Documenting Roman Citizenship

6 Documenting Roman Citizenship Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 Anna Dolganov O n a fateful day in the reign of Nero, violence broke out in Jerusalem outside the Jewish temple. As soon as the Roman garrison reached the scene, those who were busy beating one particular man halted their blows in midair. The Roman soldiers seized the man and marched him to their military head- quarters, followed by an angry crowd clamoring for the man’s arrest. The man spoke to the Roman commanding officer (tribunus) in Greek, identifying him- self as “Paulus, a Jew of Tarsus in Cilicia, a citizen of no insignificant city.” Without dwelling on this statement, the tribunus ordered Paulus to be beaten and tortured to investigate the cause of the violence. As he was being tied up, Paulus questioned whether it was permissible to beat a Roman citizen. The soldiers notified the tribunus, who confronted Paulus:  “Tell me, are you a Roman?” “Yes!” Impressed, the tribunus informed Paulus that he had obtained his own Roman citizenship for a large sum of money, to which Paulus replied that he was born a Roman citizen. By virtue of this status, Paulus was transferred to the court of the governor in Caesarea with a military escort to protect him. Insisting on his right to be tried by a Roman court, he ultimately appealed to the emperor and was sent off to Rome. “To Caesar you have appealed,” so pro- nounced the governor, “to Caesar you shall go.”1 This well-​known story, told at some point near the turn of the second cen- tury, elicits several basic questions about Roman citizenship as an exclusive and privileged status in the Roman Empire before the Constitutio Antoniniana of 212 CE. What are the implications of a claim of Roman status being made with con- fidence by a Hellenized Jew traveling far beyond his province of origin and being accepted, apparently without question, by Roman provincial authorities? When Roman citizenship was asserted by individuals throughout the empire, what could imperial officials do to verify these assertions? To what extent was status claimed on the basis of official documents? To what extent were administrators able to detect erroneous or fraudulent status claims? Investigations into these questions have tended to be skeptical about the role of documentation and the ability of the Roman state to verify the civic 1. See Acts 21:30–​25:12. On Paul’s Roman citizenship, see Acts 16:37–​38 (in Philippi) and 22:25–​28 (in Jerusalem), with Sherwin-​White 1963, 144–​171. Anna Dolganov, Documenting Roman Citizenship In: Roman and Local Citizenship in the Long Second Century CE. Edited by: Myles Lavan and Clifford Ando, Oxford University Press. © Oxford University Press 2021. DOI: 10.1093/​oso/​9780197573884.003.0007 186  Anna Dolganov and legal status of individuals. As one study concludes, “the procedures were there for punishing infringements, once detected. But the possibility of fraud, which an insistence on documentary evidence from some sort of official source would have greatly reduced, seems virtually unchecked.”2 This view of Roman state record keeping continues to be widely accepted. A recent monograph on Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 Roman technologies of information argues that Roman archives “could con- tain considerable quantities of material, but rarely collected systematically and not organized well for later use,” and concludes that calling these technologies primitive is “not an entirely unfair description.”3 There is broad agreement that Roman authorities did not require individuals to present official records of their status (such as birth declarations) or regard such records as a privileged form of evidence.4 Instead, it is argued that “the use of documentary proof of status was (always excepting Egypt) only patchy and not a legal requirement.”5 Why, exactly, Egypt should be treated as an exception is far from obvious. In this uniquely well-​documented Roman province, we find abundant evidence for the Roman state’s elaborate procedures of registering the population and policing membership in privileged status groups. Papyri also show individuals being regularly asked to prove their status on the basis of official records. In this and other respects, Egypt appears to be exceptional mainly in its wealth of documentary evidence illustrating Roman administrative practice, with fun- damental implications for our understanding of the institutional framework of the empire.6 This chapter will analyze the institutions and procedures by which the Roman imperial state documented and kept track of Roman citizens. I  will begin with an overview of the census—​ the core demographic apparatus through which the Roman state exacted its fiscal demands and exercised over- sight of fiscally privileged groups. I will then examine a series of institutional contexts in which administrators registered and verified the status of Roman citizens. 2 . Gardner 1986a, 11–​12; see also Reinhold 1971 and Brunt 1971, 42–​43, 170–​171, 208–​209. 3. Riggsby 2019, 203. 209. This study draws its conclusions from a limited sample of evidence comprising several dozen citations of predominantly Latin authors and only thirty-​six Latin inscriptions, five Greek inscriptions, and eight passages from legal texts. Of hundreds of relevant papyri, only eight are mentioned. This omission is regrettable, since the documentary evidence speaks resoundingly against the conclusions reached by the author. 4. A generally accepted thesis; see Schulz 1943, 63–​64; Gardner 1986a, 4–​5; Parkin 2003, 178–​182; Sánchez-​ Moreno Ellart 2001 and 2010, 119, “the Roman proof system did not recognize the value of privileged evidence.” 5. Gardner 1986a, 12. The exceptionalism of Egypt continues to shape historical scholarship on the Roman Empire; accordingly, the most recent comprehensive treatment of the Roman imperial census devotes a spe- cial section to “la spécifité égyptienne” (Le Teuff 2012, 325–​339). 6. In this respect, I build on the important work of Rudolf Haensch on Egypt as a rich source of evidence for Roman imperial institutions and administration; see in particular Haensch 1992 and 1994. Documenting Roman Citizenship  187 Population Records in the Roman Empire The Census in Republican Italy The Roman Republic had a strong demographic tradition. Every five years, a pop- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 ulation census was conducted in the city of Rome, a procedure that culminated in a ritual of civic purification (lustrum) and produced an official number of Roman citizens (capita civium). The census was the basis of taxation, military recruit- ment, and the division of the population into property classes and voting tribes.7 Supervision of the census, conducted by two censors, was the most prestigious and senior magistracy in a Roman public career. By virtue of their prerogative of moral policing (regimen morum), the censors could downgrade the status of individuals, cancel their voting rights, and expel them from their political class, including the Senate and the equestrian order.8 In the traditional procedures of the census, adult male citizens who were legally independent (sui iuris) presented themselves before the censors and declared under oath all members of their household, including wives, children, freedmen, and slaves, specifying their name, age, filiation, and tribe and the value of their property.9 It can be deduced that Roman widows and orphans were registered as well.10 It is an essential point that the operations of the republican census were not limited to the adult male citizens who made up the official lists of capita civium but were a demographic enterprise that was comprehensive in nature. Accordingly, the censors collected information about families, slaves, and freedmen, as well as non-​citizens and Latini residing at Rome and in territories subject to the Roman census.11 The census could also be imposed on new ter- ritories, a fate suffered by twelve Latin colonies that were forced to submit to a Roman-​style census (ex formula ab Romanis censoribus data) after failing to pro- vide military recruits to the Roman army during the lustrum of 209–​204 BCE.12 7. On the census of the Roman Republic, see the balanced discussion of de Ligt 2012, 79–​134, with the remarks of Marcone 2014; see further Hin 2013, 261–​297, 351–​353 (earlier version published as Hin 2008), and Nicolet 1991, 123–​130 and 2000, 19–​69. 8. Accordingly, the census was wielded as a means of social control. We are informed by Cicero that it was the duty of the censors to record the births of legitimate children and discourage bachelorhood among the elite; see Cic. De legibus 3.7. 9. On the procedures, see note 7. 10. Both Hin (2008, 215–​218) and de Ligt (2012, 85) point out that the exclusion of widows and orphans from the lists of capita civium (Livy 3.3.9; Per. 59.7) implies their registration, as further indicated by the finan- cial obligations imposed on them, such as the aes equestre; see Livy 7.41.8 (341 BCE), Plut. Cam. 2 (376 BCE), and Festus 183L, with Hin 2008, 209. 11. On the broad demographic scope of the republican census, see de Ligt 2012, 82–​105. The registration of freedmen and slaves is implied by the term familia; Cic. De leg. 3.7: censoris populi aevitates suboles familias pecuniasque censento. The registration of Latini is illustrated by a lex of 177 BCE ordering registered Latini and their descendants to leave Rome and return to their cities of origin; see Livy 41.9.9, 39.3.4–​6, 41.8.6–​ 12, with Laffi 1995. 12. See Livy 29.15.9–​10, with Le Teuff 2012, 183, 200; Kremer 2006, 81–​91. This may constitute early evidence for the use of decentralized census procedures, as noted by Nicolet 1991, 128–​129. 188  Anna Dolganov The demographic data generated by the census flowed into a number of dif- ferent registers and lists. During the Second Punic War, two thousand men were purportedly expelled from their voting tribes for avoiding military service during the previous lustrum, a story that implies that officials cross-​checked in- formation between census and recruitment records.13 The lex Voconia of 169 Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 BCE, prohibiting individuals worth more than 100,000 sesterces from naming women as heirs, indicates that registers for each property class could be readily consulted.14 Under the second triumvirate, such registers made it possible for the triumvirs to consider imposing a tax on the 1,400 wealthiest women at Rome.15 Livy mentions that the censors of 184 BCE ordered slaves under the age of twenty, who were purchased since the previous lustrum of 189 BCE for 10,000 asses or more, to be assessed at ten times their cost and taxed threefold. This story implies that census officials cross-​checked incoming declarations with those of the pre- vious census cycle.16 The concern of officials to keep track of the acquisition of slaves was presumably linked to Rome’s policy of granting freed slaves access to Roman citizenship. A  senatus consultum of 177 BCE ordered manumitters of slaves to swear an oath that the manumission did not serve the purpose of altering a person’s civic status, which suggests that voluntary enslavement and manumission were employed as a strategy to acquire Roman citizenship.17 The extension of the civitas Romana to the Italian peninsula in 89–​87 BCE as a consequence of the Social War strained the Roman political system and the institutions of the Roman census, which did not resume a regular cyclical pattern for several decades. In the ensuing period, Italian enfranchisement did not cease to be a matter of bitter controversy, reflected by the lex Papia of 65 BCE, which established a standing tribunal for prosecuting illegitimate claims of Roman cit- izenship.18 The decentralized census procedures implemented throughout Italy in the aftermath of the Social War are documented by the Tabula Heracleensis, a set of regulations published in the southern Italian municipium of Heraclea ca 45 CE, which includes a detailed description of how local censuses in Italian municipalities (municipia, coloniae, praefecturae) were coordinated with the census at Rome.19 One hundred and twenty days before the Roman census was 13. See Livy 24.18.1–​12. 14. On the lex Voconia, see Manthe 1999. 15. See Appian, BC 4.32–​33. The women were requested to submit an inventory of their property and valuables. Livy tells a similar story, that the censors of 184 BCE instructed their staff to increase taxes on women’s luxury items (Livy 39.44.3). We also hear of taxes being imposed on widows and orphans, of whom there were evidently lists; see note 10. 16. See Livy 39.44.3, with Brunt 1971, 15. 17. See Livy 41.9:  ciuitatis mutandae causa manu non mittere. For another reference to this strategy, see Petronius, Sat. 57: quia ipse me dedi in servitutem et malui civis Romanus esse quam tributarius. 18. On the lex Papia of 65 BCE, see Brunt 1971, 41–​42. Its consequences are illustrated by Cicero’s defense of Archias, who was prosecuted under this law in 62 BCE; see Cic., Pro Archia  6–​10. 19. See Crawford, Roman Statutes I: 355–​391, with Nicolet 1987 and 1991, 127–​129. For the argument that these procedures were an innovation of the age of Caesar, see Lo Cascio 1990 and 2001. For the more per- suasive argument that the Tabula Heracleensis probably reproduces earlier regulations, since there is some Documenting Roman Citizenship  189 due to be completed, municipal magistrates had sixty days to conduct their local census, taking declarations according to the same format as employed by the Roman censors. They were to enter these declarations into their public records (tabulae publicae) and send a copy of these records in the form of libri (pre- sumably papyrus rolls) to Rome. During the ensuing sixty days, the staff of the Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 censors was to examine the municipal records and deposit them in the archives at Rome (ll. 142–​156). Individuals with domiciles in multiple cities were to be regis- tered at Rome (ll. 157–​158), while holders of urban property at Rome were to file declarations (professiones) to establish the location of their properties, their obli- gation to repair the streets adjacent to them, and their ineligibility—​presumably by virtue of their double residence—​for the grain dole (annona civica, ll. 1–​20). This portion of the document finds an echo in Suetonius’s description of Caesar’s efforts to keep track of Rome’s urban population and its eligibility for the annona, ordering that the census be conducted “not in the usual manner and location, but street by street [vicatim] through the owners of houses within each street-​block [per dominos insularum].” Clearly, at some point in the late republic, the census came to be linked to cadastral records and street-​by-​street surveys of the city of Rome.20 Altogether, republican sources suggest a diachronic development whereby the expansion of the census within Italy gave rise to decentralized procedures where census records produced at the local level flowed into central record offices at Rome.21 This system of documentation and the topographical organization of the census described by Suetonius would be employed in the extension of the census to the provinces, to which we now turn. The Roman Provincial Census In the late republic and early principate, a centralized infrastructure of admin- istrative record keeping was progressively implemented throughout the Roman Empire.22 A  major pillar of this infrastructure was the institution of a cyclical census in every province.23 Already in the late republic, the provinces of Sicily and Bithynia observed a five-​year census on the Roman model that was supervised evidence for decentralized census procedures in Roman municipalities before the Social War, see de Ligt 2012, 106–​112; Nicolet 1987; Brunt 1971, 519–​523. 20. See Suet. Div. Iul. 41: recensum populi nec more nec loco solito, sed vicatim per dominos insularum egit. On the topographical turn in Roman administrative record keeping in the late republic and early principate, see Nicolet 1987 and 1991, 123–​148. 21. A significant shift from the circulation of people to the circulation of documents, as emphasized by Nicolet 1987 and 1991, 129. 22. See Nicolet 1991, 123–​208. 23. On the Roman provincial census, see Brunt 1981, 329–​335; Nicolet 1991, 123–​148; Jördens 2009, 62–​94; Le Teuff 2012, 248–​264, where skepticism is due regarding “la specificité égyptienne.” It is simply not true that in Egypt there were no declarations of property or cadastral registers (Le Teuff 2012, 327), see Jördens 2009, 103–​108. 190  Anna Dolganov by the governor.24 In the Augustan province of Egypt, the census was initially conducted in seven-​year cycles, then in fourteen-​year cycles from the reign of Tiberius, a periodicity determined by the traditional age at which the population became eligible for capitation taxes.25 In line with what is known about procedures in republican Sicily, the census in Egypt was initiated by an official pronounce- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 ment of the governor (praefectus Aegypti) and conducted by local officials within each of the forty-​five Egyptian districts (“nomes”). Census declarations were made “house by house” (kat’oikian) and contained the name, age, filiation, and civic and legal status of every person for whom the house served as a legal resi- dence. Some declarations were submitted by property owners for themselves and their families, while others listed tenants living in rented properties; property owners also declared houses in which no one permanently resided. In Egypt, the population census—​laographia, literally a “record of the people”—​was distinct from the assessment of landed property: declarants listed their urban property holdings but not their landholdings, which were subject to separate surveys and procedures of declaration. Household declarations were issued in multiple copies, collated and reviewed by officials, and deposited in local archives. Additionally, a complete copy of these records was sent to be reviewed and deposited in the central archives of the provincial capital.26 The similarity of these procedures to those described in the Tabula Heracleensis and the remarkable correspondence between the system of household declarations and Suetonius’s description of the Roman census being conducted street by street (vicatim) through the owners of houses (per dominos insularum) provide strong evidence that the census in Roman Egypt was broadly reflective of Roman administrative practice in Italy and elsewhere in the empire.27 The evidence for the provincial census beyond Egypt, although limited, supports the conclusion that a cyclical census was implemented in every Roman province. As far as we can tell, census cycles in different provinces were not co- ordinated and may have varied in length.28 The nature of the localities within which the census was conducted—​a city and its territory or a nome, prefecture, toparchy, or other type of district—​varied as well. It is reasonable to infer, on the basis of the Egyptian evidence, that the length of census cycles did not exceed the 24. On the five-​year census in Sicily and Bithynia, see Cic. Verr. 2.138–​139 and Plin. Ep. 79–​80, 112–​115, with the discussion of Le Teuff 2012, 211–​217. 25. In personal communication, Andrew Monson confirms my suspicion that, considering that the Roman census was introduced after the Romans had already instituted the poll tax building on Ptolemaic practices, the fiscal age in Egypt was most probably of Ptolemaic origin. Contra Le Teuff 2012, 331–​332. 26. On the procedures, see Hombert and Préaux 1952, 53–​147; Bagnall and Frier 1994, 1–​30; Jördens 2009, 68–​87; Le Teuff 2012, 332–​333. 27. On the link between the Tabula Heracleensis and the provincial census of the principate, see the important analysis of Nicolet 1987. 