Version classiqueVersion mobile

L’identification des personnes dans les mondes grecs

 | 
Romain Guicharrousse
, 
Paulin Ismard
, 
Matthieu Vallet
, 
et al.

Première partie. Administrer l’identité

Tracing the Elite from Ptolemy to Diocletian

Identifiers as Clues of Privileged Status in Graeco-Roman Egypt

Yanne Broux

Texte intégral

  • 1 See Rudolf Haensch, “Die Provinz Aegyptus: Kontinuitäten und Brüche zum Ptolemäischen Ägypten. Das (...)
  • 2 All the data mentioned in this paper was collected through the Trismegistos People (www.trismegisto (...)

1The impact of the Roman annexation of Egypt has been the subject of much debate. It has long been considered an exceptional province, since a large part of the administrative framework set up by the Ptolemies seems to have been preserved at first sight. Others have claimed a complete institutional rupture.1 Sources attesting to the earliest years of Julio-Claudian rule are scarce, but it is now agreed upon that in the course of the first century AD things did change. The population was restructured, tax systems were reformed, and an administrative network evolved that centered on the nome capitals, now called metropoleis. These measures are clearly reflected in the changes regarding specific elements of personal identification: double names, metronymics and designations of origin.2 Each in its own way may provide clues about a person’s identity and status, as well as the organization of society in the Egyptian chora at large.

From Ptolemy I to Diocletian: social history in a nutshell

2To put it (perhaps slightly too) simply, the Ptolemaic population was divided into “Greeks” and “Egyptians.” The Greek citizens of Alexandria, Naukratis and Ptolemais were more or less autonomous. The Egyptian population, on the other hand, resided in the chora, the “countryside.” However, from the very beginning of Ptolemaic rule, a steady stream of foreigners found its way to all corners of the country as a result of the government’s active settlement policy. Many who took up a career in the army or administration were awarded grants of land for their services, and thus settled alongside the indigenous population.

  • 3 See Sandra Coussement, “Because I am Greek”: Polyonymy as an Expression of Ethnicity in Ptolemaic E (...)

3These settlers were not all Greeks emigrating from what we today call Greece. People from all over the Mediterranean flocked to Egypt: besides Athenians and Argives, there were Macedonians, Thracians, Cyrenaeans, Cretans… Yet for taxation purposes, they were all considered Ἕλληνες. In Ptolemaic society, status was therefore determined by ethnicity. On an institutional level, distinction was furthermore promoted by establishing a separate Greek legal system alongside the Egyptian one, for example. On the personal level, however, these borders could be crossed in certain spheres. Pretty soon, immigrants started to marry into local families, while Egyptians worked their way up the administrative or military ladder resulting in multi-ethnic identities. Ethnics therefore evolved into socio-legal categories, and the connection to geographical origins was lost.3

4For the Romans, the situation in Egypt was no doubt confusing and impractical, and so these ethnic status markers were abandoned after the regime change. Instead, the population was classified into legal categories based on citizenship, in accordance with the other eastern provinces: Roman citizens, citizens of the Greek poleis, and non-citizens (peregrini Aegyptii or even simply “Egyptians”). This reorganization carried with it a rather significant flaw, however: given the low number of poleis in Egypt, the local elite body was not large enough to administer the province, especially not one as important and economically vital as Egypt. Moreover, it implied a considerable social downgrade for the Hellenes living in the chora, as they were suddenly robbed of their privileged position over the rest of the population.

  • 4 Peter van Minnen, “Αἱ ἀπὸ γυµνασίου. ‘Greek’ Women and the Greek ‘Elite’ in the Metropoleis of Roma (...)

5To appease the Hellenized populace, on whom the Romans needed to rely to fill this gap of qualified (from a social perspective) officials, a unique substructure was built within the ranks of the Aegyptii. Two distinct, but partly overlapping groups were created to demarcate the Hellenic population: the µητροπολῖται (incorporating the inhabitants of Egypt’s district capitals) and οἱ ἀπὸ γυµνασίου.4 Their higher social status was reflected in their poll tax rates. This tribute was imposed on all Aegyptii, but members of these privileged groups were charged at a much lower rate.

  • 5 For a more detailed discussion of the origins and early evolution of these two groups, see Yanne Br (...)
  • 6 SB 20 14440 (TM 14882).

6When this new social order was introduced exactly remains unclear.5 The poll tax seems to have been levied immediately after the annexation of the region.6 Since Roman and Greek citizens were exempt from the poll tax, and, consequently, from submitting census declarations, some sort of hierarchy must therefore already have existed within the class of Aegyptii by then.

  • 7 I. Portes du désert 25, l. 2–3 (TM 88338).
  • 8 As is clear from several texts from the end of the first century, e.g. P. Oxy. 2 257 (TM 20527, AD (...)

7The distinction between “those from the metropolis” and “those from the nome” is first attested in AD 17, and was a Roman novelty. The apo gymnasiou, on the other hand, were recruited from the (in Ptolemaic times private) gymnasia that were scattered all over the chora. The Romans converted membership into an official status, which, for Oxyrhynchos at least, can be traced back to AD 4–5.8 A plausible scenario is that initially, only the metropolitans were singled out, as Romans liked to award status upgrades to distinct communities (compare with the coloniae and municipia in the West). As stated above, however, in Ptolemaic Egypt social boundaries were not tied to residency. Many Hellenes lived in villages, and so again, prominent families were left out. The drafting of the first official gymnasium membership lists in AD 4–5 then rectified this blunder, and taken together, these two groups incorporated the former Ptolemaic Hellenes.

8The Romans made another miscalculation here though. Traditionally, gymnasial membership was passed down through the male line only, and this rule was not altered once this status was made official. This resulted in a steady influx of new members, and, consequently, in a significant loss of poll tax income for the government. This group was therefore “closed” in the second half of the first century AD. General status assessments were held throughout the entire province, and a definitive selection was made on the basis of previous lists. From then on, gymnasial status also required membership in both the paternal and the maternal line.

