Explicit Name Change in Roman Egypt

Y. Broux N AME CHANGE IN R OMAN E GYPT Depending on the circumstances, people are addressed or referred to in different ways. In western society, the surname functions as identifier in more formal circumstances, while in close family or friendship circles, nicknames and hypocoristics prevail. These latter, informal types of names were probably no less widespread in Egypt1, while for official occasions the patronymic was appended to the name. Less common, however, is the practice of actual name change, the discarding and replacement of one name in favor of another. We generally receive our names at birth and stick with them until the end. A name can be officially changed in the wake of some dramatic event (witness protection program, ridicule because of the name, …), but this is seldom. In several societies, however, name change also reflects a change in identity, e.g. at the beginning of a new life phase or with a change in status. In Britain and the US, women often still adopt their husbands' family name, while in Bali both men and women are designated with their children's name upon parenthood ('father-of-X', a practice called teknonymy). The adoption of a Latin papal name is a well-known tradition symbolizing the acquisition of a new role2. There is some evidence for name change in Roman Egypt, albeit scarce and the reasons behind it not often very clear. In some cases, the elevation of status seems to have been the decisive factor. It was standard upon receiving Roman citizenship, for instance, and van Minnen suggests that the examples encountered in P. Amst. 1 72 were the result of the introduction of the βουλή in the metropoleis in AD 201 (see below)3. In many cases the context is too vague to uncover any underlying motivations though. Most of the expressions introducing a change of name have hitherto been treated as double names4. This categorization does not do them justice, however, as double names are a fundamentally distinct concept: they consist of two official names are meant to be used simultaneously as a unity5, while a name change implies the use of a new name instead of an old one. Morever, I generally do not agree with the proposed explanations for these formulas. This article therefore provides a new overview of the different contexts of name change, in which I will suggest some amendments to previous interpretations and even strike a category off the list. It starts with a brief survey of the procedure of name change, followed by different occasions as a result of which name change could take place, i.e. elevation to Roman citizenship or "Greek" status. I have also included examples in slave sale contracts. Although these slaves do not obtain a better social position, the transition from one owner to another does place them in a new environment and they therefore also start a new phase in their life. Finally, the name changes for which no immediate 1 Y. BROUX, Elite Strategy vs Popular Characterization: Double Names in Roman Egypt (Studia Hellenistica), Leuven (forthcoming). 2 R. D. ALFORD, Naming and Identity: A Cross-Cultural Study of Personal Naming Practices, New Haven, 1987, 85-94. 3 P. VAN MINNEN, 'A Change of Names in Roman Egypt after A.D. 202?', ZPE 62 (1986), 87-92. 4 R. CALDERINI, 'Ricerche sul doppio nome personale nell’ Egitto greco-romano', Aegyptus, 21 (1941), 221-260. 5 BROUX, Elite Strategy vs Popular Characterization (forthcoming). 1 Y. Broux cause can be found are discussed. Most examples occur in tax-related documents (introduced by ἀνθ' οὗ or ὁ διὰ λόγων), while some are encountered in the context of land tenure (ὁ δι'αἱρέσεως and the abbreviations starting with ὁ δι( ), found exclusively in P. Bouriant 42). Requesting a name change The name changes discussed here only concern official names, that is, names that are formally acknowledged by an official institution. A person's unofficial names (i.e. bynames or nicknames) also change in the course of a lifetime, but they usually evolve spontaneously and do not need official ratification. In Egypt an official name generally consisted of a single name; double names were only bestowed in slect circles6. The notion of a hereditary family name, such as the Roman nomen gentilicium, was unknown. Instead, ancestors and titles (if applicable) were added to form a sort of official 'identification cluster', e.g. Σεραπίωνι τῷ καὶ Ἀπολ̣λ̣ωνιανῷ Σπαρτᾶ γυμνασια̣[ρ]χησαντα τῆς Ὀξυρυγχειτῶν πόλεως γενομένῳ ἐπισκέπτῃ Ὀάσε[…]7. Sometimes, if a cluster was too long, elements were left out; this is indicated by the expression καὶ ὡς χρηματίζει, as in Τιβέριος Κλαύδιος Δίδυμος καὶ ὡς χρηματίζει8. If one of the names in the cluster had changed, this was declared as well (see, for instance, Πεκμηις ἀνθ’ οὗ Εὐρήμων below). Proper identification was of vital importance for the assessment of status, which brought with it welcome privileges, such as tax reductions or exemptions. Acting under false identity was therefore not taken lightly, as the Gnomon of the idios logos attests: οἱ ἀκαταλλήλως χρηματίζ[ον]τ̣ες τεταρτολογ̣ο̣ῦ̣ν̣τ̣α̣ι̣ καὶ οἱ εἰδότες καὶ συνχρηματίσαντες αὐτοῖ̣ς̣ [τε]ταρτολογοῦνται ('those who officially act in a non-regulatory manner will face confiscation of one-forth [of their property] and those knowing and acting together with them will face confiscation of one-fourth [of their property]')9. Not much is known about the official procedure to obtain a new name. The only direct evidence is found in SB 18 13175, col. 5 (TM 18337, AD 194). This is a copy of a letter by Eudaimon requesting that his parents' names be changed, from Psois to Heron and from Tiathres to Didyme (for the formula used, see below). The letter is addressed to Claudius Apollonius, head of the department of the ἴδιος λόγος, one of the highest officials in the province. Originally, this equestrian magistrate was in charge of state property, gradually acquiring authority over all matters concerning inheritance and also status10. The department's reply to Eudaimon’s request is added to the letter: the name change is granted, μηδενὸς [δη]μοσίου ἢ ἰδιωτικοῦ καταβλαπ[το]μένου ἐφίημι (‘I grant it, since there is no public or private damage’)11. The document gives no clues to whether the head of the idios logos himself settled requests for name change, or if other were consulted. Moreover, since Eudaimon’s letter is very short, giving 6 A distinction should be made between 'actual' double names, introduced by the formulas ὃς καὶ and ὁ καὶ, and formulas introducing bynames (such as ἐπικαλούμενος and ἐπικεκλημένος), formerly treated as one and the same: BROUX, Elite Strategy vs Popular Characterization (forthcoming). 7 P. Oxy. 36 2793, l. 3-6 (TM 26873, AD 100-299). 8 P. Oxy. 6 908, l. 6-8 (TM 20371, AD 199). See Y. BROUX, S. COUSSEMENT and M. DEPAUW, 'καὶ ὡς χρηματίζει and the Importance of Naming in Roman Egypt', ZPE 174 (2010), 159-166. 9 BGU 5 1210, § 42. 10 P. R. SWARNEY, The Ptolemaic and Roman Idios Logos (American Studies in Papyrology 8), Toronto, 1970. 11 Line 21. 