28. On the evidence for census cycles in other provinces, see Bérenger 2009, 154–​158; Le Teuff 2012, 248–​264. It is possible to identify patterns of fifteen years in the testimonia for Thrace and the Gallic provinces; see Le Glay 1981 and Jacques 1977, respectively. Documenting Roman Citizenship  191 age at which the population of a given province became liable to taxes. Likewise, the existence of separate procedures for the population census and the assess- ment of landed property was not limited to Egypt but is attested in Arabia and arguably implicit in the Roman legal sources.29 Through household declarations, local officials obtained detailed demographic Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 data that were linked to the topography of a particular locality.30 This informa- tion was instrumental for the collection of capitation taxes, including the poll tax (laographia, tributum capitis) to which most provincials were liable.31 Through land surveys, officials obtained similarly detailed records of landholdings and landowners, some of whom did not permanently reside in the locality.32 The demographic records produced by the census, combined with information fur- nished by land surveys and property declarations, enabled officials to estimate the total wealth and income of individuals, which, in addition to determining membership in the ruling classes (ordines), was the basis for the Roman system of compulsory public services (munera publica).33 Between census cycles, population records were supplemented by addi- tional declarations of birth and death. The submission of birth declarations was incentivized as an authoritative means of documenting free birth, filiation, and status, helping to secure the child’s status and inheritance claims for the future, especially in the eventuality that one or both parents should die before the next census. Documentary evidence shows that birth declarations were, in prac- tice, a requirement for members of hereditary status groups, who account for the majority of surviving declarations.34 Declarations of death were necessary for remission from capitation taxes and for claiming an inheritance from a de- ceased relative. Accordingly, papyrological evidence consistently shows death declarations being submitted promptly after the death of the individual. The census was compulsory, and attempts to evade it incurred heavy penalties.35 Failure to register oneself or members of one’s household resulted in 29. Separate procedures are suggested by two census declarations from the province of Arabia in 127 CE (P. Yadin 16, P.Hever 62) where individuals list their landholdings without providing information about their family. Ulpian’s use of the phrase in censum referre with reference to the declaration of land may imply a similar distinction; Dig. 50.15.4.pr. (Ulpianus 3 de cens.). 30. Local officials had at their disposal topographical surveys, see P.Oslo III 111 (Oxyrhynchus, 235 CE), with the remarks of Bagnall and Frier 1994, 23. The epikrisis pediakē (topographical inspection) attested in P.Oxy. X 1287 (205–​206 CE), PSI V 450 (early third century CE) and several other papyri illustrates precisely this: a detailed record of the urban topography within which individual homeowners and tenants were inscribed. 31. Other capitation taxes included a tax on unmarried Roman women of childbearing age (see BGU V 1210, ll. 87–​88) and annual contributions to the fiscus Iudaicus (see P.Lond. II 260 [ca 73 CE], ll. 432–​451). 32. See, e.g., P.Tebt. II 343 (Arsinoite, second century CE). 33. See, e.g., P.Cair.Isid. 6 (Arsinoite, 304–​305 CE), a register of declarations of land belonging to landholders near the village of Karanis, and BGU I 6 (Arsinoite, 157–​158 CE), a list of candidates for local offices with their net wealth (poros). 34. In Roman Egypt, birth declarations played an instrumental role in the status verification (epikrisis) of Roman, Alexandrian, and Antinoite citizens, while their absence undermined claims of citizenship and inheritance, see the following sections on birth and epikrisis. 35. See Bagnall and Frier 1994, 14, 25, 29; Hombert and Préaux 1952, 97–​99. 192  Anna Dolganov the confiscation of one-​fourth of one’s property for every unregistered person.36 Failure to register slaves resulted in their confiscation, illustrated by a hearing where a Roman woman who failed to declare her houseborn slaves sees them confiscated by a Roman procurator.37 At each census, individuals were routinely requested to produce evidence of previous census declarations—​or, if unable to Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 do so, to explain why a certain declaration was missing.38 The practice of cross-​ checking declarations was presumably an effective deterrent from attempts to evade registration or submit inconsistent information. In papyri, we occasionally find lists of individuals who failed to show up for the census and were recorded as absent from their legal residence (aparastatoi). Sometimes their whereabouts are specified; elsewhere, it is noted that their absence is pending investigation.39 Other documents mention lists of persons who have permanently abandoned their place of residence (anakekhōrēkotes).40 A few papyri attest to persons not registered in the census (anapographoi) being reported and caught as undocu- mented.41 The term anapographos itself implies an administrative expectation that the majority of the population would be captured by the census. Overall, the Roman census as we observe it in the towns and villages of the Egyptian hinterland appears to have been effective at capturing and tracking the population. Where this task could become problematic was in very large cities such as Alexandria. This is illustrated by an edict of the prefect Vibius Maximus announcing the start of the census and ordering individuals to return to their place of legal residence and “resume their duties of land cultivation.” Those who believed they had a legitimate reason to remain in Alexandria had to register themselves with a military officer and obtain permits of residence within thirty days.42 This stipulation suggests that during the period of the census individuals could expect to be stopped on the street and frisked for their papers. The edict of Maximus appears to have been part of a broader tendency of Roman 36. See BGU V 1210, ll. 150–​154. For Romans and Alexandrians, the total penalty was limited to one-​fourth. 37. See M.Chr. 372 = P.Catt. recto col. vi (135 CE). The confiscation of unregistered slaves is likewise stipulated in BGU V 1210 (Arsinoite, 149–​161 CE), l. 155. 38. See Bagnall and Frier 1994, 20, 28. In PSI XII 1230 (Oxyrhynchus, 203 CE), a metropolite explains to local officials that he is missing from the last census of 201 because he spent time abroad. In SB XX 15337 (161 CE, Arsinoe), ll. 13–​17, the head of a household explains that the resident Ptolemaios son of Diogenes, who now goes by the name Gaius Valerius Capito, was registered in the previous census as Ptolemaios son of Marcus Anthestius (presumably the military name of his father). 39. In the tax records of P.Sijp. 26 (51 CE), an individual is recorded as returning from abroad (ll. 115–​119). In the administrative records of P.Lond. II 260 (ca 73 CE), individuals are recorded as traveling beyond Egypt (ll. 545–​549). See also P.Berl.Cohen 17 (second century CE), where a list of absent persons is annotated by a second hand that gives their whereabouts or states that this is being investigated. 40. See P.Oxy. II 251 (44 CE), where a father requests that his son be registered as having permanently moved abroad. 41. See P.Sijp. 26 (51 CE), ll. 109–​114, a list of undocumented fourteen-​year-​olds subject to the poll tax, and PSI III 229, 232–​233, where warnings are issued to unregistered persons. See also P.Kramer 7 (second cen- tury CE), where a private individual reports an undocumented person, and PSI XIII 1326 (181–​182 CE), where an undocumented person plaintively states that he was orphaned as a child and never registered. 42. See P.Lond. III 904 col. ii = W.Chr. 202 (104 CE), with Le Teuff 2012, 178–​180. Documenting Roman Citizenship  193 administrators to exploit the census as an occasion to identify and expel migrants from major cities in the empire.43 The demographic data yielded by the census and supplemented by additional declarations flowed into a variety of registers and lists.44 One remarkable docu- ment is a contract for nine scribes hired by local officials to process incoming Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 census returns. First, they were to produce lists of persons subject to the poll tax, ordered by name (and, presumably, place of residence). Second, they were to create updated lists of the Hellenic class (katoikoi), whose members paid a reduced poll tax, as well as lists of minors and other individuals exempt from cap- itation taxes. Finally, they were to update the population lists of the previous year on the basis of the current census. From other documents, we learn that local officials kept registers of births and deaths and population lists organized by year of birth, which made it possible to keep accurate lists of taxpayers—​including minors due to enter the age of fiscal liability in any given year.45 The presence of roll and column numbers in census rolls and the practice of citing these numbers when referring to census declarations indicate that the rolls were meant to be consulted in the archives when necessary.46 These indexing marks—​tomos (roll), deltos (tablet), kollēma, and selis (page)—​are not attested in Ptolemaic documents but emerge in diverse genres of administrative records in the Roman period. They appear to reflect the implementation of a distinctly Roman model of archival practice, employed at Rome as early as the second cen- tury BCE.47 How these records were used is significant:  when individuals cite official records in administrative proceedings, it is stated (or can be deduced) that they present copies of the records in question. It is often specified that copies have been made from records that have been “looked up” (epeskemmena) in the archives. There are numerous surviving copies of this sort, some displaying signs of correction by the archivist.48 In copies of records, we consistently find 43. Compare the edict of Caracalla expelling migrants from Alexandria, likewise on the occassion of the census, in P.Giss. I 40 col. ii, ll. 16–​29 (215 CE). See also the sources mentioning expulsions of Jews and Isiac worshipers from Rome, collected in Rutgers 1994. One such expulsion resulted in the shipment of four thousand freedmen to Sardinia (Tac. Ann. 2.85). 44. See Hombert and Préaux 1952, 135–​147; Bagnall and Frier 1994, 27–​28. 45. See P.Mich. XI 603 (Arsinoite, 134 CE), with Bagnall and Frier 1994, 27–​28. 46. See Bagnall and Frier 1994, 28, noting that census records included the roll and column numbers of previous declarations. For a striking example, see P.Lond. II 324 = W.Chr. 208 (161 CE), containing two extracts from archival records of census declarations dating twenty-​five and thirty years back, ordered by roll (tomos) and column (kollēma). These declarations had been retrieved from the archives to prove that two siblings were the children of the same mother. 47. See, e.g., RDGE 12 (ca 129 BCE) and RDGE 23 (73 BCE), two Greek inscriptions containing extracts from the acta of Roman officials with cataloguing marks indicating their location in the archives at Rome. These records were ordered by deltos and selis, the Greek rendition of the Latin tabula and pagina, the terms used in birth declarations of Roman citizens in Egypt, discussed in the next section. 48. For references to authenticated copies (epeskemmena) from archival records, see, e.g., BGU I 73 (Arsinoite, 135 CE) and W.Chr. 77 (Arsinoite, second century CE). Surviving examples of such copies include CPR I 18 (Arsinoite, 124 CE), SB IV 7362 (Arsinoite, 188 CE), and SB III 6995 (124 CE), all corrected in red ink by archival officials. 194  Anna Dolganov cataloguing marks specifying the archival location, such as the year, locality, roll, and column number of a census declaration or the year and locality (city quarter, street block) under which a person’s entry could be found in population registers. Altogether, the evidence supports the conclusion that individuals who sought to make claims on the basis of administrative records were expected to possess the Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 relevant information that would enable these records to be checked (episkeptein) in the archives and were generally expected to submit authenticated copies of these records to officials. Accordingly, during a hearing in 87 CE, an association of Egyptian priests presented a judicial record from the year 46 that had been duly verified (epeskemmenos) in the archives at Alexandria.49 Revealing insight into the mentality underpinning Roman archival institutions is furnished by the correspondence of a local official (royal scribe) in Roman Egypt who was filling in for his senior colleague as strategos of the nome. From his position as deputy strategos, he continued to draft official letters to himself as royal scribe, even though he was occupying both positions.50 Clearly, the pur- pose of the correspondence was not simply to convey information but to create an official record of the communication, every component of which could later be verified, even after the memory of the fact that for a short time the royal scribe and the strategos were functionally one and the same had faded. As an example, we observe a villager traveling to the archives of the regional capital to procure an extract from the correspondence of a local official as proof that his land had suffered damage from an inadequate water supply.51 A  record-​keeping system in which data could be looked up and retrieved over periods of several decades established Roman administrative records as an abundant and highly authorita- tive source of legal evidence.52 It is no coincidence that the strong notion that the existence of a claim was predicated upon it being “on record” (en katakhorismō) and capable of being “looked up” (epeskemmenon) in the archives first emerges in documents from the Roman period. The documentary evidence shows that the Roman provincial census was much more than a periodic collection of demographic information at intervals of ten to fifteen years. It would be more accurate to think in terms of an infrastructure or matrix into which demographic data from various sources systematically flowed (household declarations, reports of birth and death, manumissions, slave sales, changes in status or residence) and with which other bodies of information (e.g., cadastral records and fiscal records) systematically communicated. These forms of administrative record keeping are not attested in Ptolemaic Egypt. Instead, they emerge in the Roman period, showing clear parallels to what is known about 49. See P.Vind.Bosw. 1 (Arsinoite, 87 CE), ll. 7–​8, 18–​19. 50. See SB XVIII 13175 = W.Chr. 52 (Nesut, 194 CE), with Kruse 2002, II: 857–​858. 51. See P.Wisc. I 31 (Arsinoite, 147 CE), with my remarks in Dolganov 2021a. 52. On the impact of Roman record-​keeping institutions on legal evidence and factuality, see Meyer 2004, 216–​250; Ando 2020b. Documenting Roman Citizenship  195 records and archives in the city of Rome. Consequently, there is no reason to assume that these practices were specific to Egypt and dramatically different in other provinces. On the contrary, what is known about the census beyond Egypt supports the conclusion that similar arsenals of records were available to local officials who administered the census—​and used the census to administer taxes Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 and compulsory public services—​elsewhere in the empire. Roman Citizens and the Census under the Principate Roman citizens were the single most privileged status group in the Roman Empire, with numerous fiscal immunities and legal privileges. In the provinces before the Constitutio Antoniniana, Roman citizens made up a minority, ranging from one-​tenth to no more than one-​third of the total imperial population.53 As with other status groups, the main apparatus through which officials collected information about Roman citizens was the census. The evidence, although limited, suggests that a cyclical five-​year census con- tinued to be conducted in Italy under the principate.54 However, the progres- sive extension of Roman citizenship to provincials meant that the census in Italy ceased to fulfill the function of a census populi Romani. How this traditional pro- cedure evolved is not well documented. In his Res Gestae, Augustus cites census figures issuing from a census populi conducted by him on three occasions, in 28 BCE, 8 BCE, and 14 CE, respectively.55 These figures, which show a fourfold in- crease in Roman capita civium from the census of 70–​69 BCE, may reflect the inclusion of Roman citizens residing in the provinces.56 This would mean that Augustus thrice conducted a coordinated imperial census of Roman citizens. An empire-​wide census populi is subsequently attested under Claudius (47–​48 CE) and once again under Vespasian (73–​74 CE).57 The Claudian census populi is re- flected by a single piece of documentary evidence from Egypt:  a Latin census declaration submitted by a legionary veteran named Pompeius Niger, the format 53. See the important work of Lavan 2016b and 2019a, who employs quantitative models to demonstrate that the total number of Roman citizens within the empire in the first and second centuries has been signifi- cantly overestimated. 54. The Augustan evidence for the census strongly suggests a five-​year cycle: a census populi Romani in 28 BCE (RGDA 8), evidence for the appointment of censors in 22 BCE (Cass. Dio, Roman History 54.2.2), two lectiones senatus in 18 BCE and 13 BCE (Cass. Dio, Roman History 54.13, 54.26.3), another census populi Romani in 8 BCE (RGDA 8), further evidence for an Italian census in 4 CE (Cass. Dio, Roman History 55.13.4) and a third census populi Romani in 14 CE (RGDA 8); see Le Teuff 2012, 348. A five-​year census and lustrum in Italy is assumed by an Antonine legal writer in the Fragmentum Dositheanum de manumissionibus 17; see Liebs 1997b. 55. See RGDA 8, with Le Teuff 2012, 22–​23, 348–​350. 56. On the Augustan census figures, see de Ligt 2012, 120–​135, and Hin 2008, with an overview of the long-​ standing scholarly debate. 57. On the census and lustrum of Claudius, see Suet. Cl. 5.16, Plin. NH 7.159, and Tac. Ann. 11.25. On the census of Vespasian and Titus, see Suet. Vesp. 8 and 17, Tit. 6, and Plin. NH 7.162–​164, with Bosworth 1973. That the latter emperors took on the censorship is taken to imply that they conducted an empire-​ wide census populi Romani. 196  Anna Dolganov of which differs from that of Greek kat’oikian declarations and corresponds to what is known about declarations in Italy.58 It appears that the Claudian census populi required Roman citizens in the provinces to submit special declarations—​ in Latin and according to a specific format—​that were collected and processed by the staff of governors and the results forwarded to Rome. Lack of evidence Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 for a coordinated, empire-​wide census populi Romani after Vespasian suggests that this was an extraordinary procedure, occasionally undertaken by emperors as an act of symbolic significance and possibly conducted for the last time in the Flavian period. Under ordinary circumstances, Roman citizens in the provinces were subject to the regular provincial census. This is evidenced by kat’oikian declarations filed by Roman citizens in the Egyptian hinterland, several of which are declarations of property where no one is registered and at least two of which are “pri- mary” declarations where Roman citizens declare themselves and their family members.59 Evidently, Romans residing in provincial communities had the pos- sibility of filing their census declarations with local officials.60 For residents of Roman municipia and coloniae, the procedure was fundamentally similar, since it can be inferred on the basis of the Tabula Heracleensis and other sources that Roman municipalities conducted local censuses that were coordinated with the provincial census.61 In view of the authority of Roman administrative records as the basis of legal claims, to what extent could individuals legitimize false claims by having them go “on record” with Roman administrators? Was it possible, for example, to usurp Roman citizenship by fraudulently asserting it in a census declaration? The evi- dence suggests that the submission of declarations was a face-​to-​face procedure conducted within small localities such as villages and neighborhoods of larger 58. See PSI XI, 1183, with Rathbone 2001 and Le Teuff 2012, 219–​220. Rathbone is surely right to link this declaration with the Claudian imperial census of 47–​48. 