9Metropolitan status, on the other hand, followed Roman social practice. Here, only children whose father and maternal grandfather were members, were qualified, so this group was already a “closed” group from its beginnings.

  • 9 TM 25306 = SB 5 8038 with BL 4 82 (first century AD).
  • 10 This resulted in impressive displays of pedigree by the third century AD, with epikrisis declaratio (...)
  • 11 Many continued to own land in village territories where their ancestors had once settled under the (...)

10To facilitate status control, a process called epikrisis was established at this time as well. When a boy qualifying for metropolitan and/or gymnasial status came of age when he was 14, his parents submitted a status declaration listing the necessary credentials. For metropolitans, proof of descent from both parents sufficed (generally obtained from previous census lists).9 Gymnasial candidates were required to present an elaborate family tree tracing back their ancestors’ membership to the revised selection.10 Moreover, it became mandatory for members of both groups to register as a resident of a district capital.11

  • 12 The latest epikrisis declaration was filed in Oxyrhynchos in AD 295 (TM 16016 = P. Oxy. 43 3137).

11The metropolitan and gymnasial groups, and the whole status-related administration built around them, survived until the end of the third century AD. Several reforms during this century undermined this social and fiscal structure, however. The introduction of the town councils in the metropoleis in AD 201 attracted more and more Greek citizens to the chora, who married into local elite families; the universal grant of citizenship in AD 212 dissolved all legal boundaries; and economic and military struggles disrupted the organization of the census and the poll tax. Once all discriminatory factors had disappeared, these groups became redundant, and indeed, references to them disappear under Diocletian.12

More names, more claims: polyonymy and double names

  • 13 Coussement, “Because I am Greek”, op. cit., 23 ff. for an overview of polyonymy in the Pharaonic an (...)

12Polyonymy (i.e. having multiple names) had been around in Egypt quite some time before the Macedonians swooped in and heralded a new era of cultural diversity. Pharaohs usually adopted multiple names (various “throne” names, combined with their birth name); and as far back as the Old Kingdom commoners are attested with a “small name” or a “beautiful name”, apart from their given (“great”) name.13

  • 14 For a more detailed analysis of the different topics concerning double names presented here, please (...)

13The term “double name” has become widely established in onomastic/prosopographic and identity studies since the phenomenon was first studied at the beginning of the twentieth century. When mentioned, scholars more or less instinctively know what is meant. Nevertheless, defining a double name is not as easy as it may seem at first sight, especially since two recent comprehensive studies, conducted by Sandra Coussement and myself,14 have revealed that the purpose of polyonymy differed greatly in the Ptolemaic and Roman periods.

14A double name will generally be described as a set of two given names. But how do you determine whether the second name is an official name, or rather a nickname or a hypochoristic? Do the two names have to be used simultaneously at all times? At least once? What if no connecting formula is used? If a translation of a name is systematically used in another language instead of a transliteration, does this qualify as a second name, e.g. Λέων in Greek and Pȝ-mȝy in Egyptian (instead of Lȝn)?

  • 15 Broux, Double Names, opcit., p. 7.

15To be able to take differences in naming practices into account, we opted for a broad definition as “a set of two or more sufficiently different personal names, each freely chosen to identify an individual, at birth or at a later stage in life, used either alternatingly or in combination. In the latter case the two names are often connected by a formula.”15 By defining the names as “sufficiently different”, we mean to exclude transliterations of a name into another language, as well as orthographic variants and hypocoristics.

  • 16 Available at www.trismegistos.org/ref. Separate prosopographies for the Ptolemaic and Roman periods (...)

16In total, we identified 4,858 individuals with multiple names: 393 from the Ptolemaic period, and 4,465 from the Roman period (up till AD 400).16 Four characteristics stand out when comparing the two periods: the difference in popularity, the linguistic origins, the use, and the connecting formulas.

Popularity

  • 17 Coussement, “Because I am Greek”, op. cit., p. 38.
  • 18 Examples from the fourth century BC are restricted to the timeframe 332–301 BC (15 exx.). One perso (...)
  • 19 www.trismegistos.org/archive/256.

17The large discrepancy regarding the number of polyonymous people is particularly striking. For the Roman period, the figure is more than 12 times as high as the Ptolemaic period. When looking at the breakdown per century, a gradual increase is visible [figure 1]. The apparent dip in the third century BC results from the introduction of Greek after the Macedonian take-over.17 The examples from the fourth18 and third centuries involve traditional Egyptian forms of polyonymy. This unfamiliar practice did not immediately find its way into the early Greek sources, which are quite numerous, thanks to the extensive Zenon archive.19 This began to change in the second century, as more and more Egyptians, and their double names, are attested in the Greek documentation, but also because of the emergence of so-called bilingual polyonymy (see below: “Linguistic origins”). The number continues to rise steadily until the end of the first century AD, after which it increases exponentially until the third. Their demise is even more swift than their upsurge: by the end of the fourth century AD, numbers are back to the level of the early Ptolemaic period.

18At the height of their popularity, a little over 4% of the population of Roman Egypt had two (or more) given names. This discrepancy with the Ptolemaic period should be put into perspective though. As we shall see further on (“Use”), the way these names were used during the earlier centuries makes them far more difficult to trace, so it is likely they were more widespread than we are aware of.

Figure 1. Chronological distribution of individuals with a double name

Figure 1. Chronological distribution of individuals with a double name

Linguistic origins

19The few examples from the early Hellenistic period are generally composed with two Egyptian names, as had been customary in the region from Egypt’s early history (see above). In the third century BC, this remained the most common combination (and may even be underrepresented in our Greek sources), but Greek names are already attested occasionally, mostly in combination with an Egyptian name [figure 2]. A major shift occurred in the second century BC, when these “bilingual” double names make up half of all examples. Greek-Greek double names only really started to catch on in the course of the first century AD, and they quickly became the predominant type for the rest of the Roman period. At the same time, Latin names were incorporated sporadically, and in the second and third centuries AD the combination of a Greek and a Latin name became the second most common composite.