2 Y. Broux no information whatsoever on his or his parents’ status, it can be assumed that a background check was performed in order that a well-considered decision could be made. The next logical step would then be to dispatch this decision to all relevant officials, such as the στρατηγός, the keepers of the public archives, perhaps even to the ἄρχοντες of a nome capital or Greek polis if the individual in question belonged to a privileged rank, to ensure proper referral in the future. SB 18 13175 might well be such a copy, for it is addressed to the βασιλικὸς γραμματεύς who was acting as deputy strategos at the time (line 1-2)12. Roman citizenship There are some examples where people indicate that their name was changed upon receiving Roman citizenship. The exact formulation varies, but the term ῾Ρωμαίων πολιτεία or an alternative designation of the citizenship is always included. As a rule, when taking on the tria nomina one's original given name was retained and "recycled" as the cognomen, while the emperor's name was adopted as a nomen. They include the following examples (table 1): Publication […]ιω ὡς δὲ πρὸ τῆς Ῥωμαίων πολιτ[είας …] ('(or) as before Roman citizenship') [Λουκίῳ Σεπτιμί]ῳ Αὐρηλίῳ Πτολεμαίῳ τῷ καὶ Ἀμμωνίῳ ὡς δὲ πρὸ τῆς ῾Ρωμαίων [πολιτείας χρημα]τίσαντι Πτολεμαίῳ τ[ῷ κ]αὶ Ἀμμωνίῳ Νίννου τοῦ καὶ Ἀνουβίωνος υἱοῦ Ἥρων̣ο̣ς ('Lucius Septimius Aurelius Ptolemaios-Ammonios, (or) as he officially acted before Roman citizenship as Ptolemaios alias Ammonios son of NinnosAnoubion son of Heron') Ἀλκίμου Ἑρμίου [τοῦ] Ἀπολλωνίου … ὕστερον τυχόντος τῆς Ῥωμαίων πολ̣[ι]τείας καὶ χρηματίσαντ(ος) Μάρκου Αὐρηλίο[υ] Ἀλκίμου τοῦ καὶ̣ Ἑρ̣μίου̣ ('Alkimos son of Hermias, grandson of Apollonios … who later obtained Roman citizenship and who acted as Marcus Aurelius Alkimos-Hermias') Μάρκου [Α]ὐρηλίου Σαραπίωνο(ς) τοῦ κ(αὶ) Ἡρακλείδου καὶ ὡς χρημ(ατίζει) … πρ̣[ὸ τ]οῦ με τυχεῖν τῆς τῶν Ῥωμαίων πολ(ιτείας) χρημ(ατίζων) Σαραπίων ὁ κ(αὶ) Ἡρακ̣λ̣ε̣ίδ(ης) Σαραπίωνος τοῦ Γαλάτ(ου) ('Marcus Aurelius Sarapion-Herakleides and as he officially P. Lond. 3 p. 144-148 no. 1179, l. 29 (TM 11764, AD 146-147?) P. Lond. 2 p. 214-215 no. 348 [1], l. 6-7 (TM 11728, AD 202203) P. Col. 10 265, l. 12-16 (TM 22266, AD 206-212) SB 18 13858, l. 1-2 and 5-8 (TM 18344, AD 211-217) acts … before I obtained Roman citizenship acting officially as SarapionHerakleides son of Sarapion grandson of Galates') Αὐρήλιο[ς -ca.?- πρὸ μὲν] τῆς θεῖας δωρ[εᾶς καλούμενος -ca.?- ] ω̣ς τοῦ Πατ̣[ -ca.?- ] ('Aurelius … who before the divine gift was called … son of …s grandson of Pat…') Αὐρήλιος Ζώσιμος πρὸς μὲν τῆς θείας δωρεᾶς καλούμενος Ζώσιμος Λεονίδου ('Aurelius Zosimos, who before the divine gift was called Zosimos son of Leonides') BGU 7 1652, l. 1-3 (TM 9532, after AD 212) BGU 2 655, l. 5-8 (TM 9293, AD 215) 12 On the fact that addressee and sender are the same person here, see U. Wilcken, ‘Aus der Straßburger Sammlung’, APF 4 (1908), 124-125. 3 Y. Broux Αὐρήλιος Αἰλ[ου]ρίων πρὶν δ[ὲ] τυχεῖν τῆς Ῥωμαίων πολιτείας Αἰλουρίων Ζωίλου ('Aurelius Ailourion, before obtaining Roman citizenship Ailourion son of Zoilos') Μάρκος Αὐρήλιος Διονύσιος ὡς δὲ πρὶν ἢ λαβεῖν τὴν πολειτείαν Διονύσιος Ἀλέξανδρου τοῦ Ἀλέξανδρου Ἀρχιστράτιος ὁ καὶ Ἀλθαιεύς ('Marcus Aurelius Dionysios (or) as before receiving citizenship Dionysios son of Alexandros, grandson of Alexandros, of the phyle of Archistratios and the deme of Althaieus') Μάρκος Αὐρή̣[λιος … πρὶν δὲ ἢ λα]βεῖν τὴν Ῥω̣[μαίων πολιτείαν …] … καὶ ἡ τού]του γυνὴ Α̣[ὐρηλία Δημητρία πρὶν δὲ ἢ] λαβε̣[ῖν τὴν Ῥωμαίων πολιτείαν Δημη]τρία Διον[υσι…] ('Marcus Aurelius … before receiving Roman citizenship … and his wife Aurelia Demetria before receiving Roman citizenship Demetria daughter of Dionysi…') Μᾶρκος Αὐρήλιος Νειλάμμω[ν … ὡς δὲ ἐχρημάτιζεν πρὶν ἢ λαβεῖν τὴν Ῥω]μαίων πολιτείαν Νειλάμμων […]ω̣νος τοῦ Νειλάμμωνος ('Marcus Aurelius Neilammon such as he officially acted before receiving Roman citizenship as Neilammon son of …on grandson of Neilammon') [… Ἡρα]κλείδου, ὡς δὲ πρὶν ἢ λαβεῖν τὴν Ῥωμαίων πολιτεί[αν ἐχρημάτιζεν …] ('… Herakleides, (or) as he officially acted before receiving Roman citizenship as …') P. Oxy. 12 1458, l. 2-7 (TM 21859, AD 216-217) JJP 40 (2010), p. 135-144 (TM 140211, AD 219-221) SB 16 12527, l. 6-7 and 9-13 (TM 16260, AD 224) PSI 5 464, l. 2-4 (TM 19304, AD 249) BGU 4 1071, l. 5-6 (TM 31015, AD 200-299) Table 1: Attestations of formulas including ῾Ρωμαίων πολιτεία Changing your name was also common in the military milieu. Roman citizenship was generally given to peregrines upon discharge from the auxilia or the legions, but there is plenty of evidence of soldiers changing their name during their service or even upon enlistment13. Apion, son of Epimachos, wrote to his father while stationed in Campania that he was from then on called Antonis Maximus (although in his salutation and in the address docket he still refers to himself as Apion)14. In an application to the ἀρχιδικαστής, Gaius Iulius Apollinarios states that he is a 'discharged soldier, but before his military service he acted as Apollonios son of Melas grandson of Ptolemaios'15. These people's official identity was altered upon obtaining Roman citizenship: instead of the typical cluster 'person – father (- grandfather – mother)' used to identify Greeks and Egyptians under Roman rule, the patronymic and other family relations were generally dropped, and a nomen was added. In daily life probably not much changed, and it can be presumed that they would still be addressed with their given name by family, relatives and friends. Officially, however, they were no longer identified by means of ancestry, as belonging to a certain family, but in legal and political terms, as a Roman with certain privileges. 13 S. WAEBENS, 'The Legal Status of Legionary Recruits in the Principate: A Case Study (Lucius Pompeius Niger, A.D. 31-64)', Y. LE BOHEC (ed.), Le métier de soldat dans le monde romain, Paris (forthcoming). 14 BGU 2 423, l. 22-23 (TM 28137, AD 100-199). 15 P. Oxy. 41 2978 (TM 30377, AD 200-299). Other examples are cited by Hagedorn in: D. HAGEDORN, 'Marci Aurelii in Ägypten nach der Constitutio Antoniniana', BASP 16 (1979), 52-53, n. 27a. 4 Y. Broux Regulations in the Gnomon of the idios logos suggest that the authorities took firm measures against acting under a false Roman identity(16). Gaius Iulius Ptolemaios and Gaius Iulius Diodoros, both sons of the Egyptian woman Tasoucharion, indeed did not escape paying the poll tax, but they were allowed to use Roman names in the official tax rolls nonetheless(17). "Greek" status In a discussion of P. Amst. 1 72, a list of people that also mentions former names of some individuals, van Minnen suggests that their name change took place after the introduction of the βουλή in the Egyptian metropoleis in AD 201(18). It is possible that wealthy Egyptians without privileged status were incorporated into the ranks of βουλευταί at this time. Most of the people who changed their name in this text sought a Greek equivalent of their Egyptian name with more or less the same meaning. As they were now members of the political elite, they required a more appropriate (i.e. Greek) name to reflect their new elevated position. The expressions used are προχρηματίζων ('previously styled as', abbreviated as προχ), διὰ προσαγγέλματος ('according to declaration') and ἀνθ' οὗ ('instead of which', see below)(19). He suspects that with προχρηματίζων the name changes always involved an Egyptian name being translated into Greek, but the text is heavily damaged, so many of his suggestions are merely hypothetical. Thus in [Ἰ]σχυρ[ί]ων [ ̣ ̣ ̣]τ̣ω̣νος μη[τρὸς …] βο̣υλ(ευτὴς) προ̣[χ(ρηματίζων)] Ν̣ε̣χθερωους and in Ἰσχυρίων Εὐδαίμονος μη[τρὸς -ca.?