178. Heidelberger Gesamtverzeichnis der griechischen Papyrusurkunden Ägyptens (HGV) https://aquila.zaw.uni-heidelberg.de (consulted throughout 2024). Hoogendijk, F.A.J. (2006) “P.Vindob. G 31933: Griechische Übersetzung einer demotisch-griechischen Hauskaufurkunde,” ZPE 156: 199–216. Husselman, E.M., Boak, A.E.R., and W.F. Edgerton (1944) Papyri from Tebtunis, Part II. Ann Arbor - London. [= P.Mich. 5] Jördens, A. (2001) “Papyri und private Archive. Ein Diskussionsbeitrag zur papyrologischen Terminologie,” in Cantarella, E. and G. Thür (eds.), Symposion 1997. Vorträge zur griechischen und hellenistischen Rechtsgeschichte (Altafiumara, 8. – 14. September 1997). Köln - Weimar - Wien: 253–268. Jördens, A. (2017) “Entwurf und Reinschrift – oder: Wie bitte ich um Entlassung aus der Untersuchungshaft,” Chiron 47: 271–302. Kruse, Th. (2002) Der königliche Schreiber und die Gauverwaltung. Untersuchungen zur Verwaltungsgeschichte Ägyptens in der Zeit von Augustus bis Philippus Arabs (30 v.Chr. – 245 n.Chr.). 2 vols., München - Leipzig). Langellotti, M. (2012) L’allevamento di pecore e capre nell’Egitto romano: aspetti economici e sociali. Bari. Legras, B. (2021) Cléopâtre l’Égyptienne. Paris. Lenger, M.-Th. (1980) Corpus des ordonnances des Ptolémées. Revised edition with addenda and corrections, Bruxelles. [= C.Ord. Ptol.] Maehler, H. (1983) “Egypt under the last Ptolemies. In memory of Sir Eric Turner,” BICS 30: 1–16. Mascellari, R. (2021) La lingua delle petizioni nell’Egitto romano. Evoluzione di lessico, formule e procedure dal 30 a.C. al 300 d.C. 2 vols., Firenze. Mascellari, R. (2024) “Il saluto conclusivo nelle lettere degli alti funzionari provinziali: ἐρρῶσθαι … βούλομαι,” APF 70: 78–96. McConnell, J.R., Sigl, M., Plunkett, G., Burke, A., Kim, W.M., Raible, C.C., Wilson, A.I., Manning, J.G., Ludlow, F., Chellman, N.J., Innes, H.M., Yang, Z., Larsen. J.F., Schaefer, J.R., Kipfstuhl, S., Nojtabavi, S., Whilhelms, F., Opel, Th., Meyer, H. and J.P. Steffensen (2020) “Extreme climate after massive eruption in Alaska’s Okmok volcano in 43 BCE and effects on the late Roman Republic and Ptolemaic Kingdom,” PNAS 117: 15443–15449 Monson, A. (2012) From the Ptolemies to the Romans. Political and economic change in Egypt. Cambridge. Monte, A. (2018) Neue Quellen zum griechisch-römisch-byzantinischen Ägypten: Erstedition von fünfzehn griechischen Papyrustexten der Berliner Papyrussammlung. Unpublished PhD thesis, Humboldt Universität zu Berlin, available online https://doi.org/10.18452/21422. [= P.Berl. Monte] Peremans, W. and E. van’t Dack (1957) “Notes concernant PSI VIII 969,” Eos 48: 121–132 [= Symbolae Raphaeli Taubenschlag dedicatae II]. Preisendanz, K. (1933) Papyrusfunde und Papyrusforschung. Leipzig. Ricketts, L.M. (1980) The administration of Ptolemaic Egypt under Cleopatra VII. Microfilm PhD diss., Ann Arbor. Ricketts, L.M. (1992) “The administration of late Ptolemaic Egypt,” in Johnson, J.H. (ed.), Life in a multicultural society: Egypt from Cambyses to Constantine and beyond. Chicago: 275–281. Roberts, C.H. and E.G. Turner (1952) Catalogue of the Greek and Latin papyri in