Figure 2. Language combinations in double names

Figure 2. Language combinations in double names

Use

  • 20 Thanks to Trismegistos’ system of “weighed dates”, designed to incorporate broadly-dated texts in s (...)
  • 21 See figure 1.

20Taking a look at double names on the level of attestations rather than on the level of individuals, reveals other interesting features. The first concerns the way double names were used. Figure 3 shows the chronological evolution of attestations of double names.20 590 references are Ptolemaic; 7,115 are Roman. It follows more or less the same pattern as the distribution of individuals.21 On the basis of this graph, the peak in the third century can be pinpointed more precisely around the middle of the third century AD. One should factor in a certain “delay”, however, since in most cases there is a gap between the moment a name is bestowed (generally at birth), and the moment a person is mentioned in a text. The actual peak should therefore rather be situated at the beginning of this century.

Figure 3. Chronological distribution of attestations of double names

Figure 3. Chronological distribution of attestations of double names
  • 22 7,115 out of a total of 8,117.
  • 23 Coussement, “Because I am Greek”, op. cit., p. 210.
  • 24 2,209 references to 393 people vs. 8,117 references to 4,465 people.

21Again, the Ptolemaic figures require some contextualization. The 590 references in this graph are attestations of polyonymous individuals where the two names are used simultaneously in a single reference, i.e. as actual double names. But in fact, these people are more often attested using only one of the two names (1,649 instances). In the Roman period, this changes completely, as 87.7% of the attestations then include both names.22 This means that the number of polyonymous individuals in the Ptolemaic period was probably (much) higher than we can estimate now, since it is very difficult to trace people if they are mentioned with different names in different texts.23 Moreover, when they can be traced, people with two names from this period are much more frequently attested (5.6 times on average) than in the Roman period (1.8 times).24

Formulas

  • 25 Coussement, “Because I am Greek”, op. cit., p. 50–55.
  • 26 Broux, Double Names, opcit., p. 146–149.
  • 27 As is, for example, the case with Νίκωνος ὃς καὶ Πετεχωνσις in P. BM Andrews 27, l. 3–4 (TM 8518, 2 (...)
  • 28 Short for ὁ καὶ καλούµενος; examples of this full form are rare, however: Coussement, “Because I am (...)
  • 29 Broux, Double Names, opcit., p. 110 ff.

22Egyptian double names were either constructed with the relative clause nty ỉww ḏd nf (“to whom they say”) or the archaic forms ḏd.tw nf and ḏd.w nf (“who is called”), in the Ptolemaic period.25 The handful of examples from the Roman period mainly use the latter formula.26 In Greek, a relative clause was initially used as well: ὃς καὶ καλεῖται (“who is also called”), but the verb was quickly dropped. Although as a relative pronoun, ὃς καί could in theory be used for any case,27 it was generally reserved for the nominative. For a casus obliquus, a declined form of ὁ καί was the norm.28 This difference according to inflection was rather confusing, however, and gradually ὁ καί was introduced for nominative forms as well. By the second century AD, it had become the standard expression for double names (around 80–90% of all attestations), regardless of their case.29

  • 30 Ibid., p. 120–123.
  • 31 P. Ashm. 1 22, FrA, l. 3 (TM 2655).

23Other Greek formulae were virtually absent in the Ptolemaic period. Ἐπικαλούµενος is only used three times for “commoners”, while it became the second most common expression of the Roman period.30 It was, however, frequently used to introduce the second names of late Ptolemaic kings and queens, e.g. Πτολεµαίου τοῦ ἐπικαλουµένου Ἀλεξάνδρου.31

Change in purpose

  • 32 www.trismegistos.org/person/398.
  • 33 www.trismegistos.org/person/5259. See also Coussement, “Because I am Greek”, op. cit., p. 159 ff. f (...)
  • 34 Ibid., p. 147–150.
  • 35 Coussement, “Because I am Greek”, op. cit., p. 210–212.

24The popularity of bilingual polyonymy and the habit of switching between names in the Ptolemaic period was a result of the ethnic borders that characterized this society. Asklepiades, who lived in Pathryis around 150–88 BC, for example, is known by his Greek name in the official documents where he acted as agoranomos. In his private Demotic contracts, on the other hand, he identified himself as Patseous.32 The same goes for Apollonia-Senmonthis, wife of the Cretan cavalryman Dryton: in Greek business documents she is Apollonia,33 who acts under the supervision of her husband according to proper Greek custom. In private accounts and letters, often written in Demotic, she is Senmonthis, who can act independently. Even in official documents a choice was often made. In tax registers, for example, those with “Greek” status were reported with their Greek name only, since in that context their Greek identity prevailed.34 Polyonymy is therefore mainly attested among those that formed a bridge between the Macedonian ruling class and the Egyptian population, particularly military, administrative and priestly officials. Since they moved between largely separated ethnic contexts, combining names of different linguistic origins was a way to negotiate their ethnic identities.35

  • 36 Broux, Double Names, opcit., p. 291–292.

25When the Romans removed the ethnic borders, bilingual polyonymy lost its purpose. The new social structure, focusing on Greek descent, provided a new framework in which double names became status markers used by the local elite to stress Greek origin and culture (hence the preference for all-Greek double names), and to distinguish them from “ordinary” Egyptians (by consistently using both names simultaneously with the standard formula ὁ καί). At the same time, this indigenous form of polyonymy resembled Roman nomenclature, where multiple names where standard. The tria nomina were, in theory at least, restricted to Roman citizens, so for some people perhaps double names provided a subtle means to associate themselves with the ruling class.36

Who’s your mommy?

  • 37 See Demosthenes, Speeches, 39, 9. Exceptional examples can be found on inscriptions for courtesans (...)
  • 38 Mark Depauw, “The Use of Mothers’ Names in Ptolemaic Documents: A Case of Greek-Egyptian Influence? (...)