- ] προχ(ρηματίζων) Νεχθε̣ρω ̣ ους Πι̣[ -ca.?- ] the verb would signify that two people called Ischyrion were previously styled as Nechtheroous (ἰσχυρός and nḫt both mean 'strong')(20). In the case of Ἱέραξ Ἱέρακος … προχ(ρηματίζων) Ἱέραξ Πιβιχιος μη(τρὸς) […](21), Hierax translated his patronymic Pibichis (the Greek rendering of the Egyptian Pȝy-byk) ỉnto the Greek Hierax, both meaning 'the falcon'. With Πιασ… …ή̣σιος μη(τρὸς) Θ[…] διὰ προσ]αγγέλματος Δίδυμος Ἰσ[…] and Ἱέραξ Εὐδαίμονος μητρὸ[ς …] διὰ προσαγγέλματος Ἱ[…] the names are broken off so it is not possible to tell if they consisted of translations as well(22). The first example concerning Pias…-Didymos is perhaps another attestation of a simultaneous change of patronymic. These are not the only people in this text with alternative identifications. Unfortunately parts of each identification cluster are broken off because the right edge of the papyrus is damaged, and more attestations are probably lost. Eudaimon's abovementioned request to the idios logos is another example of Egyptian names being replaced by a Greek translation(23). In this case, the preposition ἀντὶ designates the name changes concerning Eudaimon's patro- and 16 BGU 5 1210, § 53 & 56. Ptolemaios: P. Mich. 4 223 + SB 14 11710, l. 431 & 670 (TM 11998, AD 172) and 224, l. 2024 (TM 11999, AD 173); Diodoros: P. Mich. 4 223 + SB 14 11710, l. 430 & 1887. 18 VAN MINNEN, 'A Change of Names in Roman Egypt after A.D. 202?', 87-92. 19 Van Minnen believes that the names following διὰ προσαγγέλματος did not substitute the original ones, but were added recently and were used simultaneously with the first ones (VAN MINNEN, 'A Change of Names in Roman Egypt after A.D. 202?', 87). 20 P. Amst. 1 72, l. 1-2 and 19-20 (TM 15496, ca. AD 202-212). 21 P. Amst. 1 72, l. 3-5. 22 P. Amst. 1 72, l. 14-16 and 24-25. 23 SB 18 13175, l. 15-16 (TM 18337). 17 5 Y. Broux metronymic. Eudaimon himself asks permission to use the Greek versions of his parents' names in his official identification: βούλομαι, κύριε, ἀπὸ τοῦ νῦν ἐπιτρ[απῆ]ναι χρηματίζει[ν] Εὐδαίμων Ἥρωνος ἀντὶ τοῦ Ψ[οι]τος καὶ [ἀντὶ] τῆς Τιαθρηου[ς μητ]ρὸς Διδύμης ('my lord, from now on I would like permission to officially act as Eudaimon son of Heron instead of Psois and instead of Tiathres (son) of the mother Didyme'). Heron is the Greek equivalent of the Egyptian god Shai (Ψοις = Pȝ-šy) and Didyme and Tiathres both mean 'the twin'(24). Eudaimon does not give a reason for this sudden change, but he adds 'no public or private interests being injured, but I will be benefitted'(25). Since he opts for Greek translations and speaks of advantage for himself, we can assume it had something to do with status and a wish to be perceived as belonging to the Greek community. It is remarkable that this request of AD 194 dates so briefly before the probably date of P. Amst. 1 72, all the more so since there is so little evidence for this practice from other periods. This sheds some doubts on the link between name change and the introduction of the βουλή in AD 201 as suggested by van Minnen. Slaves’ names in sales contracts In sales contracts it is often specified that slaves were known by different names, as these were often changed according to a new master's fancy. In the case of Sambatis she received the new name Athenais (table 2, first example). This is the only Egyptian example where the aorist participle passive of μετονομάζω ('to call by a new name', 'to receive a new name') is used, although it must be noted that the contract was originally written in Pamphylia. The formula is very specific and indicates that her owner deliberately renamed her(26). κοράσιον Σαμβατίδα τὴν μετονομασθεῖσαν Ἀθηναίδα ἢ εἴ τινι ὀνόματι καλεῖται, γένει Φρυγίαν ('the maiden Sambatis who is called by the new name Athenais or with whatever name she is called, of Phrygian origin') [δούλης ὀνόματι … ἐπικεκ]λημένης Φιλαδελφίου ε̣ἰ̣ καὶ τ̣ινι ἑτέρ[ῳ ὀνόματι καλεῖται ἢ κληθήσεται] ('slave with the name … who is called Philadelphias (?) or with whatever other name she is called or will be called') δούλης ὀνόματι Στεφ[άν]ης ἐπικεκλημένης Σ[τ]εφανοῦ[τος] ἢ καὶ τίν[ι] ὀνό[μα]τι κ[α]λεῖται ἢ κληθήσεται, γένει Κρητικῆς ('slave with the name Stephane who is called Stephanous or with whatever name she is called or will be called, of Cretan origin') δούλην Α[ἰλα]νοῦν ὀνόματι Καλημέραν ὡς ἐτῶν εἴκοσι οὐλόκομον [ἢ] καὶ τίνι ὀνόματι καλεῖται ἢ καὶ κληθήσεται ('slave Ailanous with the name Kalemera, 20 years old, with curly hair, or with whatever name she is called or will be called') Publication Date BGU 3 887, l. 3 & 14 (TM 20070) AD 151 Provenance* Side, Pamphylia (w); Egypt (f) SB 6 9216, AD 285 Hermopolis (w&f) l. 12 (TM 17872) P. Lips. 1 4, col.1, l. 11-12 & col. 2, l. 6 (TM 22325) AD 293 Hermopolis (w&f) SB 5 8007 (B), AD 300 Hermopolis (w&f) l. 4 (TM 32912) 350 24 U. WILCKEN, 'Aus der Straßburger Sammlung', 128-129. SB 18 13175, l. 16-17. 26 V. TCHERIKOVER, A. FUKS and M. STERN, Corpus Papyrorum Judaicarum. Volume 3, Cambridge (Mass.), 1964, 73, n.3. 25 6 Y. Broux δοῦλον αὐτοῦ ὀ[νόματι] Ἀργοῦτιν ἢ καὶ εἴ τινι ἑτέρῳ ὀνόματι καλεῖται ἢ κληθ[ήσεται] γένει Γάλλον ('his slave with the name Argoutis or also with whatever other name he is called or will be called, of Gallic (?) origin') δοῦλον αὐτῆς [ο]ἰκογενῆ Ἀψαλμαν ἢ καὶ εἴ τινι ἑτέρῳ ὀνόμ[ατι] καλεῖται ('her homebred slave other Apsalmas or with whatever name he is called') δούλην ἀργυρώνητον ὀνόματι Ἰμμεδαβου ἢ καὶ εἴ τινι ἑτέρῳ ὀνόματι καλ̣εί̣ ̣τ̣α̣ι̣ ἢ̣ κληθήσεται, γένει περιχώραν Νεσιβει ('slave bought with money with the name Immedabos or with whatever other name she is called and will be called, of origin from the Nesibis countryside') BGU 1 316, l. 11-13 (TM 20204) AD 359 SB 24 16167, l. 2-3 (TM 23926) AD 249 Markopolis, Mesopotamia (w); Syria (f) SB 24 16169, l. 10-12 (TM 23928) AD 251 Palestina (w); Arsinoites (f) Beth Phouraia, Syria (w); Syria (f) Table 2: Attestations of ἢ (καὶ εἴ) τίνι (ἑτέρῳ) ὀνόματι καλεῖται (ἢ κληθήσεται) * w = written; f = found The expression ἢ (καὶ εἴ) τίνι (ἑτέρῳ) ὀνόματι καλεῖται (ἢ κληθήσεται) (‘or also with whatever other name he is called or will be called’) on the other hand suggests indifference. The name given in the contract is just one of many possible identifiers, suggesting that slaves did not have official names. Since the choice of the name was up to the master, slave names could indeed be changed each time they were sold and were thus by definition temporary. However, slaves were included in census declarations as they were also liable to the poll tax; therefore the name mentioned here must have had some value for the administration. For four out of five of the slaves mentioned in the Egyptian documents, an