59. For declarations of property where no one is registered, see SB XII 10788 (Oxyrhynchus? 61 CE), BGU I 53 (Dionysias, 133 CE), BGU VII 1581 (Philadelphia, 147 CE), and P.Stras. IV 2680 (Philadelphia, 174–​175 CE,). For “primary” declarations by Roman citizens, see the fragmentary SB VI, 9573 (Karanis, 175 CE), where a Roman woman declares herself, her sister, and her children, and SB XX 15337 (Arsinoe, 161 CE), where a Roman veteran registers himself, his son, two female slaves, and a resident named Ptolemaios. These declarations provide unequivocal evidence for the participation of Roman citizens in the provincial census. The argument of Rathbone 2001, 109–​111, that Romans and Alexandrians were exempt from the poll tax and therefore exempt from the census—​hence the two “primary” Roman declarations were made as an “error of safety”—​is implausible. The purpose of the census was not merely fiscal but demographic (see Bagnall and Frier 1994, 11–​13), and the participation of Antinoites (likewise exempt from the poll tax) surely implies that Alexandrians participated as well. 60. The evidence from Egypt shows that Roman citizens (in particular, veterans) could declare any provincial community their legal residence (idia  =  origo), as illustrated by the formula boulomenos parepidēmein (“wishing to reside in”; see, e.g., BGU I 113 = W.Chr. 458, after 143 CE). The acquisition of local citizenship was, of course, a separate question. 61. On the coordination of municipal censuses in Italy with the census at Rome, see the discussion at note 19. On the coordination of civic censuses with the provincial census in Sicily and Bithynia, see Cic. Verr. 2.138–​139 and Plin. Ep. 79–​80, 112–​115, respectively, with Le Teuff 2012, 211–​217. Documenting Roman Citizenship  197 cities. The Roman subdivision of cities into residential districts, in line with a well-​ attested practice in the city of Rome, ensured that declarations were collected in localities of manageable size under the supervision of local magistrates.62 A face-​ to-​face dynamic no doubt enhanced the ability of officials to verify the identity of declarants and the information submitted by them. As papyrological evidence Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 shows, census declarations were subject to routine cross-​checking with previous censuses and population registers.63 This presumably made it difficult to declare oneself a Roman citizen in a community where one was previously registered as a non-​citizen without providing evidence of how one acquired this status. To reg- ister for the first time in a new community as a Roman citizen likewise involved the submission of documents, as in the case of veterans who settled in the hinter- land of Egypt and presented their military diplomas (or other evidence of their honorable discharge) and copies of their status verification (epikrisis) to local officials.64 Even if one managed to secure the cooperation of officials, it is unlikely that registration in the census would suffice to make lasting claims of Roman citizenship. In numerous administrative contexts, ranging from sales of real pro- perty to entry upon inheritance to procedures of epikrisis, the lack of more sub- stantial evidence such as a Roman birth declaration or evidence of family history was likely to make the defective status visible to officials. Beyond the sphere of Roman municipalities, from the perspective of local officials in most places in the empire before the Constitutio Antoniniana, Roman citizens were a conspicuous minority with fiscal exemptions that distinguished them from the rest of the local population. In the hinterland of Egypt, officials assiduously collected information pertaining to Roman citizens. This data was compartmentalized and entered into lists: surviving administrative records in- clude lists of Roman citizens who resided or owned property in the locality; Roman citizens also feature in tax lists as a separate group.65 It is clear that officials in Egypt had a very precise sense of who the local Roman citizens were, where they resided and owned property, when (if relevant) their Roman status was acquired, and whether it had been formally verified by the governor (the 62. On the Roman subdivision of cities into districts (in Egypt, amphoda, laurai), see Alston 2002, 138–​165, who draws a connection to the system of curiae and vici at Rome. Indeed, it is intuitive to link amphoda and amphodarchoi in Egypt with Roman vici and vicomagistri; see Tarpin 2001 and 2008. 63. See notes 38 and 39. 64. See, e.g., P.Oxy. VII 1023 (second century CE, Oxyrhynchus), a certificate issued by local officials that a Roman veteran has provided evidence of his honorable discharge and epikrisis. See also PSI V 447 (166–​ 167, Oxyrhynchus), a copy of an epikrisis record submitted by a Roman veteran to the local strategos. On epikrisis, see the final section of this chapter. 65. For Roman citizens as a separate category in population registers and tax lists, see SB XX 14433 (second century CE, Diospolis Magna); P.Lond. II 260 (ca 73 CE, Ptolemais Euergetis, city quarter [amphodon] of Apolloniou Parembolē), ll. 330–​349, 626–​631; BGU IX 1894 (ca 157 CE), ll. 154, 213, 217; and P.Oxy. III 597 (108–​109 CE). The document from the Arsinoite amphodon refers to a “register [graphē] of Romans and Alexandrians” compiled by the amphodarchos for the fifth year of Vespasian. 198  Anna Dolganov procedure of epikrisis, to be discussed in detail).66 Much of this data derived from declarations submitted by the individuals themselves, but the evidence also shows officials proactively requesting and verifying information.67 For example, the evidence for village tax collection shows that tax collectors went from house- hold to household, while several references to public registers of “topographical Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 inspection” (epikrisis pediakē) appear to indicate that officials in third-​century Oxyrhynchus went from door to door to collect detailed information about the de facto inhabitants of each urban dwelling.68 The compartmentalization of records at the local level went hand in hand with their centralization in the provincial capital. It is clear from papyrological sources that governors disposed of detailed lists of fiscally privileged groups—​Roman citizens, citizens of Greek cities (the so-​called astoi), priesthoods, citizens of the mētropoleis. The availability of lists of Roman citizens is illustrated by a curious piece of literature produced by Phlegon of Tralles, a freedman at the court of Hadrian. In his Makrobioi, a study of longevity, Phlegon includes a list of men and women in the Roman Empire who had attained the age of 100 or more. The majority are freeborn Romans (with a smaller number of freedmen) from cities in Italy. They are classified by age, starting with those who reached 100, followed by several whose age ranges from 100 to 120. Like Pliny the Elder be- fore him, Phlegon states that he has laboriously obtained this information from census records.69 The shape of his list suggests that he looked through population registers of Italian cities—​probably lists organized by year of birth, well known from papyrological sources—​that had been deposited in the archives at Rome. Even more striking, however, is Phlegon’s inclusion of individuals—​both Roman citizens and peregrini—​from cities in the provinces of Achaea, Bithynia, and Lusitania. To obtain this information, one would imagine that Phlegon wrote to several governors to ask their staff to search the provincial census records and produce lists of persons older than 100.70 The Makrobioi of Phlegon thus provides 66. As stated in the Arsinoite amphodal register of Roman citizens (mentioned in the previous note): “and the total for the fifth year of Vespasian amounts to  .  .  .  of whom two are Roman citizens and one has obtained Alexandrian citizenship, for a total of three, of whom the aforementioned Aulus Valerius Dextrus has undergone status verification [epikekristhai] under Ponticus, while Lucius Valerius Fronto remains ­unverified and was not present for the census.” 67. For example, we find a list of Roman men who have “demonstrated” that their status has been verified by the governor and that this information has been conveyed to the regional procurator (epistrategos, P.Wash. Univ. I 3, third century CE, Oxyrhynchus). This list includes men of different ages who underwent epikrisis over several decades, which suggests that at some point, local officials asked a series of Roman men to submit evidence of their epikrisis. 68. On the topographical organization of village tax collection, see Hanson 1978 and 1994. For the epikrisis pediakē, see note 30. 69. See Phlegon, Peri makrobion (FGrH 257 F 37) in the edition of Stramaglia 2011, sections 1–​4, with his commentary. 70. Pace Braunert 1957, the evidence does not support the notion that provincial census records were available at Rome. Accordingly, the emperor asked governors to provide the census data of candidates for imperial citizenship grants; see notes 148 and 149. Documenting Roman Citizenship  199 remarkable testimony for the organization and retrievability of Roman adminis- trative records, which enabled officials to obtain very specific information about their provincial subjects. Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 Documenting Citizenship and Its Acquisition Documenting Birth It is generally accepted that regular procedures for declaring birth and death were instituted during the reign of Augustus. For Roman citizens, the procedures of birth declaration were outlined in the Augustan laws regulating marriage and manumission—​the lex Aelia Sentia of 4 CE and the lex Papia Poppaea of 9 CE.71 A number of extant birth declarations by Romans in Egypt (60–​242 CE) document these procedures in detail. The majority are authenticated copies (descriptum et recognitum) of declarations (professiones) submitted at Alexandria and entered into the Roman birth register, the so-​called album professionum. A few declarations are affidavits (testationes) attesting to the births of illegitimate Roman children who were excluded from the album by the Augustan laws. Both types of documents were drafted in Latin on waxed wooden diptychs (tabulae, the traditional medium of Roman legal documents) with an exterior text and a hidden interior text, with the names and seals of seven witnesses, all of them Roman citizens.72 In light of these documents, the procedure for registering the birth of a Roman citizen in Egypt may be described as follows. For a legitimate child born to a Roman father, the birth declaration (professio) had to be registered in the provincial capital—​most likely within a stipulated period of thirty days—​by the father in person or through a representative.73 The term professio implies that the declaration was made before an official. If it was accepted, a record of the declaration was created that included the name of the father, the name and gender of the child, the name of the mother, and the Roman date of birth, for 71. Both laws are mentioned in extant birth declarations. On declarations of birth and death, see Geraci 2001 and Haensch 1992, 283–​290. Roman birth registration is the subject of detailed studies by Schulz 1942–​ 1943 and, more recently, by Sánchez-​Moreno Ellart 2001 and 2004. 72. For an appendix of Roman birth declarations (professiones and testationes), see Haensch 1992, 306–​313. For the texts themselves, see Sánchez-​Moreno Ellart 2001, 113–​165. On Roman tabulae as a documen- tary medium, see the useful overview of Cooley 2012, 73–​82, 171–​172, and the extended treatment of Meyer 2004. 73. A time limit of thirty days is mentioned in the Historia Augusta (HA Marc. 9.7). While this source cannot stand alone, it is corroborated by all seven dated examples of professiones in albo by Roman citizens. Further evidence is provided by an Antinoite birth declaration (P.Ant. I 37, 208–​209 CE) filed “within twenty-​nine days in accordance with . . .”—​presumably with Roman regulations. From the reign of Septimius Severus, new procedures allowed older children to be registered; see Sánchez-​Moreno Ellart 2001, 158–​165. The majority of our evidence shows Roman birth declarations being submitted by a parent in person. For three cases where a father acts through a representative, see BGU VII 1694 = CPL 157 (Philadelphia, 163 CE), P.Mich. III 168 = CPL 153 (145 CE), and SB VI 9228 (Syene, 160 CE). 200  Anna Dolganov example: “M. Valerius Turbo declares his son M. Valerius Maximus, born from Antonia Casullutis on the fifteenth day before the Kalends of September that have recently passed, to be a Roman citizen.”74 These records were inscribed on a whitewashed wooden board, the so-​called tabula albi professionum, displayed at the location of the governor’s tribunal at Alexandria.75 After a period of dis- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 play, the tabula albi was taken down, copied, and archived; the declarant then had the possibility to make an authenticated copy (descriptum et recognitum) of the declaration from the archival records.76 In addition to the text of the declara- tion, each copy contained precise information about the archival location of the record: the consular and imperial year, the name of the governor, the number of the tabula, and the specific pages (paginae) on which the record could be found.77 It is clear from these indexing marks that Roman birth declarations were organ- ized in the archives by consular year, under which one could locate the text of the relevant tabula albi and find the correct pagina.78 It is noteworthy that copyists made an effort to give a visual description of the record, noting prescripts and marginal annotations, as well as the sizing of the letters. Some scribes copied down the exact date written on every pagina as they leafed through them: “ta- bula 8, on pagina 2 it is written in letters of larger size: ‘in the second consulship of L.  Nonius Torquatus Asprenas and M.  Antonius Libo’ and subsequently on pagina 9: ‘on the sixth day before the Kalends of April’ . . .” followed by the record itself.79 These annotations clearly indicate that records of birth declarations were meant to be consulted in the archives when necessary. Several of our documents are copies of birth records that were made many months after their deposition in the tabularium of the governor.80 74. BGU VII 1692 = FIRA III 3 = CPL 152 (Philadelphia, 144 CE), ll. 16–​19. For a detailed description of the format of Roman birth declarations, see Schulz 1942, 85–​87, and 1943, 57–​59. The concluding formula CREADK or CREAK has been variously interpreted. It seems likely that CRE stood for c(ivem) R(omanum) e(sse). The interpretation of ADK and AK as ad K(alendas) or ad K(alendarium) is more controversial; see Schulz 1942, 88, and Haensch 1992, 286. 75. On the atrium magnum (see BGU VII 1691  =  CPL 150, Philadelphia, 109 CE) and forum Augusti (see P.Mich. III 166 = CPL 151, Arsinoite, 128 CE) as locations of the governor’s tribunal, see Capponi 2010, 258–​260, and Kelly 2011, 170. 76. Schulz 1942, 87–​88, assumes that the tabulae albi were copied onto papyrus rolls. However, the citation of birth records by tabula and pagina rather suggests that the tabulae albi were copied onto bundles of waxed tablets, as were judicial records in the West; see CIL X 7852 = ILS 5947 (69 CE). 77. The same cataloguing marks are attested—​in Greek (deltos, selis)—​in archival records from the city Rome; see note 47. 78. In two instances, a tabula begins with an entry from the previous consular year, then transitions into the next year; see BGU VII 169 = CPL 150 (109 CE) and P.Mich. III 166 = CPL 151 (Arsinoite, 128 CE). This suggests that the tabula albi had a standard size: when a given tabula had only one entry at the end of the year, it was continued into the next year until it was full. Accordingly, the scribal practice of describing each pagina is observable in instances where the consular year at the beginning of the tabula was different from the consular year of the entry. 79. P.Mich. III 166 = CPL 151 (Arsinoite, 128 CE), pag. iii, ll. 1–​4. 80. That Roman birth declarations were archived in the governor's tabularium is tentatively concluded by Haensch 1992, 296. It is difficult to imagine that the declarations as we have them—​drafted in Latin and registered in the provincial capital under the authority of the governor—​could have been archived anywhere else. Documenting Roman Citizenship  201 Under the Augustan marriage laws, only children born into a legitimate Roman marriage could be registered in the album professionum. For an illegit- imate child born to a Roman mother, the birth declaration took the form of an affidavit (testatio) in which the mother declared the birth before seven Roman witnesses. Although excluded from the album, testationes of birth could likewise Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 be made before officials and entered into the public records—​in actis profiteri, in the words of the Roman legal sources.81 This procedure is documented by a testatio of a Roman woman named Sempronia Gemella, who sailed to Alexandria from the Arsinoite village of Karanis to declare her illegitimate twins one month after their birth. Clearly, Gemella undertook the journey to register the twins in the public records of the provincial capital. Her testatio was drafted in Latin on tabulae before seven Roman witnesses in two copies, one of which was kept by Gemella and the other presumably deposited in the governor’s archive at Alexandria. An analogous procedure is mentioned by Apuleius, who describes the birth declaration (professio) of his wife Pudentilla being drafted on two sets of tabulae, one of which was kept by her family and the other deposited in the tabularium publicum.82 Without exception, the evidence from Egypt shows the registration of Roman birth declarations (both professiones in albo and testationes) taking place in the provincial capital.83 Even during periods of the year when the governor was typi- cally at Memphis or elsewhere in Middle Egypt on his assize tour, Roman citizens traveled to Alexandria to register the births of their children.84 The lack of any evidence for Roman birth declarations being submitted beyond Alexandria does not seem to be a coincidence. It follows that, in Egypt, the birth registration of Roman citizens could only take place in the provincial capital.85 A Roman birth declaration was a highly authoritative official document that was regarded by officials as strong proof of Roman citizenship. This is clear from proceedings of status verification (epikrisis) before the prefect of Egypt, where—​ without exception—​Roman citizens present their registered birth declarations.86 81. A professio denotes a declaration before an official; hence a testatio of birth that was registered with of- ficial authorities was a professio in its own right. This is rightly emphasized by Sánchez-​Moreno Ellart 2001,  78–​79. 82. For the testatio of Gemella, see P.Mich. III 169 = FIRA III 4 = CPL 162 (Karanis, 145 CE), with Strassi 2001, 1224–​1226. For the Apuleius passage, see De magia 89.1–​4—​whether this tabularium publicum was a mu- nicipal archive or the archive of the governor at Carthage is unclear. 