26The use of mothers’ names (“metronymics”) for personal identification is another Egyptian phenomenon that found its way into Ptolemaic and Roman practice. In the Greek world, this was inconceivable.37 Normally the father’s name sufficed; the paternal grandfather could be added to avoid confusion in severe cases of homonymy. Even in Egypt, metronymics were added only in specific contexts: mainly contracts and funerary texts, where precise identification was paramount for legal and “magical” purposes respectively. The first attestations in Greek texts do not appear until the middle of the second century BC. On closer inspection, however, most of these examples turn out to be translations from Egyptian.38

  • 39 Mark Depauw, “Do Mothers Matter? The Emergence of Metronymics in Early Roman Egypt”, in Trevor V. E (...)

27In the Roman period, long strings of “genealogical identifiers”, as the names of these family members can be called, came into vogue, such as the cluster “Arsinoe, daughter of Herakleides the younger son of Didymos, and of Ptolema daughter of Ptolemaios.” On the basis of the manually manageable corpus of data from the first century AD, Depauw surmised that the increase in metronymics was related to the first-century reforms discussed above.39

  • 40 PSI 10 1159 (TM 13863, AD 132).

28Thanks to Trismegistos People, more accurate calculations can now be made, encompassing the entire Greco-Roman period. During the semi-automated extraction of the personal names from the Greek documentary texts in the Duke Databank of Documentary Papyri [http://papyri.info], the role of each person mentioned in a person’s so-called “identification cluster” was marked: Ἀρσινόῃ Ἡρακλείδου νεωτέρου τοῦ Διδύµου µητρὸς Πτολέµας τῆς Πτολεµαίου,40 for example, is interpreted as person – father – [filiation marker] – paternal grandfather – [filiation marker] – mother – [filiation marker] – maternal grandfather. Initially, this was done to recreate genealogies, but it soon became clear that these cluster mark-ups were of particular use when studying identification practices.

  • 41 See Yanne Broux, Mark Depauw, “The Maternal Line in Greek Identification: Signalling Social Status (...)

29In total, Trismegistos People counts 14,514 identification clusters mentioning the mother’s name (260 BC–AD 410).41 Only 105 of these appear in texts dated to the Ptolemaic period, an imbalance that is clearly visible in figure 4.

  • 42 I’ve counted the number of texts, not the individual number of identification clusters, to avoid ma (...)

Figure 4. Chronological distribution of texts containing identification clusters with a metronymic42

Figure 4. Chronological distribution of texts containing identification clusters with a metronymic42
  • 43 Depauw, “The Use of Mothers’ Names in Ptolemaic Documents”, art. cit., 25, table 3.
  • 44 Depauw, “Do Mothers Matter?”, art. cit., p. 124–125 and 132.

30The exceptional peak in 160–131 BC can be explained by a group of contracts that were translated from Demotic.43 The addition of a mother’s name was indeed rare in the Hellenistic period. Figures start rising fast from the end of the first century BC onwards. In the beginning these are mostly documents related to the census and taxes, but in the second half of the first century AD the practice is taken up in other types of official texts as well.44 By the middle of the third century AD metronymics are attested in no less than 32% of all texts. Just like double names, their numbers drop spectacularly in the fourth century AD, after which they remain steady around 8%.

31In comparison, the addition of the maternal grandfather (“papponymic”) was typical of a very specific time frame [figure 5]:

Figure 5. Chronological distribution of texts containing identification clusters with a maternal papponymic

Figure 5. Chronological distribution of texts containing identification clusters with a maternal papponymic

32The 863 identification clusters with a maternal papponymic are mostly concentrated in the period AD 51–110, when 34–39% of all metronymics are followed by a maternal papponymic. There are no examples from the Ptolemaic period, and after AD 110, the practice quickly declined again.

  • 45 Depauw, “Do Mothers Matter?”, art. cit., p. 122–123, table 7.1.
  • 46 Broux, Depauw, “The Maternal Line in Greek Identification”, art. cit., p. 476–477.

33As is the case with double names, the “rise and fall” of the maternal line nicely reflects the changes described above. The addition of metronymics picked up right around the time the metropolitai and apo gymnasiou become visible in our sources, and precisely in those types of documents that were closely related to status, i.e. census lists and declarations, and tax receipts (apart from contracts, a clear continuation of Egyptian practice).45 Paternal papponymics were introduced at a second crucial moment, around the same time the gymnasial entry regulations were tightened. To prove membership through both the paternal and maternal line in these early years, it was necessary to revert to the maternal grandfather, since previously the only requirement had been that the mother were freeborn. Once the definitive membership lists were established and new generations were born, this element became less essential and was soon dropped from the identification cluster again.46

  • 47 This format accounted for 60% of all identification clusters with at least one genealogical identif (...)

34A kind of official identification cluster therefore evolved, including a person’s given name(s), patronymic, paternal papponymic and metronymic.47 Once the social hierarchy was dissolved under Diocletian, however, this manner of identification no longer had any practical impact, and metronymics faded out quietly as well.

Location, location, location

  • 48 Citizens of Greek poleis, for example, can be identified by the tribe and deme they were enrolled i (...)

35The last section of this paper presents some preliminary results regarding designations of origin. These can take on various forms,48 but the focus here is on those markers that could possibly point to local elite status in the chora, i.e. ethnics (adjectives derived from a foreign city or region, e.g. Athenaios or Thraix), and designations pointing to registration in a metropolis.

  • 49 www.trismegistos.org/geo.
  • 50 www.trismegistos.org/ref.

36Since toponyms are stored in a separate database (Trismegistos Places49) than people and names (Trismegistos People50), and there is no direct link between the two yet, tracking down these designations when used in identification clusters is not as straightforward as looking for types of names or genealogical identifiers. After targeted searches for references to ethnics and metropoleis in the Places database, these had to be checked manually to determine whether an attestation is part of an identification cluster, or a stand-alone toponym. So far, I have thus collected 6,787 identification clusters with either an ethnic (3,817) or a reference to a district capital (2,970) dated between 332 BC and AD 410.