additional name is already mentioned, making the expression ἢ (καὶ εἴ) τίνι (ἑτέρῳ) ὀνόματι καλεῖται (ἢ κληθήσεται) seem somewhat redundant. According to Straus it was added in contracts to protect the client, and its relative frequency from the end of the third century onward fits within the general tendency of adding more and more safeguard precautionary measures in sales contracts, such as the presence of a joint-guarantor (συμβεβαιωτής)(27). He attributes the occurrence of the expression in documents written outside Egypt to the fact that these were brought back into the country for ἀνάκρισις (examination of slaves)(28). Incidentally, in more than half of the cases the origin of the slaves is mentioned, stipulating that they were not οἰκογενής, i.e. they were not born from a slave mother (table 2)(29). Almost all documents are from Hermopolis and Syria; perhaps a Syrian slave dealer sold a batch of slaves in Hermopolis around the end of the third century AD? Unknown reasons for name change and alleged examples For many name changes, neither the expression introducing the change nor the context provides enough information on the cause. Two formulas, ἀνθ' οὗ and ὁ διὰ λόγων, appear in documents from all over Egypt, suggesting that they were official 27 J. A. STRAUS, L'achat et la vente des esclaves dans l'Égypte romaine. Contribution papyrologique à l'étude de l'esclavage dans une province orientale de l'empire romain (APF Beiheft 14), Munich – Leipzig, 2004, 133134. 28 Ibid., 134. 29 D. FEISSEL, J. GASCOU and J. TEIXIDOR, 'Documents d’archives romains inédits du moyen Euphrate (IIIe s. après J.-C.)', JournSav, 1997, 27. 7 Y. Broux terminology. P. Bouriant 42, on the other hand, provides us with a curious case, as it is the only text in which the expression ὁ δι'αἱρέσεως and the abbrevations starting with ὁ δι( ) occur. These formulas will each be discussed separately to gain some insight into their meaning and use. Ἀνθ' οὗ Ἀνθ’ οὗ is a combination of the preposition ἀντὶ with the masculine relative pronoun; it occurs 66 times, the feminine counterpart ἀνθ’ ἧς three times. The earliest example is Ὀρσευς ἀνθʼ (οὗ) Σβηβο( ) sometime after AD 14(30). Most attestations (79%) are from second century Arsinoites, although 46% of all attestations concern one person, a certain Πεκμηις ἀνθ’ οὗ Εὐρήμων, who appears 38 times as the father of Heron and Paesis in the tax rolls from Karanis(31). After eight attestations in the third century the formula dies out completely. Outside the Arsinoites it is found in Oxyrhynchos (four exx.) and Hermopolis (three exx.) in Upper Egypt, and Thmouis in Lower Egypt (three exx.). Ἀνθ’ ἧς accounts for less than 4% of the attestations of the formula, all found in tax lists from Theadelpheia(32). In 56% of the cases, the expression is abbreviated to simply ἀνθ’. The second name is never relegated to the end of the identification cluster, ἀνθ’ οὗ always directly following the first name. The formula was first recognized by Wilcken, who corrected Σατύρου Ἄνθου Σύρου in P. Grenfell 2 62, l. 2 into Σατύρου ἀνθ’ οὗ Σύρου and interpreted Syros as Satyros' shortname instead of his papponymic(33). Crönert agreed with the reading, but proposed an alternative explanation. He cites the example of Πανεφρεμμις Στοτοητεως ἀνθ’ οὗ Σαταβουτος πρεσβυτέ[ρο]υ(34), claiming that the addition πρεσβύτερος after the second name would not make sense if this was indeed a double name. Since all examples known to him were patronymics, he concluded that it was a fixed expression introducing the adoptive father's name, something like καθ’ υἱοποίαν δέ and θέσει δέ(35). This last expression was typical of second century Oxyrhynchites. Crönert saw ἀνθ’ οὗ as its Arsinoite equivalent and proposed the translation 'adoptive son of'(36). Viereck saw nothing wrong in interpreting Στοτοητις ανθ’ οὗ Σαταβους πρεσβύτερος as a double name: the father was called either Stotoetis or Satabous the elder, to make the distinction with a younger Satabous(37). Moreover, since Crönert's digression new texts have been published where ἀνθ’ οὗ clearly introduces the identified person's own names, e.g. Ἑρμίας ἀνθʼ οὗ Ἑρμῆς ὁ καὶ Εὐδαίμων(38), which refutes his theory(39). 30 P. Tebt. 2 401, l. 20 (TM 13555, after AD 14). Pnr 171216 in TM People (http://www.trismegistos.org/ref/ref_list.php?pnr=171216). Four times the scribe writes τοῦ instead of ἀνθ' οὗ between the two names (Πεκμηιτος τοῦ Εὐρήμωνος). They are definitely a scribal errors and not alternative identification clusters with the patronymic as a single name followed by the papponymic, since in one case the two names are reversed in P. Mich. 4 223 + SB 14 11710, col. 22, l. 648 (TM 11998, AD 172): Εὐρή(μονος) το(ῦ) Πεκμη(ιτος). 32 BGU 9 1896, l. 45 and 133 (TM 9467, after AD 166) and BGU 9 1898, l. 327 (TM 9469, after AD 172). 33 U. WILCKEN, 'Zu P. Grenfell I und II 62, 2', APF 3 (1903), 124. 34 BGU 2 406, col.2, l. 16 (TM 28195, ca. AD 192-200). 35 W. CRÖNERT, 'Zur Kritik der Papyrustexte', C. WESSELY (ed.), Studien zur Palaeographie und Papyruskunde IV, Leipzig, 1905, 102. 36 Ibid., 103. 37 P. VIERECK, 'Der Gebrauch von ΑΝΘ’ ΟΥ', C. WESSELY (ed.), Studien zur Palaeographie und Papyruskunde XIII, Amsterdam, 1967, 4. 38 BGU 4 1062, l. 1, 5 and 9 (TM 18507). 39 VIERECK, 'Der Gebrauch von ΑΝΘ’ ΟΥ', 5. 31 8 Y. Broux Calderini agrees that names connected by ἀνθ’ οὗ constitute a closed entity identifying the same person(40). If two people were involved it would be strange that the first is identified by his given name only while the second name is followed by a series of extra designations such as the patronymic and papponymic(41). She ascribes a different connotation to the expression, however, translating it as 'instead of', while Viereck saw it as an equivalent of ὁ καὶ(42). She also notes that it was not limited to double names where the second name is a shortname of the first one, as Wilcken first thought(43). The editors of P. Petaus also see ἀνθ’ οὗ as a formula typical of double names, referring to Preisigke and Calderini(44). But in P. Petaus 14 (line 10) the expression is used in a different context: … (τρίτον) μέρος οἰκί(ας) ἀνθ’ (οὗ) ὅλη οἰκ(ία) καινὴ … ('a third of a house, in its place a complete new house'). Here the formula implicates that a formal change took place, and the information following ἀνθ’ οὗ is newer and decisive(45). In other texts where the expression is used without names, it is also translated as 'in place of which'(46). Yet on ἀνθ’ οὗ κωμογ[ρ(αμματέως)] in P. Prag. 2 129 (line 2) the editor remarks that it should be interpreted as 'instead of what was stated by the κωμογραμματεύς', meaning that the current state of the property had changed since the last inspection by the κωμογραμματεὺς(47). This is exactly the opposite, as the information preceding ἀνθ' οὗ is newer here. The examples in P. Lond. 3 p. 92-103 no 1170 Ro. (lines 71, 316, 447, 467, 547, 566, 623, 727 and 748) and P. Col. 2 1 Ro. (4), col. 18 (lines 20 and 21) should also be understood this way(48): they are not name changes, but ἀνθ’ οὗ merely implies that the payment has been to someone in place of someone else. The same probably goes for εἰς Κολλευθιν ἀν̣θʼ (οὗ) Ἀπολλωνοῦς Θηβαίου in SB 18 13878 (line 9-10). Finally, in a list of names from Oxyrhynchos the formula seems to indicate that certain liturgists or magistrates were replaced by others, but unfortunately a large part of the papyrus is damaged and the context has been lost(49). In analogy with this use of ἀνθ’ οὗ in contexts not dealing with double names, Calderini’s connotation 'instead of' seems justified. Perhaps sometimes the expression was used to rectify a mistake made in earlier documents. Several examples where the two names bear a great resemblance certainly make this an attractive theory: Σοχωτης ἀνθ' οὗ Σουχίων(50), Ἑρμίας ἀνθ' οὗ Ἑρμῆς(51), Ἁρποκρᾶς ἀνθ' οὗ Ἁρποκ( )(52), Σάτυρος ἀνθ' οὗ Σύρος(53), etc. Another reason for the use of ἀνθ' οὗ could be that a person’s original name was recently replaced by another one and that they were not meant to be used simultaneously. But perhaps to avoid confusion or because the change had not yet 40 CALDERINI, 'Ricerche sul doppio nome personale', 242. Ibid., 243. 