the John Rylands Library, Manchester. Manchester. [= P.Ryl. 4] Salmenkivi, E. (2002) Cartonnage papyri in context. New Ptolemaic documents from Abu Sir al-Malaq. Helsinki. [= P.Berl. Salmen.] Sarischouli, P. (2000) “Spätptolemäische Urkunden: ΣΥΓΧΩΡΗΣΙΣ-Vertrag und Eide,” in Melaerts, H. (ed.), Papyri in honorem Johannis Bingen octogenarii. Leuven: 213–228. [= P.Bingen 45–47] Sarri, A. (2018) Material aspects of letter writing in the Graeco-Roman world. Berlin - Boston. Schäfer, D. (1933) “Zu den ptolemäischen pisteis,” Philologus 88: 296–301. Schram, V. (2023) L’arbre et le bois dans l’Égypte gréco-romaine. Paris. Schubart, W. (1913) “Alexandrinische Urkunden aus der Zeit des Augustus,” APF 5: 35–131. Schubart, W. and D. Schäfer (1933) Spätptolemäische Papyri aus amtlichen Büros des Herakleopolites. Berlin. [= BGU 8] Schubert, P. (2018) “Warrants: some further considerations on their typology,” BASP 55: 253–274. Schubert, P. (2019) “Transmission of cleruchic land. A model to describe the procedure,” APF 65: 280–316. Schubert, P. (2025, forthcoming), “Comment s’adresser poliment à une reine: une note à propos de SB VI 9065,” CÉ 100. Skeat, T.C. (1962) “Notes on Ptolemaic chronology,” JEA 48: 100–105. Thomas, J.D. (1982) The epistrategos in Ptolemaic and Roman Egypt. Part 2: The Roman epistrategos. Opladen. Thompson, D.J. (1994) “Egypt, 146–31 B.C.,” CAH2 IX: 310–326. van Minnen, P. (2000) “An official act of Cleopatra,” AncSoc 30: 29–34. van Minnen, P. (2001) “Further thoughts on the Cleopatra papyrus,” APF 47: 74–80. van Minnen, P. (2003) “A royal ordinance of Cleopatra and related documents,” in Walker, S. and S.-A. Ashton (eds.), Cleopatra reassessed. London: 35–44. van Minnen, P. (2022) “Romans in Egypt before the Constitutio Antoniniana: When, where, and why? ,” in Capasso, M., Davoli, P. and N. Pellé (eds.), Proceedings of the 29th international congress of papyrology. Lecce, 28 July – 3 August 2019. Lecce: 1002–1025. van Minnen, P. (2023) “The diplomatics of synchoreseis that are not drafts,” JJP 53: 81–114. Vierros, M. (2012) Bilingual notaries in Hellenistic Egypt. Coll. Hell. 5, Brussel. Wasserstein, D.J. (2014) “Cleopatra’s tongues,” SCI 33: 275–281. Wessely, C. (1900) Papyrorum scripturae Graecae specimina isagogica. Leipzig. Wilcken, U. (1899) Griechische Ostraka aus Aegypten und Nubien. Vol. I. Leipzig - Berlin. Wilcken, U. (1927) Urkunden der Ptolemäerzeit (ältere Funde). Vol. I. Berlin - Leipzig. [= UPZ 1] Wolff, H.-J. (1975) “The so-called private protocol,” in Rea, J.R. and E.G. Turner, Proceedings of the XIVth international congress of papyrologists, Oxford, 24–31 July 1974. London: 349–356. Wolff, H.-J. (1978) Das Recht der griechischen Papyri Ägyptens in der Zeit der Ptolemaeer und des Prinzipats II: Organisation und Kontrolle des privaten Rechtsverkehrs. München. Yiftach-Firanko, U. (2008) “Who killed the double document in Ptolemaic Egypt?,” APF 54: 203–218. Youtie, H.C. (1973) “P. Mich. Inv. 6051 = Sammelbuch V 7532,” ZPE 12: 161–171. Zimmermann, K. (2002) “P. Bingen 45: Eine Steuerbefreiung für Q. Cascellius, adressiert an Kaisarion,” ZPE 138: 133–139.