Ethnics

  • 51 The percentages presented here for the Ptolemaic period are lower than the results calculated by Fi (...)

Figure 6. Chronological distribution of ethnics in identification clusters51

Figure 6. Chronological distribution of ethnics in identification clusters51
  • 52 P. Hamb. 2 168, l. 5–10 (TM 4322, 275–225 BC) and BGU 14 2367, 5–12 (TM 2367, 225–200). For this so (...)
  • 53 Uri Yiftach-Firanko, “Did BGU XIV 2367 Work?”, in Mark Depauw, Sandra Coussement (eds.), Identifier (...)

37303 cities or regions outside Egypt are attested as ethnics identifying individuals. As expected, they are clearly a Ptolemaic phenomenon [figure 6]. Their use was regulated by a royal decree issued early on under Ptolemy II.52 In contracts and litigations, for example, anyone who was not a soldier or a citizen of a Greek polis was required to identify himself with his patronymic, ethnic and genos. The purpose of the decree was to provide an official framework for the identification of the new immigrant population in light of the creation of the dikasteria, the Greek law courts that operated alongside the Egyptian laokritai.53

  • 54 Katelijn Vandorpe, “Zenon son of Agreophon”, in Katelijn Vandorpe, Willy Clarysse, Herbert Verreth (...)

38Ethnics were most often used during the third century BC, when these designations still had an actual geographical value for first-generation immigrants. In absolute terms (the gray bars), the number of attestations of individuals mentioning an ethnic rises considerably until 191 BC. When comparing these numbers to the total amount of references to people, a temporary setback seems to occur around 280–251 BC (the black line), but this is actually the result of our “skewed” corpus: the bulk of documents from the enormous Zenon archive falls in this period, and mainly consists of letters and accounts,54 two types of texts in which ethnics were generally not used.

  • 55 Coussement, “Because I am Greek”, op. cit., p. 153.
  • 56 Ibid.
  • 57 The Ptolemies clearly favored those population groups that promoted Greek culture: for a full list (...)

39In the second century BC a slow but steady decline sets in. Not only does the number of attestations diminish, there is also less variety than in the previous century.55 Since many descendants of immigrants had no more actual connections to their forefathers’ homeland, only the most common ethnics survived in large numbers. As these were often already associated with specific population groups, they evolved into broader, “standardized” status markers. Initially, the army was mostly made up of Macedonians, for example, so the term Makedon became a generalized ethnic for men serving in the military.56 The terms Hellen, Perses and Araps, on the other hand, came to denote individuals benefitting from tax privileges.57

  • 58 See e.g. Katelijn Vandorpe, “Persian Soldiers and Persians of the Epigone: Social Mobility of Soldi (...)

40In the Roman period, 68% of all attestations are Πέρσαι τῆς ἐπιγονῆς. Originally a (fictitious) ethnic used by a class of soldiers, it evolved into a juridical label somewhere under the later Ptolemies, and was used exclusively as such in Roman contracts to designate a debtor with full personal liability.58 Other ethnics are extremely rare, and by the end of the second century AD, hardly any are attested at all. Status was now determined by internal geo-political criteria, as members of the local elite were concentrated in the pseudo-civilian environment of the metropoleis.

Metropolitan designations of origin

  • 59 Yiftach-Firanko, “Did BGU XIV 2367 Work?”, art. cit., n. 17, p. 108.

41Ethnics did, of course, not apply to the native population. Egyptians were simply identified by their name and patronymic; in some cases they added their domicile (the so-called “Heimatsvermerk”)59, but it was not mandatory, since it had no legal or social value.

  • 60 As noted above, villagers could also add their origin, but collecting all designations of origin is (...)

42Once location became a key element to determine metropolite and gymnasial status in the first century AD, however, references to metropolitan domicile became a regular component of the identification cluster.60 Three types of metropolitan designations can be distinguished: those ending in –ιτης (e.g. Ὀξυρυγχίτης) and the generic µητροπολίτης; ἀπὸ X πόλεως (e.g. ἀπὸ Πανὸς πόλεως) and the generic ἀπὸ µητροπόλεως; and references to the amphodon a person was registered in (e.g. ἀναγραφόµενος ἐπ’ ἀµφόδου Βιθυνῶν Ἰσίωνος).

Figure 7. Chronological distribution of – designations in identification clusters
  • 61 P. Oxy. 27 2476, l. 22 (TM 17010, AD 288).

43Figure 7 shows the chronological distribution of the –ιτης expressions. It is attested in the Ptolemaic period as the counterpart of foreign ethnics, but only rarely. This type is used somewhat more often from the end of the first century AD onwards. The sudden peak at the end of the third century AD is tied to a specific context: around this time, it was common for athletes to list all cities, including metropoleis, of which they were (honorary) citizens, and the –ιτης designation was the preferred format. Aurelius Herakleios-Nikantinoos, for example, is styled as Antinoeus, Panopolites, Hermopolites, Lykopolites and Oxyrhynchites,61 but this is not a standard identification cluster.

44Adding ἀπὸ Χ πόλεως or ἀπὸ µητροπόλεως to one’s identification cluster was the most common way to denote origin. Again, Ptolemaic examples are scarce, but this type is picked up very quickly once epikrisis became mandatory in the second half of the first century AD [figure 8]. Once the membership lists were firmly established, numbers drop again in the second century but then remain stable. In contrast to the other forms of identification discussed in this paper, this type of designation does not die out after the Diocletian reforms, but goes through a remarkable revival at the end of the fourth century AD. I leave the interpretation of these results to others who are more familiar with this time frame, however.