42 VIERECK, 'Der Gebrauch von ΑΝΘ’ ΟΥ', 4. 43 CALDERINI, 'Ricerche sul doppio nome personale', 243. 44 U. HAGEDORN et al. (eds.), Das Archiv des Petaus (Papyrologica Coloniensia 4), Cologne – Opladen, 1969, 102, n. 10. 45 Ibidem. This argument also applies to the ἀνθ’ οὗ in P. Petaus 15, l. 11, likewise concerning houses. 46 E.g. P. Oxy. 59 3979, l. 10 (TM 19265, AD 266) and P. Oxy. 64 4441, col. 4, l. 10 (TM 23666, AD 315-316). 47 R. PINTAUDI, R. DOSTÁLOVÁ and L. VIDMAN (eds.), Papyri Graecae Wessely Pragenses, Firenze, 1995, 28, n. 2. 48 TM 11761 and 10457, AD 144 and 155 respectively. 49 P. Oxy. Hels. 15, l. 3 and 4 (TM 24972, AD 1-99). I have excluded the examples cited in this paragraph from the statistics, since they do not concern name changes. 50 P. Petaus 15, l. 9 (TM 8754, AD 194-185). 51 BGU 4 1062 Ro., l. 1, 5 & 9 (TM 18507, AD 236). 52 BGU 9 1896, col. 1, l. 5 (TM 9467, AD 166). 53 P. Grenf. 2 62, l. 2 (TM 11364, AD 211). 41 9 Y. Broux been officially ratified this construction was (temporarily?) applied. There may then be a link with the 14-year census cycle: between two censuses, the old name as expressed in the previous declaration was still added in official documents. Almost all attestations are found in documents related to taxes, and as seen, they are limited to the first three centuries of Roman rule, which accords with the time span of the census declarations and the attestations for the λαογραφία, the poll tax(54). A similar chronological pattern is found with the expressions καὶ ὡς χρηματίζει(55), ἀπάτωρ and χρηματίζων μητρὸς(56), also concentrated in the second and third centuries AD. The immense administration revolving around taxes and liturgies needed detailed information, especially in regard to (partial) exemptions. The poll tax was still levied even after the Constitutio Antoniniana of AD 212, of which Roman citizens before this date were exempt, but it disappeared around the middle of the third century AD. After the introduction of a new taxation system by Diocletian a couple of decades later, no more trace is found of these formulas. Since ἀνθ' οὗ literally reads 'instead of which', the name preceding it should be the old reading, and the second name the new one. This does not accord well with the examples where the second name is abbreviated though. One would expect an official correction or new name to be written in full to avoid future mistakes or confusion, Moreover, considering the official nature of the documents, the correct official name logically comes first. As regards content, therefore, the more obvious interpretation seems to be the opposite: 'new name vs. old name'. In any case, I do not think ἀνθ' οὗ should be put on par with formulas to connect double names, such as ὁ καὶ or ἐπικαλούμενος, since these imply that both names were in use at the same time. Ἀνθ' οὗ on the other hand either introduces a person's old name after he took on a new one, or it was a correction of a scribal error committed in an earlier document, and therefore has more of an adversative meaning in the sense of 'instead of'. Ὁ διὰ λόγων Ὁ διὰ λόγων is used eleven times for men only, always in the second century AD Arsinoites(57). In five cases the context is too damaged to provide any useful information about its meaning. In three texts the formula is undeclined, ὁ being used while the names are written in the genitive and the accusative respectively: Νεσε[υν] Νεσευτος τοῦ Ἡρακλᾶ ὁ διὰ λόγων πάππου Νεσευτος μη(τρὸς) Σαραποῦτος, [Ἀπυγ]χεως ὁ διὰ λόγων […] and Ἁρπαλον ὁ διὰ λόγων Νικηφόρον(58). In P. Meyer 9, ὁ διὰ λόγων is interpreted as 'nach Ausweis der (im Hause vom Besitzer geführten) Listen'. The text is a census declaration listing the tenant of a house together with his resident family members and other inhabitants. Ὁ διὰ λόγων introduces the double names of the two sons of one of his slaves: Πασίων ὁ διὰ λόγων Εὐτύχης ('Pasion, nach Ausweis der (Haus-)Liste den Namen Eutyches führend') and the abovementioned Ἁρπαλον ὁ διὰ λόγων Νικηφόρον ('Harpalos, 54 R. S. BAGNALL and B. W. FRIER, The Demography of Roman Egypt (Cambridge Studies in Population, Economy and Society in Past Time 23), Cambridge, 1994, 9-11. 55 BROUX, COUSSEMENT and DEPAUW, 'καὶ ὡς χρηματίζει and the Importance of Naming in Roman Egypt', 160. 56 M. MALOUTA, 'Fatherlessness and Formal Identification', Sabine R. HÜBNER and D. M. RATZAN (eds.), Growing up Fatherless in Antiquity, Cambridge, 2009, 132-135. 57 In P. Tebt. 2 609 Vo descr. (TM 28482, AD 100-199) λόγου is written instead of λόγων. 58 P. Petaus 10, l. 11 (TM 8803, AD 184); Stud. Pal. 22 96, l. 11 (TM 27646, AD 100-199) and P. Meyer 9, l. 11 (TM 11963, AD 147) respectively. 10 Y. Broux nach Ausweis der (Haus-)Listen den Namen Nikephoros führend')(59). Though the standard formulas ὁ καὶ and ἐπικαλούμενος were also used for slaves' double names, according to Meyer they merely introduce a so-called 'Rufname' or pet name by which they were called in the household(60). By using ὁ διὰ λόγων on the other hand, the second name was not only used in the domestic sphere, but was also officialy recorded in the household declaration. Meyer's explanation seems somewhat forced, however. He points to BGU 2 493, a list of households where the formula is (supposedly) also used for slaves. In this text formula is followed by τῆς μητρὸς […] though, which is not exactly a double name. Moreover, other examples of ὁ διὰ λόγων do not apply to slaves. Nor does Meyer explain why, if only slaves' double names introduced by ὁ διὰ λόγων were officially recognized in the census declaration, another slave girl is enlisted with her unofficial pet name as Ἰ[σ]ιδώραν τὴν καὶ̣ Ἡδίστην in the same document(61). Calderini only briefly refers to these two examples and Meyer's interpretation of them(62). In the edition of the archive of Petaus, Meyer's hypothesis is discarded for the identification cluster Νεσευν χρηματίζειν Νεσε[υν] Νεσευτος τοῦ Ἡρακλᾶ ὁ διὰ λόγων πάππου Νεσευτος μη(τρὸς) Σαραποῦτος, but the translation offered remains more or less the same: 'nach Ausweis der Akten (heißen)'(63). But where Meyer sees