83. The locally drafted testationes of auxiliary soldiers are a different matter, as they were not Roman citi- zens and could not legally marry. These “constraints of military service” (districtio militiae; see BGU VII 1690 = FIRA III 5 = CPL 160, Philadelphia, 131 CE) barred soldiers from registering birth declarations with Roman authorities in the provincial capital. 84. See, e.g., P.Mich. III 166 = CPL 151 (March, 128 CE) and P.Mich. III 169 (April, 145 CE). 85. This is also the conclusion of Haensch 1992, 287. The possibility of appointing a representative in itself provides indirect evidence for the requirement that registration take place in the provincial capital. 86. There is not a single known instance where a Roman citizen undergoing epikrisis does not present a birth declaration. On the epikrisis (status verification) of fiscally privileged groups in Roman Egypt (discussed in the final section), see Nelson 1979; Foti Talamanca 1974, 56–​68; Haensch 1992, 290–​293; Kruse 2002 I: 252–​271, II: 638–​640; Kruse 2013; Kruse 2019, 123–​126. 202  Anna Dolganov For freeborn Roman women, a birth declaration is the main document submitted as evidence of Roman status in epikrisis proceedings.87 By contrast, census declarations and extracts from population registers are never cited as evidence of Roman citizenship, which reflects their inferior—​and, one would guess, implic- itly derivative—​authority with respect to birth declarations.88 Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 To what extent were Roman officials able to verify the accuracy and authen- ticity of birth declarations? The professiones themselves inform us that they were submitted at Alexandria “without investigative proceedings” (citra causarum cognitionem). While this means that the governor did not give a formal audi- ence (cognitio) to declarants—​and, indeed, may even have been absent from the capital—​it does not imply that Roman birth declarations were accepted without verification.89 In order to register a child in the album professionum, the declarant had to be a Roman citizen in a legitimate marriage—​all others were excluded by the Augustan laws. In view of the exclusive nature of the album, one would expect that declarants were required to present some evidence of their identity, Roman status, and legitimate Roman marriage. While the texts of professiones give little insight into the procedures of declaration, it is relevant that birth declarations by Alexandrians and Antinoites always involved three guarantors who person- ally vouched for the identity of the parents, their marriage—​often with a copy of the marriage contract—​and the identity of the baby as their natural child.90 One would not expect the status of Roman citizens to have been verified less rig- orously. In view of the consistent presence of three guarantors at the epikrisis of Roman citizens before the governor, it is reasonable to infer that guarantors were involved in the process of birth registration as well. In view of the centralized procedure requiring Romans to travel away from their residence to the provincial capital, it is likely that declarants brought documents (such as military diplomas or their own birth declarations) to certify their status. The registration of birth declarations in the governor’s tabularium and the practice of documenting copies on sealed tabulae, with indexing marks that made it possible for the record to be located, further contributed to the authority of these documents. Indeed, it is 87. See P.Oxy. XII 1451 (Oxyrhynchus, after 175 CE), where a woman with the ius trium liberorum registers her children and presents her own professio as proof of her Roman status. See also P.Diog. 6–​7 (Arsinoite, after 142 CE), where a freedwoman presents her tabulae manumissionis and the birth declaration of her patrona. For freeborn Roman boys undergoing epikrisis, a birth declaration is always presented; for adult men, the epikrisis record replaces the birth declaration as proof of status. 88. The invariable use of Roman birth declarations (but never census records) as proof of status in epikrisis proceedings speaks against the central thesis of Sánchez-​Moreno Ellart 2001 that birth declarations were not more authoritative than other types of declarations. See contra Sherwin-​White 1973a, 314, “but in- scription on the [census] rolls provided only a presumption of citizen status. It was not a proof in itself. That depended normally on evidence of birth,” etc. 89. That citra causarum cognitionem meant “to accept the professio without controlling its truth” was argued influentially by Schulz 1942, 87–​88 and continues to be widely accepted; see, e.g., Parkin 2003, 178–​182. 90. See in particular SB V 7602 (Antinoupolis, 151 CE), which spells out that the guarantors certify the mar- riage and birth. On the procedures of birth registration for Antinoites and Alexandrians, see Bell 1933 and Delia 1991, 55–​58, 68–​69, respectively. Documenting Roman Citizenship  203 very difficult to imagine how Roman birth declarations could have been routinely accepted by governors as evidence of Roman citizenship in epikrisis proceedings if Roman officials were not confident in their ability to verify their contents. On the basis of this evidence, we may conclude that a Roman birth decla- ration (professio or testatio) registered with the governor and deposited in his Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 tabularium, with an authenticated copy drafted on tabulae before seven Roman witnesses, was the single most authoritative document demonstrating the status of a freeborn Roman citizen.91 As such, Roman birth declarations also constituted authoritative evidence of filiation and age, evidenced by Apuleius’s use of the professio of his wife Pudentilla to demonstrate that she was not older than the maximum age of legal marriage stipulated by the Augustan laws.92 When the Roman jurist Modestinus states that “age is proven on the basis of birth regis- tration or on the basis of other legitimate proofs,” it is implicit that a registered birth declaration was the first point of recourse, lacking which one could resort to other evidence.93 This is corroborated by numerous citations of Roman legal writers who discuss how age or status can be proved in the absence of a birth declaration—​implying that a birth declaration would have provided the neces- sary evidence, which now had to be obtained from other sources.94 By contrast, when a birth declaration was available, it was treated as authoritative.95 In view of the authority of birth declarations as evidence of Roman status, one would expect birth registration to have been documented in a centralized fashion in every Roman province. Whether all provincials had to travel to the capital is less clear. In Egypt, where there were no Roman municipalities, the only state archive in which Roman birth declarations could be registered was appar- ently the tabularium of the governor at Alexandria. Could Roman municipalities register the birth declarations of their own citizens? This question is left open by Apuleius, who describes his wife’s declaration being deposited in a tabularium publicum—​was this the governor’s archive at Carthage or the local tabularium of the municipium of Oea?96 It is reasonable to suppose on the basis of evidence such 91. In this respect, the documentary evidence speaks against the received notion that the Roman state did not regard birth declarations as more authoritative evidence of age and status than other documents; see the references cited earlier in note 4. 92. See De magia 89.1–​4, mentioned earlier in note 82. 93. Dig. 27.1.2.1 (Modestinus 2 excus.). The interpretation of Sánchez-​Moreno Ellart 2001, 52, that Modestinus is arguing that birth declarations were not the only legitimate proof misses the point, I think. 94. An imperial rescript reassuring one female petitioner that the loss of her birth declaration did not a priori “mutilate” her status, Cod. Iust. 4.21.6 (286 CE), underscores the declaration’s perceived importance. See also Dig. 1.5.8 (Pap. quaest.), Cod. Iust. 5.4.9, and Cod. Iust. 7.16.15 (293 CE). 95. See Cod. Iust. 2.42.1 (223 CE), where a birth declaration is produced out of the archives (tabulis . . . oblatis tibi) to prove the exact age of a woman. In Dig. 22.3.29.1 (Scaevola 9. Dig.), an ex-​husband is able to vin- dicate his paternity of children who had been declared by his ex-​wife as illegitimate, which reflects the inherently weaker authority of testationes with a missing father. 96. Roman municipia and coloniae had their own tabularia publica attested in epigraphic sources; see, e.g., CIL III, 1315 = IDR 3.3 364 (region of Sarmizegetusa, Dacia) and CIL XII, 525 = ILN 3, 30 = CAG 13.4 (Aquae Sextiae, Gallia Narbonesis). 204  Anna Dolganov as the Tabula Heracleensis that Roman citizens residing in municipia and coloniae had the possibility of registering birth declarations with local magistrates, who deposited the records in the local tabularium and sent a copy to Rome or the cap- ital of the province.97 By contrast, Romans residing in non-​Roman communities apparently had to travel or send representatives to the capital. Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 How widespread was birth registration among Roman citizens? Apuleius speaks about his father-​in-​law declaring the birth of his daughter “like everyone else” (more ceterorum), while the Historia Augusta refers to Gordian registering his son’s birth “in the Roman way” (more Romano).98 The long journeys to Alexandria undertaken by Romans in Egypt are powerful illustrations of the im- portance attached to this procedure. In provincial landscapes where Roman citi- zens were a minority, a registered Roman birth declaration was probably viewed as a safeguard of a status that many felt to be precarious. One vulnerable group were the children of veterans, whose ability to acquire Roman status from their fathers could not be taken for granted and therefore had to be documented. A veteran who did not register the birth of his children could undermine their claims to Roman status and ability to inherit—​if, for example, it were alleged that they were not his children at all or were illegitimate by virtue of being born during his military service.99 In one document, a Roman freedwoman presents her young son by a deceased veteran for status verification (epikrisis) before the governor. To prove his legitimate birth, she provides his birth declaration—​ registered by the father at Alexandria—​along with the father’s epikrisis record and death certificate. Had the veteran failed to declare his son’s birth before his own death, with no census intervening, the woman is likely to have had trouble demonstrating his paternity of the child.100 Another vulnerable group were the children of Roman freedmen, for whom the absence of a birth declaration could call not only their Roman status but also their free birth into question. This sce- nario is attested in the Roman legal sources and illustrated by the case of Petronia Iusta, the daughter of a Junian Latin freedwoman at Herculaneum whose lack of a birth declaration enabled her mistress to claim that she was born into slavery.101 In a second-​century document, we encounter a Roman freedwoman who had been manumitted in the same year as the birth of her illegitimate twins. Had 97. For the procedures documented by the Tabula Heracleensis, see note 19. 98. Apul. De magia 89.2 and HA Gord. 4.8. 99. In the property dispute between the Roman widow Tertia Drusilla and the veteran Iulius Agrippinus, Drusilla faced allegations that her marriage had been contracted during her husband’s military service, which called its validity into question; see M.Chr. 372 (Arsinoite, 140s CE), with Phang 2001, 395–​401. On the Drusilla affair, see Rupprecht 2001. 100. P.Diog. 6 (Arsinoite, after 143 CE). See also M.Chr. 372 col. iv, ll. 1–​15 (115 CE), where a woman tries to convince the governor that her son’s father is a deceased soldier who had neglected to make a birth declaration. 101. At the same time, her patrona’s lack of a manumission document enabled Iusta to insist that she was free- born; see Metzger 2000 and Weaver 1997, 69–​71. See also Dig. 4.2.8.1 (Paulus, ad edict.), where official documentation (instrumentum) of an individual’s free status is destroyed to enable him to be enslaved. Documenting Roman Citizenship  205 she failed to file birth declarations for her children, she would presumably have faced similar problems demonstrating their free birth.102 In the interesting case of Sempronia Gemella, a Roman woman who was de facto married to a Greco-​ Egyptian named Socrates, registering the birth of her children as illegitimate (spurii) was a strategy for them to receive her Roman citizenship.103 Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 For auxiliary soldiers, who were not Roman citizens but could expect to be awarded citizenship after their honorable discharge from the army, an affidavit of birth (testatio, drafted in the Roman fashion in Latin on tabulae with the seals of seven Roman witnesses) was a crucial document for claiming paternity and requesting citizenship for children born during military service.104 The testationes of soldiers could be drafted locally in the army camp and apparently did not need to be registered at Alexandria—​perhaps because, unlike the testatio of Sempronia Gemella filed at Alexandria, they were not birth declarations by Roman citizens but documents anticipating future claims of citizenship. The absence of a testatio could undermine the question of paternity; in a second-​century legal case, an anxious woman appears before the prefect of Egypt, worried that her son’s lack of a birth declaration will prevent him from inheriting from his father, a deceased soldier.105 The significance of such testationes is spelled out by the soldier Marcus Lucretius Diogenes, who states his intention to demonstrate (adprobare) that the child is his natural son at his future status verification (epikrisis).106 Similarly, in- formally manumitted freedmen (the so-​called Junian Latins) could claim Roman citizenship for themselves and their children by demonstrating (adprobare) by means of a testatio of birth that they had entered a legitimate marriage with a partner of Roman or Latin status and produced a one-​year-​old child—​anniculi probatio, as the procedure is called by Roman legal writers.107 For both soldiers and Junian Latins, the adprobatio had to take place before an official with high jurisdictional powers—​the Roman praetor or provincial governor.108 For enfranchised civic elites in the provinces, birth registration was a ques- tion not only of Roman status but also of legitimacy and lineage, which deter- mined inheritance rights and membership in the officeholding class. It can be deduced from the provisions of the Augustan marriage laws that legitimate birth 102. See SB I 5217 = FIRA III 6 (Theadelphia, 148 CE). 103. See Strassi 2001, cited earlier in note 82. 104. For testationes of birth by soldiers, see BGU VII 1690 = FIRA III 5 = CPL 160 (Philadelphia, 131 CE), P.Diog. 1 = CPL 159 (Contrapollonopolis, 127 CE), and P.Mich. VII 436 = CPL 161 (Pselkis, 138 CE). 105. See M.Chr. 372 col. iv, ll. 1–​15 (115 CE), mentioned earlier in note 99. The soldier, an Alexandrian citizen, had not made a birth declaration. While recognizing the boy’s legitimacy as an heir under a military will (where soldiers were free to name whomever they pleased), the prefect informs the woman that he cannot be regarded as a legitimate son of the soldier. 106. P.Diog. 1 = CPL 159 (Contrapollonopolis, 127 CE). 107. On anniculi probatio and other procedures for Junian Latins to attain Roman citizenship, see Weaver 1997, 57–​62. See further the section on manumission in this chapter. 108. On the power of legis actio, necessary to approve claims of Roman citizenship, see the discussion of man- umission at note 118. 206  Anna Dolganov and registration in the album professionum were necessary for entry into the local senate (ordo decurionum) and other elite ordines. The same standard ap- plied to the bouleutic class of Greek cities, as we observe in the examination of ephebic candidates at Alexandria and Antinoupolis, where families presented the registered birth declarations of their sons as proof of their pedigree and legiti- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 mate birth.109 Roman citizens who were simultaneously citizens of Alexandria or Antinoupolis accordingly registered the births of their children with the civic magistrates of their Greek city and in the Roman album professionum in the pro- vincial capital.110 How common was birth registration among the general population? The only places in the empire where Roman citizenship would have been relatively wide- spread before the Constitutio Antoniniana were Italy and communities of Roman citizens (coloniae and municipia civium Romanorum) in the provinces. If it is cor- rect to think that these communities had local procedures of birth registration, one would expect birth declarations to have been a common practice. Even in Italian cities, freeborn Roman citizens were a privileged sector of the population, distinct from slaves, freedmen, Junian Latins, and migrants of peregrine or Latin status. As a parallel, one may adduce the citizens of the mētropoleis of Egypt, who had an active practice of birth registration that enabled them to secure local citizenship and fiscal privileges for their children. Judging from the behavior of metropolites, as long as Roman citizenship remained an exclusive and fiscally privileged status, one would expect Romans to have been active in registering the births of their children.111 The authority of birth declarations as proof of free birth—​which was, among other things, a requirement for service in the Roman army—​provided a further incentive for birth registration. The instructive ex- ample of P.Petaus 1–​2 (185 CE), containing two copies of a birth declaration filed by Egyptian villagers who did not possess any sort of fiscally privileged status but clearly did not want to wait until the next census to register the birth of their daughter, encourages us to think in terms of widespread birth registration among the general population.112 109. See P.Diog. 8 (Arsinoite, 217 CE), P.Oxy. III 477 = W.Chr. 142 (Oxyrhynchite, 132–​133 CE), and P.Flor. III 382 = W.Chr. 143 (Hermopolite, 223 CE). On the ephebate at Alexandria and Antinoupolis, see Nelson 1979, 47–​59; Whitehorne 1982; and Delia 1991, 71–​88. 110. See Schubert 1990, 19–​33, on P.Diog., an archive of papers belonging to a family of Antinoites with Roman citizenship. Among the documents are P.Diog. 2 (Antinoite birth certificate), P.Diog. 8 (Antinoite epikrisis), and P.Diog. 5 (Roman epikrisis). 111. On birth registration by metropolites, see Sánchez-​Moreno Ellart 2010. His assertion (116–​117) that birth declarations are never cited by metropolites in epikrisis proceedings does not seem to be correct; in at least three published epikrisis applications, parents refer to the registration of their sons in the birth registers (en epigegenēmenois): BGU I 109 (Arsinoite, 121 CE), l. 17; SB XIV 11270 (Arsinoite, 96–​97 CE), ll. 19–​20; and P.Gen. I 19 (Arsinoite, 148 CE), ll. 12–​13. It may be inferred that copies of birth declarations were submitted with the applications; see the section on epikrisis. 