Figure 8. Chronological distribution of ἀπὸ X πόλεως in identification clusters

Figure 8. Chronological distribution of ἀπὸ X πόλεως in identification clusters

45Finally, a person could also mention the metropolitan district where he or she was registered. The typical formulation reads ἀναγραφόµενος ἐπ’ ἀµφόδου X, but variations with ἀπογραφόµενος and ἐπικριθείς are also attested, as well as a shortened version ἀπὸ ἀµφόδου X.

  • 62 Richard Alston, The City in Roman and Byzantine Egypt, London/New York, Routledge, 2002, p. 138–140

46The division of metropoleis into official districts or amphoda was most likely related to the early Imperial reforms regarding the privileged metropolitan groups. Attestations of the word ἄµφοδον before AD 51 are scarce and often uncertain.62 Once registration became mandatory for both metropolitai and apo gymnasiou, the number of attestations in identification clusters rises quite steeply [figure 9]. The temporary setback in AD 171–200 can be explained by the tax rolls of Karanis: since these long lists of individuals include villagers only, they distort the results. When leaving these texts out, the use of designation remains stable until the end of the third century AD. Once the privileged groups quit the scene and status is no longer related to domicile, these districts are no longer specified in a person’s identification.

Figure 9. Chronological distribution of references to amphoda

Figure 9. Chronological distribution of references to amphoda

47Which type of designation was preferred, seemed to depend on the metropolis in question. Table 1 presents a breakdown according to the four most well-documented metropoleis: Oxyrhynchos, Krokodilopolis, Hermopolis and Herakleopolis. In Oxyrhynchos, the phrase ἀπ’ Ὀξυρύγχων πόλεως was used almost exclusively, and (to a slightly lesser extent) ἀφ’ Ἡρακλέους πόλεως in Herakleopolis. In the other two metropoleis, there was more variation. Inhabitants of Krokodilopolis generally combined ἀπὸ µητροπόλεως with the amphodon until the fourth century AD, when they switched to ἀπὸ τῆς Ἀρσινοιτῶν πόλεως. Hermopolitans, on the other hand, preferred Ἑρµοπολίτης in combination with the amphodon, but they too went for ἀπὸ Ἑρµοῦ πόλεως after AD 300.

48The main question I am interested in is: can these metropolitan designations be used as status markers? Tracking down metropolitai and apo gymnasiou is difficult: since they are all legally Aegyptii, they do not form a separate class, and are therefore generally not singled out in our documentation. The only exceptions are epikrisis declarations, tax registers and census returns, as this is the only context where their status was relevant for the government. But these types of documents do not provide much background information on these people. Therefore, if these designations are not just merely a kind of “address”, but point to privileged status, the number of relevant texts would increase considerably.

Table 1. Designations of origin according to the four main metropoleis

Oxyrhynchos

Krokodilopolis

Hermopolis

Herakleopolis

-ιτης

35

34

223

8

ἀπὸ µητροπόλεως

14

291

0

7

ἀπὸ

Χ πόλεως

1,440

188

574

94

ἄµφοδον

93

545

203

12

  • 63 Giovanni Ruffini, “Genealogy and the Gymnasium”, Bulletin of the American Society of Papyrologists, (...)
  • 64 Laurens E. Tacoma, Fragile Hierarchies: The Urban Elites of Third-Century Roman Egypt, Leiden/Bosto (...)

49But how to asses whether this is the case? When the metropolitai were first established, this privilege was no doubt extended to all inhabitants at that time. Populations never remain static, however. Over time, people move, and outsiders, i.e. villagers, make their way in. Were these “immigrants” allowed to define themselves as “from the metropolis” as well, or did they continue to mention their village, even if they lived permanently in the metropolis? If the former was the case, then why are there so few examples? Oxyrhynchos counted some 4,000 privileged members on average at any given time,63 while the total population of this city is estimated between 20,000 and 42,000 inhabitants.64 With only 2,970 attestations of individuals mentioning a metropolis over a period of more than 600 years, it seems unlikely that these designations could be used by simply anybody. A more detailed study encompassing all designations of origin, including those referring to the poleis and villages, would help put these metropolitan labels in perspective and thus perhaps answer the question whether they point to elite status.

50The subjects highlighted in this paper show that although indigenous traditions found their way to official Ptolemaic and Roman identification methods, they were, sometimes quite heavily, adapted to accommodate the specific needs imposed by institutional changes. The longstanding tradition of polyonymy in Egypt provided an excellent solution for those navigating between Greek and Egyptian segments of Ptolemaic society, while the local elite of the Roman period used them to stress their “Greekness” and to single them out from the broad mass of Aegyptii. Mothers’ names offered extra surety in Egyptian contracts and texts concerning the afterlife, a practice that was retained in Greek translations, but was not deemed necessary for general identification purposes as long as privileged status was merely patrilineal. Once the Romans proclaimed that eligibility through both lines was mandatory, the maternal line became an important component of the identification cluster. Finally, geographical identifiers could take on different forms. The use of ethnics has been discussed extensively in scholarly literature, as they were an expression of the institutional segregation established by the Macedonian rulers. In contrast, other designations of origin used as identifiers have received little or no attention, even though they became a regular component of the identification cluster in the Roman period. The variety of metropolitan designations shows that, in contrast to double names and genealogical identifiers, there was no standardization at the provincial level when it came to expressing origin. The exact purpose of these geographical identifiers remains unclear though. Further research will hopefully shed some light on if and how they were used by the local elite, and consequently on the composition and no doubt heterogeneous characteristics of this enigmatic section of the Egyptian population.

Notes

1 See Rudolf Haensch, “Die Provinz Aegyptus: Kontinuitäten und Brüche zum Ptolemäischen Ägypten. Das Beispiel des administrativen Personals”, in Ioan Piso (ed.), Die römischen Provinzen. Begriff und Gründung, Cluj-Napoca, Mega, 2008, p. 81–105.

2 All the data mentioned in this paper was collected through the Trismegistos People (www.trismegistos.org/ref) and Trismegistos Places (www.trismegistos.org/geo) databases and can be consulted in these respective repositories.