the first name as the official name and the name following ὁ διὰ λόγων as a byname that is officially put down in writing, the editors of P. Petaus prefer the opposite interpretation: Heraklas is merely used by his grandson Neseus when defining himself in official documents, but the grandfather's official name as stated in official documents was simply Neseus: 'Neseus, officially registered as Neseus son of Neseus and Sarapous, grandson of Heraklas, the grandfather who according to identification in official documents is called Neseus'. In any case, it is generally accepted that ὁ διὰ λόγων refers to a person’s official identification, a notion that is not explicitly expressed by the previously discussed formulas. All preserved examples are indeed found in official documents: census declarations(64), official correspondence(65), a list of canal workers(66), etc. This contradicts the interpretation given in P. Petaus. In official documentation one would expect an individual's official identification to come first. Especially in the examples consisting only of a person's name, without patronymic or other additional information, it surprises that the byname is given before the official name. The use of the verb χρηματίζω in P. Petaus 10 also implies official identification(67), so the subsequent identification cluster should include the official names by which Neseus is recorded in the administration. I am not inclined to interpret ὁ διὰ λόγων as a formula introducing double names. I agree with the connotation put forward by both Meyer and the editors of Petaus, that λόγων refers to official documentation, but I think the expression was employed to indicate that a new official name (the first one mentioned) had 59 P. Meyer 9, l. 11 (TM 11963, AD 147). P. M. MEYER, Griechische Texte aus Ägypten, Berlin, 1916, 59-60, note 35. 61 P. Meyer 9, l. 13. 62 CALDERINI, 'Ricerche sul doppio nome personale', 243. 63 HAGEDORN et al. (eds.), Das Archiv des Petaus, 90. 64 P. Meyer 9; BGU 2 493 (TM 28155, AD 100-199). 65 P. Petaus 10; SB 12 11082 (TM 14404, AD 138-160). 66 BGU 13 2263 (TM 29222, AD 125-175). 67 BROUX, COUSSEMENT and DEPAUW, 'καὶ ὡς χρηματίζει and the Importance of Naming in Roman Egypt', 163-164. 60 11 Y. Broux replaced the old official name (the second one) that had previously been used in official documents. Πασίων ὁ διὰ λόγων Εὐτύχης would thus be 'Pasion, who according to official documentation (was previously styled as) Eutyches' (slaves' names were often changed by their owners, see above under ἢ (καὶ εἴ) τίνι (ἑτέρῳ) ὀνόματι καλεῖται (ἢ κληθήσεται)). Concerning the Neseus mentioned above, his identification cluster could be interpreted in two ways, depending on whether ὁ διὰ λόγων is attributed to him, or to his grandfather. In the former case his identification should be translated as 'Neseus, officially registered as Neseus son of Neseus and Sarapous, grandson of Heraklas, who according to official documentation (was previously styled as the one) of the grandfather Neseus'. If ὁ διὰ λόγων refers to his grandfather Heraklas, however, the construction should be interpreted as 'Neseus, officially registered as Neseus son of Neseus and Sarapous, grandson of Herakles, the grandfather who according to official documentation (was previously styled as) Neseus'. The grandfather's original name was Neseus, but either he himself, his son or his grandson had changed it to Heraklas. It is added in this description to facilitate identification, as in older records he would not be found under the name Heraklas. In BGU 2 493 then, Εὐτυχὴς δοῦλ(ος) αὐτοῦ ὁ διὰ λόγων τῆς μητρὸς [ -ca.?- ] can be interpreted as 'Eutyches his slave, who in official documentation (was styled as) of the mother …'. In two texts, the abbreviation οδιαλ is supplemented as ὁ διὰ λ(όγων)(68). The household declaration BGU 2 495 (line 2) reads [ ̣ ̣ ̣]δώρας Ἰσχυρίωνος ὁ διὰ λ(όγων) δοῦλ(ος) […](69). The formula is not followed by a name, but in analogy to the previous example of Eutyches, this identification cluster can be translated as '…doras son of Ischyrion, who in official documentation (was styled as) the slave [of …]'. Since the text is severely damaged, the purpose of this construction remains unclear though. In a tax collector’s synopsis from AD 179 on the other hand the editor reads Pakysis' complete identification cluster as Πακυσ̣[ις Σαταβου]τ̣ος τοῦ Στοτοητεως μητ(ρὸς) Ταουητιος ὁ διὰ λ(όγων) [Πα]νεφρεμμεως(70). If ὁ διὰ λόγων refers to the subject, Pakysis, the subsequent gentive is puzzling. One could consider reading [Τα]νεφρεμμεως, but his would then be the only feminine attestation of the expression, moreover construed with the masculine ὁ. Either way the cluster is grammatically inconsistent. A last possibility is that ὁ διὰ λόγων refers to the father or grandfather, although one would then expect it to be incorporated immediately after the first name in question. Alleged examples: the enigmatic ὁ δι and ὁ δι' αἱρέσεως In her survey of the formulas used to introduce double names, Calderini mentions an expression abbreviated in various ways but always starting with ὁ δι. Because of the close resemblance to ὁ διὰ λόγων, she classifies the examples as belonging to this formula(71). She remarks that they are rather odd, but does not examine them any further. These abbreviations do not always seem to have the same purpose though. The majority (11 exx.) is found in one and the same text, namely P. Bouriant 68 These two examples should not be put on par with the abbreviation ὁ διαλ in P. Bouriant 42 (see below), a register of tenants of public land. They occur in a totally different context, namely documents concerning population control. 69 (TM 28157, AD 100-199). 70 SB 16 12816, l. 5-6 (TM 14676). 71 CALDERINI, ‘Ricerche sul doppio nome personale', 243. 12 Y. Broux 42 (AD 166-167, Hiera Nesos). This document consists of two parts, a land register (l. 1 -285) and a register of inhabitants of Ptolemais Nea and Karanis subject to land tax (l. 286-710). It is in this second part that the abbreviations occur. The extent of abbreviating differs widely: from ὁ δι over ὁ δια, ὁ διαλα, ὁ διλ ως, to ὁ κ διαι ως. The editor of P. Bouriant 42 interprets them differently though, as abbreviations of ὁ δι' αἱρέσεως, an expression again limited to this text only. The following table gives an overview of all the abbreviations and ὁ δι'αἱρέσεως attested in P. Bouriant 42. In many cases, a relationship between the two parties is probable; if so this has been indicated as such: Formula ὁ δι ὁ διλ ὁ διαλα ὁ δι̣ ως ὁ διλ ως ὁ κ δι ως ὁ κ δια ως ὁ κ διαι̣ ως ὁ δε διαι ὁ διαρο (?) Identification cluster [Χ]α[ι]ρή(μων) ὁ δι( ) Μυσθαρίω(ν) Ἥρων Πτολεμ(αίου) ὁ δι( ) Ἰσχυρίωνος Πεθεως ὁ δι( ) Πτολεμαῖ(ος) Μάρωνος Δῖ[ο]ς ὁ κ(αὶ) Σαβῖνο(ς) Ὡριτ(ος) ἀπ(ὸ μητροπόλεως) ὁ δι( ) Δῖος Ὡριτ̣ο̣ς Ἀφροδ(ίσιος) Ἡρακλείδ(ου) ὁ διλ( ) Ἀφροδί(σιος) Δημητρίο(υ) Κάστωρ μη(τρὸς) Ταπασμουτ(εως) [ὁ] διλ( ) Κάστωρ Ἰσίων(ος) υ(ἱ)ο(ῦ) Διδᾶ ἀ(πὸ μητροπόλεως) Πισκᾶς Ἰσχυρ(ί)ω(νος) ὁ διλ( ) Ἰσχυρίω(ν) ἀ(πὸ μητροπόλεως) ἐπικ(αλούμενος) Ἀνουβ Ἰσχυρίω(ν) (ὁμοίως) ἐπικ(αλούμενος) Ἀπώνιος ὁ διλ( ) Πασίω(ν) νεώ(τερος) Ἰσχυρίω(νος) ἐπικ(αλούμενος) Ἀσβο Χαιρήμων ὁ διαλα( ) Μάξιμος Πασίωνος τοῦ Χαιρήμ(ονος) ὁμ(οίως) … ὁ δι̣( ) ως κ… Ἰσχυρίω(ν) Πετεσουχ(ου) ὁ διλ( ) ως Μελανᾶς Δημητ(ρίου) ἀπὸ (μητροπόλεως) Ἀυνης Δημᾶ πρεσ(βυτέρου) Θέωνο(ς) ὁμ(οίως) ὁ διλ( ) ως Ἀυνης Δημᾶ Πανοντ(ως) Μυ(σ)θ(αρίωνος) … [ὁ δ]ιλ( ) ως Πανοντ[ως] Δημᾶς πρεσβ(ύτερος) Θέωνο(ς) ἀ(πὸ μητροπόλεως) ὁ κ( ) δι( ) ως Δημητρίο(υ) Μαμερτῖνο(ς) Κεφαλᾶ ὁ κ( ) δια( ) ως Μαμερτῖνο(ς) Ἀντιόχ(ου) Πα[σ]ίων Πλαντ(ᾶ) τ̣ο̣(ῦ̣) Ἀρείο(υ) ἀ(πὸ μητροπόλεως) ὁ κ( ) διαι̣( ) ως Πασίο(νος) Πλαντ(ᾶ) Ἰσχυρίω(ν) ὁ κ(αὶ) Πασίω(ν) Πετεσουχ(ου) υ(ἱ)ο(ῦ) Θέ[ω]νος ἀ(πὸ μητροπόλεως) ὁ κ( ) διαι( ) ως Πασίων Πετεσουχ(ου) Ἀχιλλᾶς ἀπάτ(ωρ) ὁ δε διαι( ) Χαιρήμων ἀπάτωρ μη(τρὸς) Οὐαλερί(ας) Πασίων Ἰσχυ(ρίωνος) ὀ διαρο̣( ) Παησις Πασίωνο(ς) Relationship parties same person? brothers? Line 386 498 525 541542 521 564 father & son? brothers? father & son? same person same person? same person? 