112. The Historia Augusta (HA Marc. 9.7) attributes to the emperor Marcus Aurelius a general requirement that the freeborn status of Roman citizens be proved on the basis of birth declarations registered in public Documenting Roman Citizenship  207 Overall, it seems likely that birth registration by Roman citizens was common throughout the empire before the Constitutio Antoniniana. While it may have been possible to register Roman children locally in Roman municipalities, Romans who resided in peregrine communities (as was the case in Egypt and elsewhere in the Greek East) apparently had to do so directly Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 in the provincial capital. From an administrative perspective, birth registra- tion “created” new Roman citizens. Consequently, one would expect copies of municipal records of Roman birth declarations to have been forwarded to the provincial capital, reviewed by officials, and deposited into the tabularium of the governor. Documenting Manumission In the Roman tradition, the manumitted slaves of Roman citizens had the possibility of acquiring citizenship. This could be achieved via three formal procedures:  manumissio vindicta (“by the rod”) in proceedings before a mag- istrate, manumissio testamento in the will of a Roman citizen, or manumissio censu whereby a Roman citizen pronounced a slave free before magistrates administering the census.113 Slaves manumitted by any other method—​by letter, for example, or informally in the presence of friends (inter amicos)—​had their freedom protected by Roman magistrates but did not become citizens.114 Under the principate, Augustan legislation further restricted the possibilities for formal manumission: the lex Fufia Caninia of 2 BCE limited the proportion of slaves in a single estate who could be manumitted in a will, while the lex Aelia Sentia of 4 CE limited formal manumission to slaves aged thirty and older who were freed by Roman masters aged twenty and older.115 The ancient sources attribute these policies to the concern of Augustus to curtail the influx of freedmen into the civitas Romana.116 While narrowing the pathway from manumission to citizen- ship, Augustan legislation did not limit the practice of manumission itself. This had the effect of expanding the category of informally manumitted freedmen, archives within thirty days. Whether or not this reflects an actual piece of imperial legislation, it is signif- icant that such an administrative requirement seemed plausible to the fourth-​century authors. 113. On the procedures of Roman manumission, see Buckland 1908, 437–​451, and Mouritsen 2011, 11–​13. On manumission as an avenue to citizenship, see Sherwin-​White 1973a, 322–​334. Manumissio censu is generally (and perhaps incorrectly) regarded as an obsolete procedure under the principate. It is briefly described by two Antonine legal writers; see Gai. Inst. 1.17, 35, 44 and 138–​140 and the Fragmentum Dositheanum 17, cited earlier in note 54. The latter text states that manumissio censu was performed ex- clusively in the city of Rome. 114. Under the republic, the freedom of informally manumitted freedmen was not recognized under Roman ius civile (see Cic. Top. 10: si neque censu nec vindicta nec testamento liber factus est, non est liber) but was protected by the praetor; see Gai. Inst. 1.22, 3.56, and Fr. Dosith. 4–​8, with Mouritsen 2011, 85–​86. Under the principate, their status became a modified form of Latin status; see notes 117 and 118. 1 15. On the Augustan manumission laws, see Mouritsen 2011, 80–​92; Treggiari 1996, 893–​897; Gardner 1991. 116. See Suet. Aug. 40; Dionysius of Halicarnassus, Roman Antiquities 4.24.4–​8; and Cass. Dio, Roman History 56.7.4–​6, 56.33.3. On the imperial vision of Augustan social legislation, see Dolganov 2021b. 208  Anna Dolganov whose legal status was assimilated to that of Latini with limitations on their transmission of property.117 Of the formal procedures resulting in citizenship, manumissio vindicta was performed before a magistrate as a fictitious trial where a third party claimed the slave without objection from the owner, thereby setting the slave free. Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 Manumissio vindicta had to be performed before a high-​ranking official with the power of legis actio—​the Roman praetor, consul, or provincial governor (or one of his legati, who could receive legis actio by delegation).118 This is exemplified by Pliny’s letter to his father-​in-​law Calpurnius Fabatus, on whose behalf Pliny had asked the incoming governor of Baetica to divert his journey to Comum so that Fabatus could manumit his informally freed slaves. Clearly, the procedure could not be performed locally but required the authority of a praetor or gov- ernor, even if the latter was not in his own province.119 Manumissiones vindicta were documented in Latin on tabulae and kept by the freedmen.120 As acts performed under the authority of a Roman official, the manumissions were pre- sumably entered into the official’s administrative records, which were stored in the archives at Rome or in the governor’s tabularium.121 In the case of Calpurnius Fabatus, one would imagine that the manumissions were documented both in the acta of the proconsul of Baetica and in the acta publica of the municipium of Comum, with copies of the record drafted on tabulae for Fabatus and his freedmen. A further level of documentation is reflected by two second-​century certificates that two freedmen manumitted vindicta have paid the 5 percent man- umission tax (vicesima libertatis). By recording payments of the vicesima tax, the Roman fiscal administration indirectly created an additional, detailed record of Roman manumissions.122 117. The assimilation of informal manumission to Latin status took place through a lex Iunia of uncertain date (see Gai. Inst. 1.29, 31; Inst. Iust. 1.5). In favor of the 17 BCE date, see Balestri Fumagalli 1985 and Weaver 1997, 58–​59. In favor of the 19 CE date, see Sherwin-​White 1973a, 328–​334. On Latini Iuniani and their prevalence, see Weaver 1990, 1997, and 1998. 118. On the authority to manumit vindicta, see the sources discussed in Buckland 1908, 453–​455. On man- umission before a consul, see Dig. 1.10.1.2 (Ulpianus 2 de off. cons.). On manumissio vindicta before a legatus of the governor, see Dig. 40.2.17 (Paulus 50 ad ed.). The emperor could dispense with the pro- cedure; his voluntas was all that was required: Dig. 40.1.14.1 (Paulus 16 ad Plaut.). In the Flavian pe- riod, manumissions performed by municipal magistrates resulted in informally manumitted freedmen of Latin status; see chap. 28 of the Lex Irnitana, with Gardner 2001b. By the third century, some municipal magistrates could apparently manumit by virtue of their legis actio (Paul. Sent. 2.36.4). 119. See Pliny Ep. 7.16. The civil jurisdiction of a proconsular governor began as soon as he left Rome (Dig. 1.16.2pr. (Marcianus 1 inst.)) and was not restricted to his province (Dig. 1.7.36.1 (Paulus 18 resp.)). 120. As described in P.Diog. 6 (Arsinoite, 142 CE), ll. 19–​22. 121. It is noted in the legal sources that manumissiones vindicta could be performed by governors in an in- formal setting, such as the baths or a private villa; see Dig. 40.2.7 (Gaius 1 rer. cott. sive aur.) and Dig. 40.2.8 (Ulpianus 5 ad ed.). Even in such cases, one would expect the manumission to be documented in the governor’s acta. For a good illustration of acta being recorded “on the go,” see the acta of a strategos of the Ombite nome preserved in P.Paris 69 = W.Chr. 41 (Elephantine, 232 CE). 122. The tax certificates are inscribed in Latin on small wooden tablets; see van Minnen and Worp 2009, 15–​18 (second to third century CE), and P.Mich. VII 462 (second century CE). One would suppose that these certificates were necessary for the procedure to be complete and for the freedmen to obtain their tabulae manumissionis. Documenting Roman Citizenship  209 Testamentary manumission likely accounted for the largest number of for- mally manumitted freedmen eligible to receive Roman citizenship. The opening of Roman wills was subject to precise rules and took place under the close scru- tiny of fiscal officials to ensure that the Roman taxes on inheritance (vicesima hereditatium) and manumission (vicesima manumissionis) were duly paid. In Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 Egypt—​and presumably in other provinces—​Roman wills had to be opened at special toll stations (stationes) for the collection of these taxes in the presence of fiscal agents.123 In a second-​century document, two agents of the procurator vicesimae hereditatium write to an imperial freedman to request his presence at the opening of a Roman will in the Arsinoite nome.124 The agents specify that the freedman’s presence is mandatory, that the procedure must take place in a specific location (presumably the statio vicesimae in the regional capital of Arsinoe), and that detailed records must be made in the freedman’s presence within ten days after opening the will. One can infer that manumissions and other provisions of the will were recorded by the fiscal agents and forwarded to the fiscal procurator at Alexandria. Precisely how testamentary manumissions were carried out is less clear. It would have been necessary to verify payment of the manumission tax125 and check the age and personal history of the slaves to determine their eligibility to receive Roman or Latin status in line with the Augustan laws. A record of the manumission could then be created, with a copy drafted on tabulae. In the provinces, one would expect this to have been a centralized procedure that took place under the authority of the governor, as was the case with manumissiones vindicta. That Roman citizens seeking to claim an inheritance had to petition the governor is evidenced by several surviving Latin requests for entry upon inheritance (agnitiones bonorum possessionis).126 A  further clue is furnished by a second-​century legal case where an official at Alexandria investigates a claim of inheritance by an Alexandrian woman.127 The official had looked for and failed to find her birth declaration in the archives at Alexandria, leading him to suspect her Alexandrian citizenship. If the testamentary claims of Alexandrians residing in the hinterland were sent to Alexandria for review, one would expect a similar level of scrutiny for the testamentary claims of Roman citizens. In a second-​century hearing before a Roman fiscal proc- urator, a freedman states that he has sailed down to Alexandria to pay his 1 23. On the procedures for opening Roman wills, see Nowak 2015, 73–​104. 124. See P.Oxy. XX 2265 (Oxyrhynchus, 119 CE), with van Minnen and Worp 2009, 20–​22. 125. See the tax payment certificates discussed earlier in note 122. On agnitiones bonorum possessionis in papyri, see Gagos and Heilporn 2001. Possibly, in Roman 126. municipalities, inheritances could be claimed through local magistrates who forwarded the records to the governor. 1 27. See P.Oxy. XVIII 2199 (Oxyrhynchus, 117–​138 CE), containing only a partial transcription. A full tran- scription and analysis of the papyrus will be published by me in a separate paper. 210  Anna Dolganov manumission tax, which may shed light on the procedure of formalizing tes- tamentary manumissions.128 As noted, the informally manumitted slaves of Roman citizens (the so-​called Junian Latins) were under some circumstances eligible to be granted citizenship, if, for example, they sponsored the construction of public buildings, served on Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 the fire brigade at Rome, or presented evidence of a child from a legitimate mar- riage with a partner of Roman or Latin status (anniculi probatio). Junian Latins could obtain citizenship exclusively from the praetor or governor—​or, in excep- tional cases, directly from the emperor.129 The procedure of anniculi probatio is documented in three tabulae from Herculaneum that were part of a dossier belonging to a Junian Latin named Lucius Venidius Ennychus. The first tabula is a testatio by Ennychus for the birth of his daughter on July 24, 60 CE. The second is a fragmentary copy of a decree of the local senate of Herculaneum (ordo decurionum) dated July 25, 61 CE. As the date coincides with the first birthday of the child, it may be deduced that Ennychus had presented his wife and daughter to the ordo to claim Roman citizenship. The third tabula is a copy of an edict by the Roman praetor on March 23, 62 CE, stating that the magistrates (duumviri) of Herculaneum have forwarded to him the ruling of the ordo that Ennychus has successfully demonstrated his legitimate marriage with a one-​year-​old child. Satisfied with the ruling, the praetor pronounces Ennychus, his wife, and their daughter Roman citizens.130 These documents demonstrate the logic of the pro- cedure. First, Ennychus’s claim was examined and approved by the local ordo on the basis of the necessary evidence—​the testatio of the daughter’s birth and pre- sumably the manumission documents of Ennychus and his wife, with guarantors confirming their identity. The results of the investigation were then forwarded to the praetor, who exercised his authority to confer Roman citizenship.131 The promotion of Junian Latins is also evidenced by three letters of Pliny to Trajan requesting the ius Quiritium for informally manumitted freedmen on behalf of their patroni. In one instance, Pliny makes the request for three Junian Latins whose patronage he has inherited from a deceased friend, presumably because the original patronus could not be present to give his consent. This also seems to be the explanation for Pliny’s other two requests: that it was not possible for the 128. See BGU II 388 = M.Chr. 91 (157–​159 CE), ll. 6–​26, with Meyer 2004, 234–​236—​a complex case where a Roman citizen dies intestate and several tabulae manumissionis drawn up by him twelve years earlier are discovered in his house. One of the freedmen brings his tabulae to Alexandria to pay the vicesima tax for himself and his children. As this was not, strictly speaking, a testamentary manumission, it presumably did not result in Roman citizenship. 129. See the later discussion at notes 132 and 133. 130. See AE 2006, 306 = TH2 5 + 99 (July 24, 60 CE); AE 2006, 307 = AE 2017, 223 = TH2 A2 (July 25, 61 CE); and AE 1959, 297 = AE 2006, 305 = TH2 89 (March 23, 62 CE), with the commentary of Camodeca 2017, 57–​84, and Weaver 1997, 68–​69. 131. The procedure was perhaps analogous for testamentary manumissions, which one would imagine were processed locally in Roman municipalities and then finalized through the praetor or governor. Documenting Roman Citizenship  211 patroni to be physically present at a hearing before the governor.132 In granting the first request, Trajan states that the grant of ius Quiritium will be duly entered into his administrative records (commentarii).133 Altogether, one may conclude that grants of citizenship to freedmen manumitted by Roman citizens were issued under the authority of the praetor Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 or provincial governor (or one of his legati) and documented in a centralized fashion, both in the acta of the official and in the records of the Roman fiscal administration by virtue of the vicesima taxes. As evidenced by the case of the freedman Venidius Ennychus, Roman municipalities could locally investigate and approve claims of Roman citizenship, but only a high-​ranking official such as the praetor could formally confer the status. In non-​Roman communities, one suspects that manumitted slaves had to approach the governor to formalize their status as Roman citizens. Documenting Enfranchisement Apart from birth and manumission, the third avenue by which individuals could enter the civitas Romana was enfranchisement—​through service in the Roman army, through an imperial grant, or by entering the ordo decurionum or holding a magistracy in a Roman municipium. Grants of citizenship to honorably discharged veterans—​those who had served in the fleet or auxiliary units (accessible to non-​citizens) or under exceptional circumstances had been transferred to a legion—​are documented by more than 1,200 surviving bronze military diplomas. These bronze tabulae, which replicated the format of Roman waxed tablets, were given to veterans as certificates of their Roman status and their right to contract a legal marriage (conubium) with one non-​Roman woman of their choosing.134 The earliest diplomas date to the reign of Claudius, who may have been the first emperor to regularize the enfranchise- ment of auxiliary veterans.135 The emergence of bronze diplomas evidences a marked intensification of administrative record keeping with regard to Roman citizenship, further reflected by the introduction of special procedures of status verification for Roman citizens and other fiscally privileged groups, discussed in the next section of this chapter. 132. See Plin. Ep. 10.104–​105, 10.5.2, 10.11.2, with Sherwin-​White 1966, 714–​715, 566–​568, 576–​578, respec- tively. If Pliny had been the original patronus, he would presumably have been able to promote his own freedmen to the ius Quiritium—​as may be inferred from Ulpian’s remark that a consul may supervise the manumission of his own slaves; see Dig. 1.10.1.2 (Ulpianus 2 de off. cons.). 133. See Plin. Ep. 10.105: referri in commentarios meos iussi. 134. On Roman military diplomas, see Eck 2003 and the edited volumes of Eck and Wolff 1986 and Speidel and Lieb 2007, as well as the useful overview of Cooley 2012, 172–​177. On the number of Roman citizens produced by the Roman army, see Lavan 2019a. On the conubium privileges of veterans, see Campbell 1978 and Phang 2001, 53–​85. 135. The earliest-​known diploma is CIL XVI, 1 (Stabiae, 52 CE), issued to a veteran of the fleet of Misenum. On the possibility of a Claudian regularization of citizenship grants to auxiliaries, see the remarks of Beutler 2007 and Lavan 2019a, 28–​31. 212  Anna Dolganov A military diploma was part of a dossier of records tracing the military ca- reer of each Roman soldier that were kept by the imperial administration.136 Such records began with the soldier’s recruitment and evaluation of eligibility and fitness for military service (probatio), followed by his formal induction into the army (signatio) and assignment to a specific unit (relatio in numeros). Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 Considering that soldiers were often sent to serve in other provinces, one would expect that records of their recruitment, probatio, and relatio were kept in the governor’s tabularium in his home province, possibly with copies forwarded to the assigned province.137 From the texts of military diplomas, one may deduce the procedures by which grants of Roman citizenship to veterans were processed and documented. Every year, the commander of each military unit had a number of soldiers who were eligible to be honorably discharged and (if relevant) to receive Roman citizen- ship. As attested in numerous documents, soldiers could request conubium with their de facto wives and Roman citizenship for children born during military service and have their names inscribed on their bronze diplomas.138 One would imagine that soldiers declared their family members to the military commander, presenting documents such as testationes of birth, after which lists of soldiers and family members were drawn up and submitted to the governor, whose staff reviewed the lists and forwarded them to Rome. At Rome, these lists would have been reviewed once again and formally approved by the emperor, who issued a collective grant of Roman citizenship and conubium privileges. The imperial grant was inscribed on a large bronze plaque with the names of the veterans (and, if relevant, their family members) and displayed in a prominent location in the civic center of Rome. A copy of each entry was then inscribed on bronze tablets, closed, and wired shut, with the names and seals of Roman witnesses. These bronze diplomas were sent to the provincial governor, whose staff compiled a record of them for the governor’s archive. The diplomas were then distributed to veterans in each military unit.139 A bronze diploma was a highly authoritative record that a veteran could present in any place in the empire where he chose to settle. As with Roman birth declarations, the authority of a military diploma derived from its being an 136. For an example of an administrative dossier documenting the career of a Roman cavalryman in the early fourth century, see Rea 1984. On documentation pertaining to the Roman army in the governor’s tabularium see Haensch 1992, 264–​276. 137. For a detailed discussion of recruitment, probatio, and documentation of these procedures, see Davies 1989,  3–​32. 