3 See Sandra Coussement, “Because I am Greek”: Polyonymy as an Expression of Ethnicity in Ptolemaic Egypt, Leuven, Peeters (StudHell, 55), 2016, p. 4–18, p. 131–137, and 140ff. for the concept of ethnicity and the various spheres in which this construct allowed for the creation as well as the crossing of borders in Ptolemaic society.

4 Peter van Minnen, “Αἱ ἀπὸ γυµνασίου. ‘Greek’ Women and the Greek ‘Elite’ in the Metropoleis of Roman Egypt”, in Henri Melaerts, Leon Mooren (eds.), Le rôle et le statut de la femme en Égypte hellénistique, romaine et byzantine. Actes du colloque international Bruxelles-Leuven (27-29 novembre 1997), Leuven, Peeters (StudHell, 37), 2002, p. 337–353.

5 For a more detailed discussion of the origins and early evolution of these two groups, see Yanne Broux, “Creating a New Local Elite: The Establishment of the Metropolitan Orders of Roman Egypt”, Archiv für Papyrusforschung, 59, 2013, p. 142–152.

6 SB 20 14440 (TM 14882).

7 I. Portes du désert 25, l. 2–3 (TM 88338).

8 As is clear from several texts from the end of the first century, e.g. P. Oxy. 2 257 (TM 20527, AD 94–95) and 10 1266 (TM 21768, AD 98).

9 TM 25306 = SB 5 8038 with BL 4 82 (first century AD).

10 This resulted in impressive displays of pedigree by the third century AD, with epikrisis declarations going back across 8 generations, e.g. TM 22165 = P. Oxy. 18 2186 (AD 260).

11 Many continued to own land in village territories where their ancestors had once settled under the Ptolemies though, and therefore probably spent a lot of their time there. See, for example, the second-century family archive of Philosarapis (www.trismegistos.org/archive/192).

12 The latest epikrisis declaration was filed in Oxyrhynchos in AD 295 (TM 16016 = P. Oxy. 43 3137).

13 Coussement, “Because I am Greek”, op. cit., 23 ff. for an overview of polyonymy in the Pharaonic and Late periods.

14 For a more detailed analysis of the different topics concerning double names presented here, please consult Coussement, “Because I am Greek”, op. cit., and Yanne Broux, Double Names and Elite Strategy in Roman Egypt, Leuven, Peeters (StudHell, 54), 2015.

15 Broux, Double Names, opcit., p. 7.

16 Available at www.trismegistos.org/ref. Separate prosopographies for the Ptolemaic and Roman periods are presented in Coussement, “Because I am Greek”, op. cit., 215ff. and in Yanne Broux, Double Names in Roman Egypt: A Prosopography (Trismegistos Online Publications 8), Leuven, 2015 (available at www.trismegistos.org/top) respectively.

17 Coussement, “Because I am Greek”, op. cit., p. 38.

18 Examples from the fourth century BC are restricted to the timeframe 332–301 BC (15 exx.). One person is dated to 400–201 BC and therefore counts for 0.5 for each century.

19 www.trismegistos.org/archive/256.

20 Thanks to Trismegistos’ system of “weighed dates”, designed to incorporate broadly-dated texts in statistical calculations, a more detailed chronological breakdown, e.g. per 30 years as is done here, can be presented for attestations of double names, since the database of attestations of individuals is linked directly to the sources. See Bart Van Beek, Mark Depauw, “Quantifying Imprecisely Dated Sources: A New Inclusive Method for Charting Diachronic Change in Graeco-Roman Egypt”, Ancient Society, 43, 2013, p. 101–114.

21 See figure 1.

22 7,115 out of a total of 8,117.

23 Coussement, “Because I am Greek”, op. cit., p. 210.

24 2,209 references to 393 people vs. 8,117 references to 4,465 people.

25 Coussement, “Because I am Greek”, op. cit., p. 50–55.

26 Broux, Double Names, opcit., p. 146–149.

27 As is, for example, the case with Νίκωνος ὃς καὶ Πετεχωνσις in P. BM Andrews 27, l. 3–4 (TM 8518, 210 BC).

28 Short for ὁ καὶ καλούµενος; examples of this full form are rare, however: Coussement, “Because I am Greek”, op. cit., p. 58, n. 89.

29 Broux, Double Names, opcit., p. 110 ff.

30 Ibid., p. 120–123.

31 P. Ashm. 1 22, FrA, l. 3 (TM 2655).

32 www.trismegistos.org/person/398.

33 www.trismegistos.org/person/5259. See also Coussement, “Because I am Greek”, op. cit., p. 159 ff. for the different domains in which these “bicultural identities” were most prominent.

34 Ibid., p. 147–150.

35 Coussement, “Because I am Greek”, op. cit., p. 210–212.

36 Broux, Double Names, opcit., p. 291–292.

37 See Demosthenes, Speeches, 39, 9. Exceptional examples can be found on inscriptions for courtesans whose mothers had also been in the trade and whose fathers were unknown: Laura K. McClure, Courtesans at the Table: Gender and Greek Literary Culture in Athenaeus, New York/London, Taylor & Francis Books, 2003, p. 76–77.

38 Mark Depauw, “The Use of Mothers’ Names in Ptolemaic Documents: A Case of Greek-Egyptian Influence?”, Journal of Juristic Papyrology, 37, 2007, p. 21–29, at p. 24.

39 Mark Depauw, “Do Mothers Matter? The Emergence of Metronymics in Early Roman Egypt”, in Trevor V. Evans, Dirk D. Obbink (eds.), The Language of the Papyri, Oxford, Oxford University Press, 2010, p. 121–139.

40 PSI 10 1159 (TM 13863, AD 132).

41 See Yanne Broux, Mark Depauw, “The Maternal Line in Greek Identification: Signalling Social Status in Roman Egypt (30 BC– AD 400)”, Historia, 64, 2015, p. 467–478 for a more comprehensive overview.