664665 574575 459460 613 475476 506507 646647 598599 602603 588590 same person? brothers? 592593 brothers? 555556 482483 father & son? 13 Y. Broux ὁ διαιρε ως ὁ διαιρέσεω ὁ δι διαιρέσεως Λούκιος Ἀνθέστιος Πασίωνο(ς) ὁ διαιρε( ) father & son? ως Πασίων Λουκίου Ἰσχυρίων Διοσκ(όρου) ὁ κ( ) διαιρέσεω(ς) father & son? Ἰσχυρίω(ν) (ὁμοίως) ἐπικ(αλούμενος) Κέων(ος) Μά[ξιμος] Π[α]σίωνο(ς) ὁ δι( ) διαιρέσεως brothers? Χαιρ[ᾶς Πασ]ίωνος 452453 494495 462463 Table 3: Attestations of ὁ διαιρέσεως and the various abbreviations in P. Bouriant 42 The formula ὁ δι' αἱρέσεως is found exclusively in P. Bouriant 42 (166-167), a document consisting of a land register (l. 1-285) and a register of people from Ptolemais Nea subject to land tax (l. 286ff). A number of variations and abbreviations for this formula also exist: ὁ δε διαι, ὁ διαρε, ὁ διαιρε ως and ὁ κ διαι ως. The last one is used twice for what Collart presumes to be ὁ κ(αὶ) δι’ αἱρέσεως. Examples include Πασίων Ἰσχυ(ρίωνος) ὁ διαρε Παῆσις Πασίωνο(ς)(72) and Ἰσχυρίω(ν) ὁ κ(αὶ) Πασίω(ν) Πετεσούχ(ου) υ(ἱ)ο(ῦ) Θέ[ω]νος ἀ(πὸ μητροπόλεως) ὁ κ διαι ως Πασίων Πετεσούχ(ου)(73). Ὁ δι' αἱρέσεως is translated by the editor Collart as ‘who by choice (is also known as)’, indicating that the person himself decided to alter his identification cluster one way or another. By being entered in the documents this change was then officially recognized(74). Calderini too recognizes that this formula does not exactly deal with double names but rather 'provides all elements required for the exact identification of an individual in official documents, for example regarding taxes, such as P. Bouriant 42'(75). But while Collart assumes that a person’s two identifications could exist side by side, Calderini poses that a new name was chosen to replace the old one. The change could concern the person himself or his patronymic, or both. Her interpretation of ὁ δι’ αἱρέσεως is consequently: ‘who has chosen (to be called)’, a subtle yet crucial turn from the abovementioned ‘who by choice (is also known as)’, a translation that emphasizes Collart's theory of coexistence. Calderini even hints at adoption for a possible explanation, but admits that this is speculative(76). The tax list of P. Bouriant 42 contains some other peculiar formulations, more precisely a whole range of abbreviations (see table 3). Some could indeed be short for ὁ διὰ λόγων, as Calderini suggested, but not all seem to have the same purpose. Sometimes an actual double name does seem intended, such as Πεθεως ὁ δι( ) Πτολεμαῖ(ος) Μάρωνος (l. 525) and Χαιρή(μων) ὁ δι( ) Μυσθαρίω(ν) (l. 386). Others seem to introduce an different patronymic, as in Ἀφροδ(ίσιος) Ἡρακλείδ(ου), ὁ δια( ) Ἀφροδί(σιος) Δημητρίο(υ) (l. 521)(77); in the case of Kastor an entirely new identification cluster is even fabricated: Κάστωρ μη(τρὸς) Ταπασμουτ(εως) [ὁ] διλ( ) Κάστωρ Ἰσίων(ος) υ(ἱ)ο(ῦ) Διδᾶ (l. 564). Ἰσχυρίω(ν) Πετεσουχ(ου) ὁ διλ( ) ως Μελανᾶς Δημητ(ρίου) (l. 475-476) seems to be a combination of both. Alternatively the second leg could be a shortened version of the first identification cluster: Ἀυνῆς Δημᾶ πρεσ(βυτέρου) Θέωνο(ς) ὁμ(οίως), ὁ διλ( ) ως Ἀυνης Δημᾶ (l. 506) and 72 ‘Pasion son of Ischyrion who by choice (is known as) Paesis son of Pasion’; P. Bouriant 42, l. 482-483. ‘Ischyrion alias Pasion son of Petesouchos who was the son of Theon, from the metropolis, who also by choice (is known as) Pasion son of Petesouchos’; P. Bouriant 42, l. 592-593. 74 P. COLLART, Les papyrus Bouriant, Paris, 1926, 169. 75 CALDERINI, ‘Ricerche sul doppio nome personale', 244. 76 Ibid., 224-225. 77 The reading διλ is also possible here. 73 14 Y. Broux Πα[σ]ίων Πλαντ(ᾶ) Ἀρείο(υ) ἀ(πὸ μητροπόλεως) ὁ κ( ) διαι( ) ως Πασίο(νος) Πλαντ(ᾶ) (l. 588-590). In this context Δῖ[ο]ς ὁ κ(αὶ) Σαβῖνο(ς) Ὡριτ(ος) ἀπ(ὸ μητροπόλεως) ὁ {Δι} Δῖος Ὡριτ̣ο̣ς (l. 542) should be mentioned. The editor of the papyrus supplemented ὁ (καὶ) before the last leg of this string. This would then be the only example of ὁ καὶ introducing an alternative identification cluster rather than an actual double name. The following 'δι' he then interpreted as a scribal error, assuming it belonged to the subsequent name Δῖος, but in light of the other abbreviations the construction should obviously be read as Δῖ[ο]ς ὁ κ(αὶ) Σαβῖνο(ς) Ὡριτ(ος) ἀπ(ὸ μητροπόλεως) ὁ δι( ) Δῖος Ὡριτ̣ο̣ς. Collart has proposed that ὁ διλ and ὁ διαλ might be abbreviations of a form of the verb διαλλάττω, ‘to (ex)change’, ‘to alter’(78). If, analogous to ἐπικαλούμενος and ἐπικεκλημένος, it is supplemented by the present participle passive διαλλάττομενος, it could be rendered as ‘A who has changed (his identification) to B’. Ὁ δι’ αἱρέσεως also occurs once in combination with the abbreviation οδι: Μάξ[ιμος] Π[α]σίωνο(ς) οδι διʼ αἱρέσεως Χαιρ[ᾶς Πασ]ίωνο(ς)(79). Collart seems to imply that this is actually the complete formula. A person would then indicate that he himself changed his personal identification (the οδι part) by choosing (the δι’ αἱρέσεως) a different name, patronymic, etc(80). Collart thus interprets the construction as an expression of double civic 81 status( ), but this seems to be pushing the matter somewhat too far. In Roman Egypt the different classes were clearly defined and involved certain rights and obligations. ‘Upgrades’ were difficult to obtain, especially since it was discouraged for Romans and Alexandrians to mix with the lower classes. This rigid hierarchy did not leave room to accumulate different statuses. The editor himself stresses that διαλλάττω is just a suggestion though, and that other solutions to the abbreviations are equally possible(82). Calderini is followed by Hobson in her interpretation of the abbreviations in P. Bouriant 42 as coming from ὁ διὰ λόγων(83). If this is the case then the aspect of change is lost and the phrase merely expresses the difference between that person’s identification as it is given in official documents. Calderini herself mentions some abnormalities, but does not discuss them in detail. Indeed, for some of the examples the interpretations offered by Collart and Calderini seem implausible. In the case of Λούκιος Ἀνθέστιος Πασίωνο(ς) ὁ διαιρε ως Πασίων Λουκίου(84), for instance, this would mean that the son changed his name into that of his father and vice versa. Moreover, the abbreviations in P. Bouriant 42, although adopted in a list of names concerning taxes, are employed in a completely different context than ὁ διὰ λόγων. Whereas the latter is applied in official documentation concerned with population control and taxes, the former is used in a text regarding land tenure, so I am not convinced that they are used for the same purpose. When taking the subject of land tenure into account though, another solution easily presents itself. For many of the lands around Ptolemais no buyer or 78 COLLART, Les papyrus Bouriant, 169. P. Bouriant 42, l. 462-463. 