138. Among many examples, see CIL XVI 24 = AE 1927, 96 = AE 2004, 389 (Egypt, 79 CE), where citizenship is granted to a fleet veteran and his wife and son. 139. On the procedures of issuing and distributing diplomas, see Eck 2003. Provincial records of citizenship grants to veterans are evidenced by epikrisis documents where confirmation of the grant is obtained from the governor’s archive; see P.Hamb. I 31 = P.Stras. V 340 (Arsinoite? 103–​107 CE), P.Diog. 5 (Arsinoite, 131–​133 CE), and SB IV 7362 (Arsinoite, 188 CE). Documenting Roman Citizenship  213 authenticated copy (descriptum et recognitum) of records that were kept at Rome and in the capital of the province where the soldier had completed his service. Its inscription in bronze bore the strongest possible message of authenticity: the bronze tabulae were issued at Rome on a single occasion and could not be al- tered or replicated.140 At the same time, when veterans presented diplomas on Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 which their wives and children were inscribed, they also presented an official confirmation of their honorable discharge by the archive keepers of the gov- ernor. It appears that, when more than one claim of Roman citizenship was at stake, additional verification was requested from the archival records of the province.141 Roman legionary veterans did not ordinarily receive diplomas, since by def- inition they had to be Roman citizens in order to enroll in the legions. It could therefore be assumed that their Roman status had been verified and confirmed during the initial recruitment (probatio). The situation was more complex for auxiliary soldiers who were transferred up to a legion, as illustrated by a peti- tion to the governor of Judaea by veterans of the legio X Fretensis who had been transferred from the fleet of Misenum to compensate for devastating losses during the Bar-​Kokhba revolt.142 Since the veterans planned to return to their native city of Alexandria, they asked the Judaean governor to confirm their honorable discharge “from this very legion and not from the fleet, so that in times of ne- cessity your confirmation may serve as official documentation [instrumentum].” Clearly, they were anxious about returning home without bronze diplomas and being suspected of failing to complete their service in the fleet. While their ini- tial recruitment would have been documented in the records of the governor of Egypt, their subsequent transfer to a legion in Judaea evidently was not. To oblige them, the Judaean governor wrote to the prefect of Egypt that their transfer and honorable discharge from the legion took place on orders from the emperor. The insistence of the veterans on official documentation (instrumentum) reflects their conviction that administrators in their home province would expect them to prove their status on the basis of official records.143 140. The non-​replicability of military diplomas is arguably implied by references to auxiliary veterans “without bronze diplomas”; see, e.g., BGU I 113 = W.Chr. 458 (after 143 CE) and BGU I 265 = W.Chr. 459 (Arsinoite, after 148 CE). This could indicate that no bronze diploma was issued or that it had been lost or destroyed. The implication was, clearly, that it could not be issued again. 141. See note 139. 142. See PSI IX, 1026 = CPL 117 = Ch.LA XXV, 784 (Caesarea, 150 CE), with Campbell 1994, 201–​202. On emergency recruitment during the Bar-​Kokhba revolt, see Eck 1999, 79–​80. 143. It is unclear why Mann and Roxan 1988, 343, assume that the standards of documentation that worried these legionaries were specific to Egypt. It is only logical to suppose that legionaries who settled away from the provinces where they had served needed to present some form of documentation of their hon- orable discharge. The problem faced by soldiers who were transferred up to a legion and returned to their home province would presumably have been the same everywhere. For soldiers who chose to settle in the province where they had served, provincial authorities could simply check their own military records, as illustrated by SB XII 11043 R = Ch.LA XI 466 R (Caesarea, 152 CE). 214  Anna Dolganov Altogether, the evidence for the enfranchisement of veterans attests to highly centralized record keeping, whereby requests for honorable discharge and Roman citizenship were examined within the military unit, then by the officium of the governor, and ultimately by the staff of the emperor—​since, in strict terms, Roman citizenship was a privilege (beneficium) conferred on veterans by the em- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 peror himself.144 It follows that the emperor’s staff had a complete record of grants of Roman citizenship to veterans and their family members in any given year. That administrators at Rome had a precise sense of the number of enfranchisements taking place through the Roman army is significant for interpreting changes in imperial citizenship policy, such as the progressive erosion of the liberal policy allowing children of veterans born during military service to receive Roman status.145 Grants of Roman citizenship could also be made by the emperor on an in- dividual basis, in response to requests from petitioners or officials and civic communities acting on their behalf.146 Several examples of viritane grants are preserved in Pliny’s correspondence with Trajan. In one exchange, Trajan grants Roman citizenship to the daughter of a centurion, whose request Pliny had forwarded.147 On another occasion, Pliny asks Trajan to grant Roman citizenship to his Egyptian doctor. Approving the request, Trajan asks Pliny to provide the doctor’s name, age, nome (evidently, his legal residence), and wealth (census) for the imperial records, after which an official letter could be drafted to notify the prefect of Egypt of the citizenship grant.148 A similar procedure is documented by an imperial letter preserved in an inscription from Banasa in Mauretania (the so-​ called Tabula Banasitana), where Marcus Aurelius and Lucius Verus respond to the governor concerning a citizenship request by a local chieftain. In granting cit- izenship to the chieftain and his family, the emperors ask the governor to provide their exact ages for the imperial records.149 Following the letter, the inscription includes an authenticated copy (descriptum et recognitum) of the relevant entry in the register of imperial grants of Roman citizenship (commentarius civitate Romana donatorum) kept in the archives at Rome. This register was evidently or- ganized chronologically by imperial reign and consular year. The entry contained the precise date and location of the grant (issued at Rome on the sixth of July), the names and ages of the Mauretanian chieftain and his family, the circumstances of the grant (private petition forwarded by the governor), and the conditions of 144. On the role of the emperor as the commander in chief of the army, see Eck 2003. The prohibition against dishonorably discharged soldiers residing in Rome and other locations with an imperial residence (Dig. 3.2.2.4 (Ulpianus 6 ad ed.)) further illustrates the centralized documentation of missiones. 145. On the privileges of veterans and their development, see Wolff 1986. On the erosion of the civitas liberorum privilege, see Waebens 2012a and 2012b. 146. On viritane grants of citizenship by the emperor, see Millar 1992 [1977], 477–​490. 147. See Plin. Ep. 10.106–​107, with Sherwin-​White 1966, 715–​717. 148. See Plin. Ep. 10.5–​7, 10.10, with Sherwin-​White 1966, 566–​571, 575–​576. 149. See IAM II 1, 94, with the commentary of Seston and Euzennat 1971. Documenting Roman Citizenship  215 the grant (Roman citizenship salvo iure gentis but without diminution of impe- rial taxes and tribute). Analogous to a military diploma, an authenticated copy of the record was drawn up at Rome on tabulae—​possibly bronze—​with the names and seals of Roman witnesses, in the case of the Tabula Banasitana by twelve members of the consilium of the emperor.150 As with military diplomas, Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 the tabulae were presumably first sent to the governor, whose staff entered the information into his records and handed the document over to the recipient. Therefore, following the general pattern of Roman state record keeping, records of individual grants of citizenship were kept in the imperial archives at Rome, in the governor’s tabularium, and in the hands of recipients, who registered these grants with local officials in their cities of residence. A further means for individuals to enter the civitas Romana was to be elected to a magistracy in a Roman municipium with the “lesser” Latin right (Latium minus) or to be admitted into the curial class (ordo decurionum) of a municipium with the “greater” Latin right (Latium maius).151 How admis- sion into the ordo of a Roman municipium functioned and was documented is suggested by evidence for the admission of Alexandrians and Antinoites into the ephebate, an institution of the civic elite. When applying for the ephebate, families brought copies of the registered birth declarations of their sons and other documents attesting to their local citizenship, family history, and suc- cessful participation in a number of civic rituals. All documentation was submitted to a Roman official—​the prefect or one of the epistrategoi—​for re- view. The families then presented themselves in a hearing before the prefect (for Alexandrians) or epistrategos (for Antinoites), who questioned ephebic candidates regarding their identity, family, residence and education.152 Other 150. In view of the prestigious context of imperial citizenship grants, it seems likely that the tabulae were bronze—​as may be implied by a reference to a diploma in another inscription documenting a viritane grant: ZPE 143, 243 = AE 1999/​00, 168 = AE 2003, 29 = AE 2007, 82 = AE 1999, 1250 = AE 2003, 1379 (Carnuntum, Pannonia, ca 151–​230 CE). 151. On the acquisition of citizenship per honorem or through adlection into the ordo of Roman municipalities, see Thomas 1996, 83–​97, and Kremer 2006, 146–​148, 169–​173. The ius Latii traditionally conferred citi- zenship only on magistrates. The Latium maius encompassing all members of the ordo was a later devel- opment, first documented for the African municipium of Gigthis, which successfully petitioned for it in the reign of Antoninus Pius; see CIL VIII, 22737 = ILS 6780. 152. On admission into the ephebate, see Nelson 1979, 47–​59, and Delia 1991, 71–​88. On the involvement of the Roman epistrategos in the examination (epikrisis) of Antinoite ephebes, see Thomas 1982, 107–​ 110. That Alexandrian ephebes were examined by the prefect of Egypt is implied by the inclusion of Alexandrians in introductory formulae of second-​century epikrisis records; see BGU IV 1033 (unknown provenance, after 117 CE) and P.Hamb. I 31 (Arsinoite? 126–​133 CE). See also P.Lond. II 260 (Arsinoite, ca 73 CE), l. 354, referring to an epikrisis of Alexandrians at the conventus, and P.Flor. III 382 (Hermopolite, 223 CE), ll. 73–​74, referring to an Antonine register of ephebes “inscribed by the prefect.” Contrary to Haensch 1992, 290–​291, and Foti Talamanca 1974, 58–​59, the involvement of Alexandrian officials does not imply that the epikrisis hearings ceased to take place before the prefect. Instead, the procedure seems analogous to the epikrisis of Roman citizens, where delegated officials received and processed the paper- work. For a record of proceedings of an ephebic examination (presumably before the prefect of Egypt), see P.Gen. II 111 = SB V 7561 (mid-​second century CE). 216  Anna Dolganov evidence suggests that admission into the curial class of a Roman municipium functioned in a similar manner. The procedure was supervised by the duoviri quinquennales, the magistrates responsible for the municipal census and album decurionum, who reviewed each candidate’s credentials, such as legitimate birth, wealth and civic status. The applications were then subject to review and Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 approval by the governor.153 A significant number of incoming decuriones were sons of officeholding families who were already Roman citizens.154 Accordingly, one would expect applications by “new men” of Latin status to have undergone closer scrutiny. To receive Roman citizenship per honorem or by entering the ordo of a municipium, individuals had to be born or adlected as citizens of the municipium.155 Adlection into the ordo was beyond the authority of Roman municipalities and required approval from the emperor. An Antonine inscrip- tion from the municipium of Tergeste praises an orator who had persuaded the emperor to allow wealthy members of local tribes to be admitted into the ordo and obtain Roman citizenship through elected magistracies.156 The inability of municipia and coloniae to confer Roman citizenship through adlection is further illustrated by the case of Dio Chrysostom, a native of Prusa in Bithynia, whose mother and maternal grandfather had received Roman citizenship and citizen- ship in the Roman colonia of Apamea from the emperor. Dio’s phrasing suggests that the imperial grant of Roman status was a prerequisite for acquiring citizen- ship in the colonia.157 In an Antonine inscription from the colonia of Saldae in Mauretania, two men with Roman names thank the emperor for permitting the ordo to adlect them as citizens of the colonia.158 Another Antonine inscription commemorates a certain C.  Valerius Avitus, former citizen of the municipium of Augustobriga, whom the emperor had allowed to obtain citizenship in the 153. On the work of the duumviri quinquennales in supervising admission into the local ordo decurionum, see the extensive treatment of Jacques 1984, 573–​618. The role of the governor in the procedure is illustrated by Plin. Ep. 79, 112, 114; see Le Teuff 2012, 204–​205. 154. On the sons of decurions, see Jacques 1984, 603–​613. 155. On adlection into the ordo, see note 153. On analogous procedures in Roman municipalities in the Greek East, see Raggi 2004a. 156. See CIL V, 532  =  ILS 6680  =  AE 1975, 423 (Tergeste, Venetia et Histria, 138–​161 CE), with Thomas 1996,  87–​89. 157. See Dio Chrys. 61.6 (Or. 41):  μὲν γὰρ πάππος ὁ ἐμὸς μετὰ τῆς μητρὸς τῆς ἐμῆς παρὰ τοῦ τότε αὐτοκράτορος φίλου ὄντος ἅμα τῆς Ῥωμαίων πολιτείας καὶ τῆς ὑμετέρας ἔτυχεν, ὁ δὲ πατὴρ παρ’ ὑμῶν, with Raggi 2004a, 61–​63. Raggi argues that the reception of honorific citizenship in Apamea by Dio’s fa- ther, Pasikrates, implies that he was already a Roman citizen. By contrast, Blanco-​Pérez 2015, 146–​148, interprets the passage as evidence for an independent local citizenship of Apamea being bestowed on a peregrinus. The notion of a separate, non-​Roman citizenship in a Roman colonia makes little sense in Roman juridical terms. The two other examples cited by Blanco-​Pérez (OGIS 567 [Attalea, 138 CE] and I.Magnesia 192 [Magnesia, 161–​169 CE]) appear to show Roman coloniae granting citizenship to individuals who were already Roman citizens and hence do not support this argument. Such grants by no means excluded that the coloniae still had to ask for imperial permission to confer their citizenship; see the documents cited in notes 158 and 159. 158. See CIL VIII, 20682 = ILS 06875 (Saldae, Mauretania Caesariensis, 152 CE). Documenting Roman Citizenship  217 colonia of Tarraconensis.159 It is unclear whether these men needed imperial per- mission because they were not, in fact, Roman citizens or whether (as seems more likely) citizenship in Roman coloniae was generally subject to central over- sight by imperial authorities at Rome. Altogether, the evidence clearly shows that Roman coloniae and municipia did not have the power to create new cit- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 izens through adlection without permission from the emperor. As in the case of the Mauretanian chieftain and Pliny’s Egyptian doctor, these imperial grants would presumably have been documented in the commentarius civitate Romana donatorum at Rome, in the records of the governor, and in the local records of the city, with authenticated copies kept by the recipients.160 It is opportune to conclude this section on enfranchisement with several well-​ known instances where illegitimate claims of Roman citizenship came to the no- tice of Roman officials. In 46 CE, it was reported to the emperor Claudius that the municipium of Tridentum in Cisalpine Gaul had illicitly adlected members of neighboring Alpine tribes, enabling them to acquire the legal status (origo) of Roman citizens.161 Since many years had passed and some were serving in the praetorian guard and the quinque decuriae of judges at Rome, Claudius issued an edict granting the individuals in question Roman citizenship.162 The emphasis on longa usurpatio and intermarriage suggests that the adlection had taken place several decades earlier, possibly during the initial phase after the foundation of the municipium under Caesar or Augustus. The timing of the discovery is signifi- cant, as it coincides with Claudius’s empire-​wide census populi Romani, described by Claudius in the Senate as a “laborious enterprise” (opus arduum) intended to “thoroughly investigate and make publicly known the resources of our em- pire” (ut publice notae sint facultates nostrae exquiratur).163 A few years earlier, Claudius had appointed a senatorial commission to supervise the collection of 159. See CIL II, 4277 = ILS 6943 (Tarraco, Hispania Tarraconensis, 138–​192 CE). Raggi 2004a, 60, follows De Ruggiero 1921 in supposing that this was a Roman citizen who required imperial permission to change his origo to a Roman colonia. 160. The skepticism of Brunt 1971, 171–​172, that “the Roman government would probably not have controlled grants of the local franchise by a Latin city” (with reference to the municipalization of Gallia Cisalpina) seems unwarranted. If Roman officials at Alexandria systematically verified claims of metropolite citizen- ship and Hellenic status in the hinterland of Egypt, it is difficult to imagine that Roman officials would “probably not have controlled” grants of citizenship by Roman municipia. 161. See the so-​called Tabula Clesiana, CIL V 5050 = ILS 206 = AE 1983, 445 = AE 2014, 45 = AE 2014, 60 (Cles, Venetia et Histria, 46 CE), with Chastagnol 1987, 21–​23, and Ando 2017a, 125–​129. 162. The citizenship grant did not extend to the three tribes but only to the adlected tribesmen. As suggested by Ando 2017a, the mingling of colonists with the surrounding population and their subsequent legiti- mation through legal fictions and grants of citizenship may have been part of the Roman colonial agenda. As parallel examples, one may adduce Tac. Hist 3.34 on Cremona (adnexu conubiisque gentium adolevit) and Domitian’s pronouncement attached to the lex Irnitana legitimizing past transgressions of conubium; see Mourgues 1987. 163. See the so-​called Tabula Lugdunensis, which records a speech delivered by Claudius in the Senate re- garding the ius honorum of Gallic elites in the context of his imperial census; CIL XIII, 1668 = AE 1955, 115 = AE 1975, 612 = AE 1983, 693 = AE 2003, 41 (Lugdunum, Gallia Lugdunensis, 48 CE), with Riess 2003 and Le Teuff 2012, 21, 33, 60–​63, 379, 439. 