42 I’ve counted the number of texts, not the individual number of identification clusters, to avoid major distortions caused by exceptionally long lists of names, such as the tax rolls from Karanis, as well as by abbreviated identification clusters in such lists where siblings are described as ἀδελφὸς µητρὸς τῆς ἀυτῆς (Broux, Depauw, “The Maternal Line in Greek Identification”, art. cit., p. 470).

43 Depauw, “The Use of Mothers’ Names in Ptolemaic Documents”, art. cit., 25, table 3.

44 Depauw, “Do Mothers Matter?”, art. cit., p. 124–125 and 132.

45 Depauw, “Do Mothers Matter?”, art. cit., p. 122–123, table 7.1.

46 Broux, Depauw, “The Maternal Line in Greek Identification”, art. cit., p. 476–477.

47 This format accounted for 60% of all identification clusters with at least one genealogical identifier; the sequence “person – father – mother” is attested in 24% of the examples (Broux, Depauw, “The Maternal Line in Greek Identification”, art. cit., p. 472, table 1).

48 Citizens of Greek poleis, for example, can be identified by the tribe and deme they were enrolled in (e.g. Σωσικόσµειος ὁ καὶ Ἀλθαιεύς); village residents added the name of their village as ἀπὸ X.

49 www.trismegistos.org/geo.

50 www.trismegistos.org/ref.

51 The percentages presented here for the Ptolemaic period are lower than the results calculated by Fischer-Bovet, since she based herself on the Prosopographia Ptolemaica, which collected only those people with a title, occupation and/or ethnic, while TM People has entered any person attested, and the denominator for this graph is thus much higher: Christelle Fischer-Bovet, “Official Identity and Ethnicity: Comparing Ptolemaic and Early Roman Egypt”, Journal of Egyptian History, 11, 2018, p. 13.

52 P. Hamb. 2 168, l. 5–10 (TM 4322, 275–225 BC) and BGU 14 2367, 5–12 (TM 2367, 225–200). For this so-called “Nomenklaturregel”, see e.g. Anne-Emmanuelle Veïsse, “L’usage des ethniques dans l’Égypte du iiie siècle”, in Laurent Capdetrey, Julien Zurbach (eds.), Mobilités grecques. Mouvements, réseaux, contacts, en Méditerranée, de l’époque archaïque à l’époque hellénistique, Bordeaux, Ausonius (Scripta Antiqua, 46), 2012, p. 57–66 (with further references to the extensive literature already written on the subject of ethnics and ethnicity).

53 Uri Yiftach-Firanko, “Did BGU XIV 2367 Work?”, in Mark Depauw, Sandra Coussement (eds.), Identifiers and Identification Methods in the Ancient World: Legal Documents in Ancient Societies, Leuven/Paris/Walpole, Peeters (OLA, 229), 2014, t. 3, p. 103–118, at 107.

54 Katelijn Vandorpe, “Zenon son of Agreophon”, in Katelijn Vandorpe, Willy Clarysse, Herbert Verreth (eds.), Graeco-Roman Archives from the Fayum (Hell, 6), Leuven, Peeters (Hell, 6), 2015, p. 455.

55 Coussement, “Because I am Greek”, op. cit., p. 153.

56 Ibid.

57 The Ptolemies clearly favored those population groups that promoted Greek culture: for a full list of groups with tax benefits, see Willy Clarysse, Dorothy J. Thompson, Counting the People in Hellenistic Egypt, Cambridge, Cambridge University Press (Cambridge Classical Studies), 2006, p. 123 ff.

58 See e.g. Katelijn Vandorpe, “Persian Soldiers and Persians of the Epigone: Social Mobility of Soldiers-Herdsmen in Upper Egypt”, Archiv für Papyrusforschung, 54, 2008, p. 87–108.

59 Yiftach-Firanko, “Did BGU XIV 2367 Work?”, art. cit., n. 17, p. 108.

60 As noted above, villagers could also add their origin, but collecting all designations of origin is beyond the scope of this paper.

61 P. Oxy. 27 2476, l. 22 (TM 17010, AD 288).

62 Richard Alston, The City in Roman and Byzantine Egypt, London/New York, Routledge, 2002, p. 138–140.

63 Giovanni Ruffini, “Genealogy and the Gymnasium”, Bulletin of the American Society of Papyrologists, 43, 2006, p. 78–79.

64 Laurens E. Tacoma, Fragile Hierarchies: The Urban Elites of Third-Century Roman Egypt, Leiden/Boston, Brill (Mnemosyne, suppl. 271.), 2006, p. 42–43.

Table des illustrations

Titre Figure 1. Chronological distribution of individuals with a double name
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-1.jpg
Fichier image/jpeg, 195k
Titre Figure 2. Language combinations in double names
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-2.jpg
Fichier image/jpeg, 212k
Titre Figure 3. Chronological distribution of attestations of double names
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-3.jpg
Fichier image/jpeg, 219k
Titre Figure 4. Chronological distribution of texts containing identification clusters with a metronymic42
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-4.jpg
Fichier image/jpeg, 396k
Titre Figure 5. Chronological distribution of texts containing identification clusters with a maternal papponymic
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-5.jpg
Fichier image/jpeg, 244k
Titre Figure 6. Chronological distribution of ethnics in identification clusters51
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-6.jpg
Fichier image/jpeg, 242k
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-7.jpg
Fichier image/jpeg, 253k
Titre Figure 8. Chronological distribution of ἀπὸ X πόλεως in identification clusters
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-8.jpg
Fichier image/jpeg, 318k
Titre Figure 9. Chronological distribution of references to amphoda
URL http://books.openedition.org/psorbonne/docannexe/image/54847/img-9.jpg
Fichier image/jpeg, 270k

Auteur

KU Leuven

© Éditions de la Sorbonne, 2019

Conditions d’utilisation : http://www.openedition.org/6540

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search