80 Which element exactly is made clear in the second identification cluster. 81 COLLART, Les papyrus Bouriant, 169. 82 Ibid., 170. 83 D. W. HOBSON, ‘PVINDOB. GR. 24951 + 24556: New Evidence for Tax-Exempt Status in Roman Egypt’, Atti del XVII congresso internazionale di papyrologia, Napels, 1984, 861, note 6. 84 P. Bouriant 42, l. 452. 79 15 Y. Broux lessee was found, so the state resorted to other means and deported farmers from Karanis to cultivate the remaining land. But before this last resort, some of the land was divided by lot for lease among the klerouchoi of Ptolemais Nea, a process designated as διαίρεσις, applied already in the Ptolemaic period(85). The resemblance with our "formula" is striking, not to say they are practically identical. It is odd that this did not attract Collart's attention, since he briefly discusses this procedure only a few pages after he explains his interpretation of official name change(86). This is too much of a coincidence to ignore. The reading as ὁ δι' αἱρέσεως should therefore be corrected into ὁ διαιρέσεως and the κ in the abbreviation ὁ κ διαι ως then does not stand for κ(αὶ) but for κ(ατα) in καταδιαίρεσις(87). What the expression means is not exactly clear though. Tenure in the form of diairesis was accorded indefinitely, but a new diairesis could be imposed at any time, at which the current farmer lost the plot and it was accorded to someone else(88). So in P. Bouriant 42 the expression could introduce the new tenant as opposed to the old one. But parcels could also be attributed to γεωγροὶ operating together, as is clear from P. Tebt. 2 376 (TM 13532, AD 162), where Cheus and Heraklas jointly hold a plot, and the former subleases his part to Saras. Family members could also constitute such an association(89). In light of this the abovementioned example Λούκιος Ἀνθέστιος Πασίωνο(ς) ὁ διαιρέ(σε)ως Πασίων Λουκίου makes sense: if Pasion, father of Lucius Anthestius is the same person as the Pasion son of Lucius in the second part, it could indicate that the parcel of land was cultivated by both father and son. With this new interpretation of ὁ διαιρέσεως in mind, perhaps a solution for the abbreviations should be sought in that direction. For ὁ δι( ) and ὁ δια( ), ὁ διαδεχόμενος ('the successor') is possible, indicating that the first person is succeeded by the second as cultivator of a plot. For ὁ διαλα( ) a verb like διαλαγχάνω comes to mind, signifying a transfer or a division by lot. A similar interpretation would also make better sense for examples such as Ἰσχυρίω(ν) (ὁμοίως) ἐπικ(αλούμενος) Ἀπώνιος ὁ διλ( ) Πασίω(ν) νεώ(τερος) Ἰσχυρίω(νος) ἐπικ(αλούμενος) Ἀσβο (l. 574-575), where ἐπικαλούμενος is already used to express a second name. In Collart's interpretation, using the verb διαλλάττω, Ischyrion son of Ischyrion would be equated with with Pasion aka Asbo son of Ischyrion, assigning them both three names! Moreover, his supplement as οδι<α>λ( ) is unconvincing, as abbreviations within words are highly uncommon. An appropriate supplement with a meaning similar to διαλαγχάνω could indicate that Ischyrion aka Asbo succeeded Ischyrion aka Aponios as tenant (perhaps they were brothers?), and therefore also the accompanying taxes due. Or a piece of land was appointed to both of them and they now had to share the tax burden accordingly(90). 85 G. POETHKE, Epimerismos: Betrachtungen zur Zwangspacht in Ägypten während der Prinzipatszeit, Brussels, 1969, 31-32 & 50. 86 COLLART, Les papyrus Bouriant, 169-170 (names) and 174 (lease of land). 87 Καταδιαίρεσις is used in P. Tebt. 2 376, l. 27 (TM 13532, AD 162) for example. 88 P. Tebt. 2 376 (TM 13532, AD 162) where a plot is held μέχρι τῆς ἐσομένης κοινῆς γεωργῶν διαμισθώσεως ('until the coming joint leasing out among the cultivators'). M. ROSTOVTZEFF, Studien zur Geschichte des römischen Kolonates (APF Beiheft 1), Leipzig – Berlin, 1910, 164. 89 Ibidem. 90 In other parts of the text people from Karanis are appointed excess public or state land in the area of Ptolemais. If the state did not find sufficient lessees from the town whose territory the land belonged to, the burden was imposed on farmers from other areas. See A. C. JOHNSON, 'The ἐπιβολή of land in Roman Egypt', Aegyptus, 32 (1952), 61-72. 16 Y. Broux There is one problem with the abovementioned theory, however. The 'ως' in the abbreviations are separated from what precedes by interspacing; once it is even reverted to the next line(91). Therefore it can hardly be interpreted as the genitive ending of (κατα)διαίρεσις. Ὡς as a conjunction, preceded by a form of the verb (κατα)διαιρέω provides a similar conundrum, creating an akward composition with a conjuntion followed by a personal name. Moreover, other examples, such as the abovementioned Ἰσχυρίω(ν) ὁ κ(αὶ) Πασίω(ν) Πετεσουχ(ου) υ(ἱ)ο(ῦ) Θέ[ω]νος ἀ(πὸ μητροπόλεως) ὁ κ( )διαι( ) ως Πασίων Πετεσουχ(ου) or Μαμερτῖνο(ς) Κεφαλᾶ ὁ κ( ) διαι( ) ως Μαμερτῖνο(ς) Ἀντιόχ(ου), still raise questions. The repetition of names obscures the relation between the two parties. Concerning Ischyrion-Pasion son of Petesouchos and Pasion son of Petesouchos one could argue that they were brothers, but the name Mamertinos in the second example is not that common in Egypt so as to readily assume that this example also involves two individuals. In any case it can be concluded that ὁ διαιρέσεως and the abbreviations do not introduce a name change. An interpretation of the expressions as indicators of joint lease or transfers of land seems the most probable, perhaps as part of the proces of διαίρεσις. In light of some of the examples where the two parties apparently refer to one and the same person (e.g. Ἀυνης Δημᾶ πρεσ(βυτέρου) Θέωνο(ς) ὁμ(οίως) ὁ διλ( ) ως Ἀυνης Δημᾶ) I consider the expression as introducing a transfer of the lease, as an individual cannot enter into a joint lease with himself. What is possible though, is that at the time of a new διαίρεσις, a plot of land is allocated to the same farmer as before, explaining why he is entered twice, both before and after the expression. The concise nature of the text makes it rather difficult to interpret, but given the context of land tenure, an interpretation of ὁ διαιρέσεως and the abbreviations in that direction makes more sense than the complicated explanations of double civic status by Collart or Calderini's change of names. Hopefully this will help to a better understanding of the process of διαίρεσις and land tenure in Roman Egypt in general. 91 Lines 602-603: Μαμερτῖνο(ς) Κεφαλᾶ ὁ κ( ) διαι( ) / ως Μαμερτῖνο(ς) Ἀντιόχ(ου). 17