218  Anna Dolganov arrears due to the fiscus.164 The Claudian imperial census and fiscal review pro- vide a plausible context for the origo of the municipes of Tridentum being closely scrutinized, revealing that a number of them did not have solid claims of Roman status going back to a legitimate source. Although the episode of the Tridentini is often cited as evidence of the Roman state’s incapacity to detect fraudulent Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 status claims,165 it arguably demonstrates something quite different: the interest of Roman administrators to trace status claims back to their source and their ability to do so, even in a remote Alpine municipium. That the Roman state had a growing interest in policing membership in privileged status groups is reflected by the emergence of centralized procedures for the verification of status in the reign of Claudius, discussed in the next section. Two further cases of illegitimate status involved the recruitment of non-​ citizens into military units of Roman citizens. During Pliny’s governorship of Bithynia, a recruitment officer found two slaves among the new recruits. The irregularity was apparently discovered quickly, before the slaves were distrib- uted into units.166 In response to Pliny’s inquiry, Trajan wrote that, if the slaves had volunteered for the army, they should suffer capital punishment; if they had been put forward in the place of others, the latter should be punished; if, how- ever, the slaves were drafted, the blame fell on the officials who vetted their eli- gibility (lecti sunt, inquisitio peccavit). Whether this was a case of error or fraud by the recruitment officials is not revealed. Some years later, in 119, Hadrian issued an edict granting Roman citizenship to members of the praetorian guard whose Roman status was defective—​either because they had been drafted by officials who failed to scrutinize their family history (ex dilectu probati parum examinata origine parentum) or because they had been transferred from other units.167 The deployment of the praetorians in Trajan’s Parthian campaign, with likely emergency levies in situ, suggests a plausible context for indiscriminate recruitment and transfers of troops from units of non-​citizens. Interestingly, Hadrian presents his citizenship grant as a preemptive remedy against future legal problems, evidently anticipating that the defective status of the praetorians would be detected by administrators. The preservation of the edict on three bronze diplomas shows that their function, like the instrumentum requested by Alexandrian oarsmen transferred to a legion in Judaea, was to provide these veterans with certification of their Roman citizenship.168 Unlike the incident 164. See Cass. Dio 60.10.4, with Le Teuff 2012, 258, 380. It is relevant that the adlection of Alpine tribesmen at Tridentum was reported to Claudius along with a series of fiscal infringements, including encroachment on imperial land, which suggests that the information came to light during a fiscal review of the region. 165. See, e.g., Brunt 1971, 208: “the fact that in the meantime they had not only enjoyed local civic rights in the municipium of Tridentum but had even been enrolled as iudices at Rome shows how negligently local records were kept and how easily they were taken at face value by the central government.” 166. See Plin. Ep. 10.29–​30. Whether the recruits were intended for legions or auxilia is debated, but legions seem more likely; see the remarks of Sherwin-​White 1966, 599–​600. 167. See Eck, Pangerl, and Weiß 2014a and 2014b. 168. See notes 142 and 143. Documenting Roman Citizenship  219 reported by Pliny, Hadrian’s edict does not illustrate status fraud within the Roman army; on the contrary, it appears to anticipate procedures of verification by Roman authorities. Status Verification of Roman Citizens (Epikrisis = Recensio, Probatio) Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 In the mid-​first century CE, papyrological sources attest to the introduction of generalized procedures of status verification, termed epikrisis, which required all male members of fiscally privileged groups to prove their status on the basis of official documents.169 The semantics of the term are significant: in classical and Hellenistic sources, epikrisis is a rare word for discernment or judgment, em- ployed in particular with reference to arbitration proceedings.170 This connota- tion prompted the use of epikrisis to render the Latin arbitratus in the Greek translation of the customs law of Asia in the early principate.171 However, epikrisis emerges with a different meaning in Roman administrative documents from Egypt. In a letter drafted during the census of 4 BCE, the Roman prefect requests complete lists of priests with documentation of their status, “so that I may verify/​ confirm [epikrinō] them among those of the twenty-​fifth year of Caesar.” Here, epikrinō clearly signifies a review of members of a particular status group.172 In the early decades of the first century, epikrisis is attested in papyri with refer- ence to the provincial census and verification of the fiscal status of land.173 This suggests that epikrisis was linked to the Latin census or recensus, a meaning that accords with the phrasing of the Augustan prefect’s letter.174 A  further clue is offered by Suetonius’s description of Caesar’s revision of the lists of Romans el- igible for the grain dole as a recensio, which corresponds to the use of epikrisis for the verification of eligibility for the grain dole in numerous documents from 169. On the papyrological evidence for epikrisis of fiscally privileged groups, see the literature cited earlier in note 86. The implications of this evidence for Roman administrative practice beyond Egypt have, to my knowledge, never been assessed. 170. See IG V2 6 = IPArk 3 (Tegea, fourth century BCE), ll. 19, 50, where ἐπίκρισις appears to be synonymous with κρίσις. See also IG XI4 1052 (Delos, third century BCE) ll. 24–​31, and I.Eph. 4 (297–​296 BCE), l. 6, with reference to the judgments of arbitrators. In Ptolemaic Egypt, see P.Hib. II 197 (Hibeh, mid-​third century BCE) with reference to an ἐπίκρισις γῆς, and P.Rev. Laws (Arsinoite? ca 259 BCE) col. xxviii, ll. 6–​8, referring to the sealed record of a judgment (ἐπίκρισις). See also Hipparchus of Nicaea (second cen- tury BCE), cited in Strabo 1.1.12, where ἐπίκρισις refers to the discernment of astronomical phenomena. 171. See epikrisis for arbitratus in the Augustan clauses of the lex portorii Asiae: SEG 39, 1180 (75 BCE–​62 CE), ll. 69 (clearly, a rendition of boni viri arbitratu), 102, 107, 110, 124, with the remarks of Cottier et al. 2008, 144–​145. See also epikrisis for a decretum of Claudius in IG XII4 1, 87 (41–​54 CE). 172. See BGU IV 1199 (Herakleopolite, 4 BCE). 173. In P.Oxy. II 288 (Oxyrhynchite, 25 CE), ll. 35–​42, and O.Brux. 14 (Thebaid, 38–​39 or 42–​43 CE), epikrisis refers to the census. The review of eligibility for fiscal privileges ordered by the prefect in BGU IV 1199 (Herakleopolite, 4 BCE) likewise takes place in the context of the census. See further CPR XV 6 (Arsinoite, 16 CE), SB I 5240 (Arsinoite, 17 CE), and BGU III 915 (Arsinoite, ca 49–​54 CE), where epikrisis refers to the status of land being earmarked for review (ἐν ἐπ̣ι̣[κ]ρ̣ίσει τετάχαμεν). 174. The Greek text of the letter strongly suggests a Latin original, which may have had probare or recensere. The expression ὅπως ἐν τοῖς  .  .  .  ἐπικρίνῳ suggests a Latin expression with inter + censere; see, e.g., Ovid Epist. ex. Pont. 1.2.137–​138: hanc probat et primo dilectam semper ab aeuo est inter comites Marcia censa suas. 220  Anna Dolganov third-​century Oxyrhynchus.175 It appears, therefore, that epikrisis in Roman ad- ministrative terminology was linked with recensio, connoting a review and veri- fication of status and eligibility for privileges. Whereas in the early first century, epikrisis is only sparsely attested in papyri, from the mid-​first century onward epikrisis and its variants (epikritheis, epikrithenai, epikrinomenos, epikekrimenos) Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 emerge as standard technical terms for procedures of status verification.176 In addition to recensio, Roman administrative use of epikrisis also appears to have been linked with the Latin probatio, as suggested by late-​first-​and second-​ century documents where the probatio of military recruits is rendered in Greek as epikrisis.177 I will return to the significance of this point.178 Distinct from the census, the procedures of epikrisis investigated the source of one’s status, in some cases tracing family histories back to the earliest holders of the status, documented up to two centuries earlier.179 These procedures were applied to all major fiscally privileged groups:  Roman citizens, citizens of Greek cities (astoi), citizens and Hellenic class of the mētropoleis, as well as Egyptian priesthoods and professional associations. The first clear evidence for a generalized epikrisis dates to the first year of the reign of Nero, attested in a Flavian administrative document from the Arsinoite nome where a number of metropolites in their forties, fifties, and sixties are referred to as “epikekrimenoi in the first year [of Nero].”180 By contrast, individuals who had undergone epikrisis in subsequent years of Nero are referred to as “epikekrimenoi in their fourteenth year.”181 This indicates that status verification was instituted for metropolites in the Arsinoite nome in the first year of Nero and was initially applied to all male metropolites and subsequently only to boys entering their fourteenth year. This accords with the evidence for the epikrisis of metropolites elsewhere in Egypt, all of which postdates the beginning of Nero’s reign. 175. See Suet. Div. Iul. 41 (recensionis causa, qui recensi non essent) and analogous use of epikrisis in Oxyrhynchite grain-​dole applications: P.Oxy. XL 2892–​2899, 2902–​2903, 2908, 2013, 2927, 2931–​2932 (Oxyrhynchus, 269–​275 CE). 176. In P.Oxy. II 288 (Oxyrhynchite, 25 CE), ll. 35–​42, and O.Brux. 14 (Arsinoite, 38–​39 or 42–​43 CE), epikrisis refers to the census. By contrast, in P.Congr.XV 13 (Arsinoite, mid-​first century CE), SB XX 15210 (Oxyrhynchus, 77 CE), P.Oxy. VII 1028 (Oxyrhynchus, 86 CE), and later papyri, epikrisis and its variants refer specifically to status verification. 177. The link between the Roman administrative meaning of epikrisis and probatio was already noted by Wilcken 1912, I: 197, 395–​396, on the basis of P.Oxy. I 39 = W.Chr. 456 (Oxyrhynchus, 52 CE), BGU I 142 = W.Chr. 455 (unknown provenance, 159 CE), and BGU I 143 = W.Chr. 454 (Arsinoite? 159 CE). 178. It is not possible to determine whether epikrisis was employed with the same administrative meaning in other eastern provinces, where papyrological evidence is lacking. In Roman-​period Greek inscriptions, epikrisis remains rare and on occasion stands for the Latin arbitratus, arbitrium, and decretum in line with its earlier meaning of “judgment” in Hellenistic sources (see note 171, as well as SEG 49, 1676 [Sardis, 189–​190 CE], and the bilingual text of I.Eph. 43 = AE 1906, 30b [375–​378 CE]). 179. See, e.g., P.Oxy. XVIII 2186 (Oxyrhynchus, 260 CE), a third-​century epikrisis application where Greek gymnasial status is documented for every generation back to the Augustan age. 180. See P.Lond. II 260 (Arsinoite, ca 73 CE), ll. 553–​576. 181. See P.Lond. II 260 (Arsinoite, ca 73 CE), ll. 507–​526. Documenting Roman Citizenship  221 As documented by papyri, Roman procedures of status verification can be described as follows. For metropolites, epikrisis was required for all males and male slaves in their fourteenth year.182 It was necessary to submit to local officials evidence of the metropolite status of both parents, typically in the form of extracts from pop- ulation registers of the city quarter (amphodon) in which the parents were registered Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 during the most recent census.183 It was also necessary to submit documentation of birth, likewise in the form of an extract from the amphodal register in which the parents had registered their son during the last census or through a separate birth declaration.184 Similar documentation was presented in the epikrisis of slaves, who were declared by metropolites in the census and registered in the amphodon.185 In turn, freedmen and freedwomen of metropolites seeking to register their sons had to demonstrate the source of their status by bringing evidence of the metropolite status of their male patrons or fathers of female patrons. In a second-​century epikrisis ap- plication, a metropolite freedwoman named Dionysous cites the registration of her son in the amphodon with evidence of the metropolite status of her late husband and of the father of her patrona in the amphodal register.186 For the Hellenic class of the mētropoleis, the standards of documentation were more elaborate. In addition to demonstrating the Hellenic status of both parents (again, on the basis of the amphodal registers), applicants had to pro- vide evidence of Hellenic status for all male ancestors on each side of the family, tracing both genealogies back to registers of Hellenes compiled in the reign of Vespasian.187 As further proof of status, families adduced birth declarations and 182. This is made explicit in epikrisis applications, where metropolites state that they are acting “in accordance with orders regarding the epikrisis of those entering the age of thirteen if they are metropolites taxed at the rate of twelve drachmas”; see, e.g., W.Chr. 217 (Oxyrhynchus, 172–​173 CE), ll. 5–​8. 183. See, e.g., SB IV 7440 (Hermopolite, 132–​133 CE) and SB XIV 11270 (Arsinoite, 96–​97 CE). On re- gional variations in the format of epikrisis applications, see Kruse 2013, 319–​322. In the Arsinoite nome, metropolites tended to submit particularly extensive paperwork, as illustrated by BGU I 109 (121 CE), documenting the registration of both parents in the amphodal lists for every census during their lifetimes, as well as the registration of the grandparents before the birth of the parents. 184. See, e.g., BGU I 109 (Arsinoite, 121 CE), ll. 15–​17; SB IV 7440 (Hermopolite, 132–​133 CE), ll. 39–​42; and SB XIV 11270 (Arsinoite, 96–​97 CE), ll. 14–​17, where parents cite their registration of their son in the amphodon during the last census. Separate birth registration is likewise mentioned in a number of epikrisis applications (see note 111). 185. See, e.g., W.Chr. 27 (Oxyrhynchus, 172–​173 CE), where a metropolite registers his thirteen-​year-​old houseborn slave and provides evidence of his own metropolite status. See further BGU I 324 = W.Chr. 219 (Arsinoite 166–​167 CE), where a metropolite woman registers two slaves (apparently not houseborn) whom she had declared in the last census. As supporting evidence, she submits a copy of the successful epikrisis of another slave several years earlier. 186. See P.Oxy. III 478 (Oxyrhynchus, 132 CE), annotated in a different hand (ll. 15a, 28a), which appears to note inconsistencies between the application and the statements made by Dionysous during her epikrisis hearing. This document illustrates the Roman principle of manumission as an avenue to citizenship being applied to the local citizenships of provincial towns. 187. On the requirement that ancestry be traced back to the Vespasianic registers, see Ruffini 2006, 75–​78. See, e.g., P.Oxy. X 1266 (Oxyrhynchite, 98 CE), SB XX 14111 (Arsinoite, after 161 CE), and P.Amh. II 75 (Hermopolis, 161–​168 CE). Some families could document their ancestry back to the registers compiled under Nero and Augustus (see Kruse 2013, 314–​316), which was not a strict requirement but 222  Anna Dolganov records documenting the marriages of parents and grandparents.188 In families with multiple children, evidence of the successful epikrisis of older sons was typ- ically provided.189 It is clear that applicants submitted authenticated copies of the records enumerated in their epikrisis applications.190 For both metropolites and Hellenes, the procedures of epikrisis culminated in a hearing where applicants Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 were examined by local officials. The proceedings were documented in local ad- ministrative records, a copy of which was kept by the applicant and his family. This information was transmitted to tax officials, who entered the name of the epikekrimenos into their tax lists as paying the relevant reduced tax rate.191 Quite remarkably, the epikrisis records of each mētropolis were forwarded to Alexandria, where officials systematically reviewed them and, if necessary, rejected the claims or sent them back for additional verification.192 In the Greek cities of Alexandria and Antinoupolis, status verification is attested for boys in their fourteenth year in connection with the ephebate.193 Like the Hellenes of the mētropoleis, ephebic candidates submitted extensive paper- work, which always included an authenticated copy of a registered birth decla- ration, evidence of the father’s ephebate and marriage, birth declarations, and (if applicable) evidence of the ephebate of all male siblings. All of this information was certified by three guarantors present at the hearing.194 In both Alexandria and Antinoupolis, ephebic applications were processed by civic magistrates, while hearings took place before the governor (for Alexandrians) or the epistrategos (for Antinoites), who certified the applicant’s examination.195 Despite an additional prestige factor. In the mid-​third century CE, Hellenic families could still trace their ancestry back to the Augustan lists; see note 179. 188. For references to birth declarations, see the documents cited earlier in note 187, with Kruse 2013 and Nelson 1979, 26–​39. For references to documentation of marriages of parents and grandparents, see, e.g., P.Oxy. II 257 (Oxyrhynchus, 94–​95 CE), ll. 25–​27, 30–​32, and P.Oxy. X 1266 (Oxyrhynchus, 98 CE), ll.  15–​20. 189. For references to the epikrisis of older sons, see SB IV 7440 (Hermopolite, 132 CE) and SB XIV 11270 (Arsinoite, 96–​97 CE). 190. See P.Amh. II 75 (Hermopolis, 161–​168 CE), containing verbatim citations of entries in the amphodal registers, as well as P.Oxy. X 1266 (Oxyrhynchus, 98 CE), ll. 15–​20, which refers to a marriage contract registered in the katalogeion at Alexandria. See also BGU XI 2086 (234–​235 CE), a copy taken from ar- chival records of epikrisis. 191. In tax lists, it is specified that individuals have been “verified [epikritoi] for the payment of twelve drachmas”; see, e.g., P.Kron. 1 = P.Mil.Vogl. II 81 = SB VI 9394 (Arsinoite, 123 CE). In other adminis- trative lists, we find references to anepikritoi (see, e.g., P.Berl.Cohen 17 [Arsinoite, second century CE], P.Lond. II 259 = W.Chr. 63 [Arsinoite, 94–​95 CE], and P.Lond. II 260 [Arsinoite, ca 73 CE], ll. 626–​631), which implies that the other individuals listed had been verified. 192. See PSI X 1109 (Oxyrhynchus, 93–​94 CE), where the status of a thirteen-​year-​old metropolite is questioned by an official at Alexandria (eklogistes), who orders the local strategos to obtain additional proof of status from both parents. The document is discussed by Kruse 2013, 328–​331. 193. On the ephebate at Alexandria and Antinoupolis, see Nelson 1979, 47–​59, and Delia 1991, 71–​88, discussed at notes 109 and 152. 194. See, e.g., P.Flor. III 382 (Hermopolite, 223 CE) and P.Diog. 8 (Arsinoite, early third century CE) for Alexandria and Antinoupolis, respectively. 195. On the involvement of Roman officials in supervising the examination of ephebes, see note 152. This has important implications for our general understanding of the ephebate in Greek cities under Roman Documenting Roman Citizenship  223 an apparent lack of published evidence for the epikrisis of slaves and freedmen of Alexandrians and Antinoites, their fiscal immunity implies that they, too, were subject to procedures of status verification, on par with the slaves and freedmen of metropolites and Roman citizens. A second-​century document shows a local official processing slave sales by Egyptian provincials to Alexandrians. The of- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 ficial took a money deposit until an authenticated record of the registered sale (chrēmatismos) was submitted to him, “in line with other cases of slaves sold to individuals immune from taxes.”196 Clearly, changes in the fiscal status of slaves had to be registered with Roman provincial authorities and the registra- tion reported to local officials before the fiscal immunities could be claimed. If the freedmen of astoi could acquire citizenship through manumission, as some evidence suggests, one would expect them to have been subject to an analogous standard of status verification to that of Roman freedmen.197 The epikrisis of Roman citizens is documented in approximately two dozen papyri, the majority of which are records of epikrisis proceedings before the Roman prefect of Egypt.198 The date of the hearings and references to a dialogismos indicate that epikrisis frequently took place at the governor’s assizes.199 The epikrisis records are prefaced by prescripts indicating that the procedure applied to veterans, freeborn Roman citizens, and their freedmen and slaves.200 In line with these prescripts, epikrisis documents show veterans, freeborn Romans and Roman freedmen registering themselves and their sons and male slaves with the governor. The procedure involved submission of documentary evidence of Roman status to a military officer (typically a tribunus legionis) appointed by the governor, followed by a hearing in the presence of three guarantors who vouched rule—​as, in effect, a curial class monitored by the Roman state. As a parallel, one may adduce the epikrisis of priests, conducted before the archiereus, who was a Roman procurator; see Kruse 2019, 126. 196. See P.Thmouis (Mendesian, 180–​192 CE) col. cxix, ll. 1–​23. 197. That the Roman state established a pathway from manumission to citizenship for the freedmen of astoi is suggested by P.Hamb. IV 270 (Alexandria, second to third centuries CE), where the freedwoman of an astos describes herself as an astē. In the Roman fiscal rulebook of the Idios Logos, astoi and their freedmen are subject to the same rules concerning marriage and inheritance; see BGU V 1210 (Arsinoite, 149–​161 CE), ll. 38–​41, 50, 133–​135. 198. On the epikrisis (status verification) of Roman citizens, see Foti Talamanca 1974, 56–​68, and Haensch 1992, 290–​293. The Roman epikrisis proceedings listed in Haensch 1992, 313–​317, can be supplemented with BGU IV 1032 (Arsinoite, 173 CE), P.Diog. 6 (Arsinoite, after 143 CE), and P.Oxy. LVIII 3920 (Oxyrhynchite, early third century CE). 199. That many epikrisis hearings date to the early spring months when the prefect was typically touring Middle Egypt was already noted by Wilcken in his commentary to W.Chr. 458–​459. For references to epikrisis taking place at the conventus (ἐν διαλογισμῷ), see SB VI 9227–​9228 (Syene, 161 CE) and P.Lond. II 260 (Arsinoite, ca 73 CE), l. 354. In a private letter, a certain Ptolema (presumably a Roman or Alexandrian) urges her brother to sail to the ongoing assize of the prefect “so that, if at all possible, we may get the little one examined [epikreinai]” (P.Hamb. I 86, Arsinoite, second century CE). 200. See, e.g., the prescript in BGU I  113  =  W.Chr. 458 (unknown provenance, ca 140 CE), ll. 1–​7:  οἱ ὑπογεγραμένοι οὐετρανοὶ  .  .  .  ὁμοίως δὲ καὶ Ῥωμαῖοι καὶ ἀπ̣ελ̣[εύθ]ερο[ι] καὶ δοῦλ[ο]ι καὶ ἕτεροι. In one instance, we find that the epikrisis of veterans was documented in a separate roll of records from the epikrisis of other Romans; see BGU I 165 = W.Chr. 459 (Arsinoite, 148 CE) and SB I 5217 = FIRA III 6 (Theadelphia, 148 CE), with Foti Talamanca 1974, 65, and Haensch 1992, 292. 224  Anna Dolganov for the identity of the declarant. For auxiliary veterans undergoing epikrisis, it was customary to present a bronze diploma as proof of Roman citizenship.201 In two instances where veterans present diplomas on which their children are inscribed, they provide letters from the tabularii of the governor confirming their honorable discharge.202 While no epikrisis proceedings of Roman legionary vet- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 erans or their children have survived, it may be presumed that legionaries, like all Roman citizens, provided evidence of their status (birth declaration, military probatio, or honorable discharge) and, if relevant, their conubium privileges.203 In the epikrisis of freeborn Roman children, fathers presented the child’s reg- istered birth declaration (professio) with evidence of their own Roman status in the form of an epikrisis document.204 In the epikrisis of illegitimate children, the mother presented the child’s testatio of birth together with evidence of her own status—​for freeborn Roman women a registered birth declaration (professio liberorum), for freedwomen a tabula manumissionis together with evidence of the Roman status of the patronus or patrona in the form of an epikrisis docu- ment or birth declaration, respectively.205 In the epikrisis of houseborn slaves, the master or mistress presented the slave’s birth declaration (oikogeneia) together with a census declaration mentioning the slave—​to show that the slave had been declared and the relevant taxes paid to Roman authorities in good faith.206 It is reasonable to suppose that in the epikrisis of slaves purchased by Roman citizens, a copy of the contract of sale was submitted in place of the birth declaration.207 201. See, e.g., BGU I  265  =  W.Chr. 259 (Arsinoite, 148 CE) and BGU III 780 (Arsinoite, 154–​159 CE). By contrast, in SB IV 7362 (Karanis, 188 CE), an auxiliary veteran presents a letter from a former prefect attesting to his honorable discharge—​either because no bronze diploma had been issued or because the diploma had been lost or damaged; see note 140. 202. See P.Hamb. I  31  =  P.Stras. V 340 (Arsinoite? 103–​107 CE) and P.Diog. 5 (Arsinoite, 131–​133 CE), discussed earlier at notes 139 and 140. A waxed tablet where the prefect of Egypt confirms the honorable discharge of an auxiliary veteran may have served a similar purpose; see W.Chr. 457 = CPL 113 (Arsinoite, 122 CE), with Mann and Roxan 1988. See also SB IV 7362 (Karanis, 188 CE), discussed in note 201. 203. For documents attesting to the honorable discharge of Roman legionaries, see the evidence discussed in Mann and Roxan 1988, with my reservations expressed earlier in note 143. P.Mich. VII 432 = CPL 105 (unknown provenance, first century CE) is a witnessed copy of an imperial edict granting citizenship and conubium privileges to the legio XXII Deioteriana with the individual’s name below. That epitaphs for soldiers often refer to the date of their probatio suggests that documentation thereof was kept by the soldiers. 204. See, e.g., SB VI 9227–​9228 (Syene, 161 CE) and BGU III 847 = W.Chr. 460 (Arsinoite, after 176 CE). 205. See P.Oxy. XII 1451 (Oxyrhynchus, 175 CE), where a freeborn Roman woman registers her illegitimate children and provides her own professio. See also SB I 5217 = FIRA III 6 (Theadelphia, 148 CE), where a freedwoman registers her illegitimate son and provides her own tabula manumissionis with the epikrisis document of her patronus, whereas in P.Diog. 6 (Arsinoite, 143–​161 CE), a freedwoman provides her ta- bula manumissionis with the birth declaration of her patrona. 206. See BGU IV 1033 (unknown provenance, after 117 CE), PSI V 447 (Oxyrhynchus, 166–​167 CE), and P.Oxy. XII 1451 (Oxyrhynchus, 175 CE). See also SB III 6995 (Memphite, 124 CE), an authenticated copy of an oikogeneia by a Roman citizen (with payment of the relevant tax, aparchē) from the archival records, presumably from the central archive of the mētropolis, corrected in red ink by the archivist. Failure to de- clare houseborn slaves warranted confiscation; see note 37. 207. For slave sales by Roman citizens, see, e.g., P.Freib. II 8 (Arsinoite, 143 CE), where shares of inherited slaves are sold among siblings and co-​heirs. It is specified that the slaves were born at Alexandria but not that they were houseborn. Presumably, this means that they had been purchased by the testator. Clearly, Documenting Roman Citizenship  225 Were the procedures of status verification (epikrisis = probatio) a requirement for Roman citizens? This certainly seems to be implied by the prescript that “vet- erans, Romans, and their freedmen and slaves” undergo epikrisis “on orders [ex enkeleuseos] from the prefect.” For metropolites, epikrisis was a prerequisite for the enjoyment of fiscal privileges: if for whatever reason the epikrisis was not performed Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 in the fourteenth year, metropolites were automatically entered into the list of laographoumenoi and taxed at the full rate.208 That Romans appear as epikekrimenoi and anepikritoi in fiscal records209 and consistently present their children and slaves for epikrisis prior to the age of fourteen210 seems to indicate that a similar standard was in place. The epikrisis of Roman slaves offers perhaps the clearest indication that epikrisis was linked to the enjoyment of fiscal immunity. This constituted a strong incentive for Romans and their family members to undergo the procedures of status verification. To what extent can the evidence for status verification in Roman Egypt be generalized for the rest of the empire, where comparable documentation has not survived? The apparent link between these procedures and the imperial census and fiscal reforms of Claudius, along with the contemporaneous emergence of bronze military diplomas, certainly suggests a broader imperial reform ca 48 CE.211 One would not expect the introduction of sweeping measures for the verification of Roman citizenship and other fiscally privileged statuses to have been limited to one province. Instead, one would presume that similar procedures were implemented elsewhere, with adjustment for regional variations such as the inclusion of both men and women in provinces where both were liable to the poll tax. What was the terminology of status verification beyond Egypt? As noted, papy- rological sources suggest that the term epikrisis was linked with the Latin recensio and probatio, illustrated by epikrisis/​recensio with reference to verification of eli- gibility for the grain dole and several papyri where epikrisis refers to the probatio of military recruits.212 The civilian and military procedures of status verification the will and subsequent contract of sale would be relevant for the epikrisis of the male slave once he entered his fourteenth year. 208. See BGU II 562 = W.Chr. 220 (Arsinoite, reign of Trajan), where a family of Hellenic katoikoi pursues the status examination of one of their sons who had been listed as anepikritos (“unverified”) and added to the list of laographoumenoi; see Kruse 2013, 326–​328. 2 09. See note 191. 210. The only exception in the surviving evidence appears to be SB I 5217 = FIRA III 6 (Theadelphia, 148 CE), where a freedwoman manumitted in the same year as the birth of her illegitimate twins registers her adult son at the age of twenty. One wonders, in this case, whether the requirements for illegitimate children were different or whether the woman had evaded registration for other reasons—​for example, to await the death of her patronus, who may have been the father of the child. In P.Oxy. XII 1451 (Oxyrhynchus, 175 CE), the supplement δέκα τρ]ε̣ιῶν is preferable to εἴκοσι τρ]ε̣ιῶν in the editio princeps at line 31, which brings the document in line with the broader pattern of children being registered before the age of fourteen. 211. A link between the emergence of bronze diplomas and the Claudian imperial census of 48 CE is likewise posited by Beutler 2007. 212. See notes 175 and 177. 226  Anna Dolganov were in many respects analogous, involving examination of documentary ev- idence for status back to its legitimate source, described in Latin as inquisitio and examinatio of the origo parentium and so forth.213 Probatio and adprobatio were technical terms for the examination of citizenship claims by Junian Latins as well as army veterans seeking Roman status for children born during mili- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 tary service.214 While one cannot generally expect to find references to routine administrative procedures in epigraphic sources, dozens of military inscriptions mention the year of a soldier’s probatio, underscoring the importance of the pro- cedure and its documentation for soldiers throughout the empire.215 There are also epigraphic testimonia for the probatio of officeholders and members of pro- fessional groups. At Beneventum, we find a man who had been probatus as a specialist in weights and measures and granted a public salary.216 In the colonia of Madauros in North Africa, we encounter a municipal officeholder adlected ex inquisitione—​“upon investigation,” presumably into his status, wealth, and family history—​as a priest of the imperial cult (flamen perpetuus).217 In Umbria, we find a sacerdos probatus, likely signifying a similar procedure.218 One also suspects that the Roman citizen and medicus from Prusa, who was probatus by clarissimi viri at Rome to serve as an assessor, had been evaluated by a senatorial commission in a procedure analogous to epikrisis.219 Further evidence for status verification may be found in provincial inscriptions mentioning administrators of the census. In the cursus inscriptions of high-​ranking senators who held governorships and served as censitores of major provinces, the function presumably designated a supervisory role in a province-​wide fiscal review.220 Elsewhere, however, we find military officers appointed to “receive the census” (ad census accipiendos), a le- gionary tribune who served as censitor Thraciae, a praefectus cohortis appointed censitor of Roman citizens in the colonia of Camulodunum in Britain, and a le- gionary officer (tribunus or praepositus) who served as censitor of Roman citizens in the assize (conventus) of Caesaraugusta in Spain.221 The latter two, in particular, 2 13. See the earlier discussion at note 167. 214. See the earlier discussion at notes 106, 107, 108, 130, and 132. 215. See, e.g., CIL VI, 2731 = MEFR 2002, 742 (Rome, third century CE). 216. RAAN 1965, 141 (Beneventum, second century CE), ILAlg 1, 2145 = AE 1907, 234 = AE 1919, 37 (Madauros). 217. Epigraphica 1996, 52 = AE 1996, 601 = AE 2013, +444 (Casuentum, second century CE). 218. CIL XI, 3943 = ILS 7789 (Etruria, 87 CE). A municipal doctor once probatus could also be deprived of his 219. status by the municipality, see Dig, 27.1.6.6 (Modestinus 2 excus.). 220. See, e.g., CIL VI, 1333  =  CIL VI, 31633  =  ILS 1077  =  AE 2016, 28 (Rome, 172–​177 CE), a senatorial career that includes the consulship and governorships of Cappadocia, Arabia, and Lusitania, the last as legatus Augusti and censitor. See further CIL V, 7783 = ILS 1128 (Albingaunum, third century CE) and CIL II, 4121 = CIL V, 1005 = ILS 1145 (Tarraco, early third century CE). The evidence is enumerated and discussed by Brunt 1981, 333–​335, and Bérenger 2009. 2 21. For military men appointed ad census accipiendos, see, e.g., AE 1975, 251 (Paestum, 71–​79 CE) and CIL VIII 5355 = CIL VIII 17493 = ILAlg 1, 282 (Calama, Africa Proconsularis). For censitores of military rank, see, respectively, CIL V, 7784 (Albingaunum, third century CE), CIL XIV, 3955 = ILS 2740 (Nomentum, second century CE), and CIL VIII, 7070  =  CIL VIII, 19428  =  ILAlg 2.1, 669  =  AE 2011, 1775 (Cirta, Documenting Roman Citizenship  227 pose a suggestive parallel to the legionary tribunes appointed by prefects of Egypt to receive and evaluate the epikrisis applications of Roman citizens.222 From an administrative perspective, the status verification (epikrisis = recensio, probatio) of Roman citizens can be described as a “backup” procedure in which official documentation produced by the procedures of birth registration, man- Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 umission, discharge from the army, and the census was collected, examined, and verified once again. If indeed epikrisis was a prerequisite for Roman citi- zens to claim fiscal exemptions for themselves and their children and slaves, as the evidence from Egypt suggests, then we are observing the introduction of a very impressive standard of documentation and proof of Roman status. That this was indeed the case is suggested by analogous procedures being applied to other fiscally privileged groups and being assiduously documented by Roman administrators. In view of the clear evidence that epikrisis applications from Greek cities and all forty-​five mētropoleis of Egypt were reviewed by Roman officials at Alexandria every year, it follows that administrators in the provincial capital disposed of detailed records of the status and family history of male citi- zens and members of the civic elite in all major cities of the province. The likeli- hood that similar procedures were implemented beyond Egypt opens a striking perspective on the Roman state’s potential for administrative and fiscal surveil- lance of urban elites throughout the empire. Conclusion Papyri provide by far the most abundant and detailed evidence for Roman pro- vincial administration. It is all the more peculiar that this rich corpus tends to be marginalized, as if Roman Egypt were an exception from the norms of Roman administration elsewhere in the empire. The present investigation has taken the opposite approach, treating papyrological evidence as providing a high-​resolution image of institutions and practices that were generally present in the Roman provinces. One result of this investigation has been to demonstrate how scattered pieces of evidence from other provinces can be comfortably inte- grated within the general picture deriving from papyrological sources. Similarly, procedures that are often regarded as specific to Egypt (such as epikrisis) can, without too much effort, be shown to be part of a broader continuum of Roman administrative practice: just as epikrisis was semantically linked with the Latin recensio and probatio and finds an important parallel in the status verification (probatio) of Roman military recruits, evidence for the population census in Egypt corresponds to procedures attested in southern Italy and in Rome itself. second to third centuries CE). In AE 1931, 36 (Sala, Mauretania, 144 CE), we find an officer who was detained in Cappadocia to supervise the census. 2 22. See, e.g., P. Diog. 6 (Arsinoite, 143 CE) and BGU I 465 = W.Chr. 459 (Arsinoite, 148 CE). 228  Anna Dolganov Similarly, Roman birth declarations in Egypt conform to the evidence of the Roman legal sources and sources from Roman North Africa. Seen from this angle, Roman provincial administration reveals itself as replicating recognizable patterns, which in their main contours hearken back to the institutions of repub- lican Rome and its administration of Italy. Downloaded from https://academic.oup.com/book/43924/chapter/372329708 by Vienna University Library user on 30 January 2023 The subdivision of people into discrete status groups, each with specific privileges and obligations in legal and fiscal terms, was a distinctive and charac- teristic aspect of Roman social life that was integral to the Roman approach to governance, fiscality, and social order within the empire. It is only to be expected that the developing apparatus of the Roman state should devote considerable ef- fort to keeping track of status and fiscal privilege within the imperial population. And this is precisely what the evidence shows. Overall, the results of this investigation underscore the importance of the Roman apparatus of administrative record keeping as a context in which status claims—​including claims of Roman citizenship, the most privileged status be- fore 212 CE—​were made by individuals and evaluated by officials in the Roman Empire. Implicit in the story of Paul’s Roman citizenship is a self-​evident no- tion that status claims could actually be verified; regardless of the truthfulness of the story, this notion is a necessary underpinning of its plausibility. Only this explains why it was sufficient for Paul to utter the magic words civis Romanus sum. Paul’s claim that he was born a citizen implies that his birth had been regis- tered. Even if he did not carry a copy of his birth declaration, it would have been possible to locate it in the archives of his province, along with evidence for his registration in the census, his name in the tax lists, his status verification, and so on. The tribunus who had purchased his Roman citizenship would in turn have had a diploma documenting the grant, registered in the imperial records at Rome and in the capital of his province of residence. That Phlegon of Tralles could get his hands on demographic records for centenarians in Italy and several Roman provinces offers a further illustration of the immense administrative knowledge wielded by the Roman state over its provincial subjects. Needless to say, then as now, knowledge is power.223 223. The sphere of administrative record keeping provides an ideal starting point for pursuing the ques- tion of the infrastructural power of the Roman imperial state, posed by Ando 2017a with reference to Mann 1984.