Exploring the patchwork of personal names in ancient Cyrenaica with the help of digital tools

Abstract

We know plenty of personal names from Cyrenaica and most of them are known to us through inscriptions. Although the bulk of them are Greek, with some local dialectal features, we also find already in the Classical period some Libyan names. Later on, two new influxes become clear in local onomastics: Jewish and Roman names which appear in great number. In 1987 (LGPN 1), these names were collected, with the main emphasis on Greek, and were studied both from a linguistic and a sociohistorical point of view. The publication of digital corpora in 2017 and 2020 and other digital resources now make it possible to update the corpus and to search it much more easily. Some new results are presented here with the stress on Libyan and Jewish names.

Introduction

This paper will concentrate on the encounter and mutual influences of Greek, Libyan and Jewish names in Cyrenaica in the Greek and Roman periods. New studies, already published or currently in development, enhanced by digital tools, shed new light on the material, as will be shown with some case studies. We know plenty of personal names from ancient Cyrenaica and most of them are known to us through inscriptions that are mainly written in Greek. They reflect the historical development of the region. The very beginning of written tradition follows the settlement of Greek people who arrived in the last third of the seventh century BC from the Aegean, mainly from the island of Thera, with the help of Cretans and, probably, Rhodians. The first inscriptions are no earlier than the beginning of the sixth century and are written in a Dorian dialect, not similar to the dialects of any of the groups of settlers.¹ Its main features prevailed until the second century AD before merging into the panhellenic common idiom, the koine. Before the Greeks came, Libyan tribes with a nomadic way of life had been living in the region and they helped the Greeks, to some degree, to settle down in Cyrene and other places. Although the cohabitation was not continuously peaceful, there were certainly marriages between Greeks and Libyans, especially between Greek men and Libyan women.² This resulted in Libyan names being in use alongside Greek names, even in the upper layers of society. Later, two changes become clear in local onomastics: Jewish personal names are mainly documented from the second century BC on, as Jewish settlements took place during the Ptolemaic domination of the country. Eventually, Roman names appeared when the region came under Roman rule during the first century BC.

New tools for ancient historians

In the past, various stocks of names were collected, although with a focus mainly on Greek, in the first volume of P.M. Fraser and E. Matthews’ Lexicon of Greek Personal Names (LGPN 1), published in 1987.³ The names were carefully collected from the existing publications and benefited in many cases from J.M. Reynolds’ revised and, at the time, published and unpublished readings. That important step has resulted in useful studies from both a linguistic and a sociohistorical point of view.

Thirty years later came the long-awaited corpora, published online: in 2017, Inscriptions of Greek Cyrenaica together with Greek Verse Inscriptions of Cyrenaica (https://igcyr.unibo.it/), by the present writer, and three years later Inscriptions of Roman Cyrenaica 2020 (https://ircyr2020.inslib.kcl.ac.uk/en/) from Reynolds’ manuscripts and photographic archive.⁴ All three collections, complying with the academic guidelines of epigraphic corpora, were produced with the same methodology and their bibliographies are shared. They were combined early on into a larger project named InsLib (Dobias-Lalou et al. 2020). InsLib aims to provide the largest documentation on ancient Libya, developing compatible tools – the first one being the very useful Heritage Gazetteer of Libya (https://slsgazetteer.org/), which collects all place names, ancient and modern, mentioned in studies on ancient Libya.⁵ The online corpora, being the result of the knowledge of their authors at the date of their respective publications, should be supplemented periodically. Although reworking a digital corpus is a deep and demanding work, its publication is easier than a second printed edition of a book. In that view, a second edition of IGCyr has been prepared during the two past years and should be available when this paper is published (https://igcyr2.unibo.it/en/).⁶ These online collections provide an updated stock of names, which can be searched easily using their respective indices.⁷ All names already mentioned in the LGPN 1 may be checked and many more are registered. Moreover, a simple click on the mention of an attestation in an index leads to the published, and in most cases revised, text with illustration, translation and commentary.

Meanwhile, from the material gathered in the printed volumes of LGPN (volume I and the seven other volumes hitherto published), covering a large part of the Greek-speaking ancient world, a databank was created, supplying the LGPN online database (https://search.lgpn.ox.ac.uk/index.html) which allows various search facets and makes the work much easier without the need to work in a library.

The study of personal names from Cyrenaica did not await the development of modern technologies. During the twentieth century, Louis Robert and, above all, Olivier Masson were deeply devoted to this topic and provided notable studies. Nor did the study of dialect neglect personal names (Dobias-Lalou 2000). The publication of the printed LGPN 1, although not dedicated to explanations about the origin, declension and history of the names, was an important touchstone in the field. A further step in the knowledge of personal names proceeds from the LGPN-Ling project, initiated in 2015 and currently under development. Under Sophie Minon's direction, a team is studying all Greek names in the LGPN as linguistic material. The focus is on etymology (study of the lexical components of a name) and semantics (a personal name, like a word, is meaningful). As with LGPN, the results will be available both in print and in a database, although the chronology of publication is different from that of LGPN: the website LGPN-Ling (https://lgpn-ling.huma-num.fr/about.html) already provides work in progress on the whole alphabet, whereas the first printed volume was published in June 2023, covering alpha to epsilon (Minon 2023). Hyperlinks between the LGPN database and LGPN-Ling allow us to shift easily from one resource to the other.

Libyan names in Cyrenaica

Libyan names have often been interpreted in a questionable way. A Libyan name does not prove that the person bearing it was considered a Libyan in his or her environment. We should distinguish between the linguistic approach (identifying a name as Libyan) and the socio-historical approach (the person, the time, the relations of the person). An important landmark, in spite of its somewhat misleading title, was a study by Masson (1976), who provided a classified gazetteer of names from Cyrenaica which are non-Greek and seem to be Libyan. We owe further insights and tentative statistics to Colin (2000, 130–39). For identifying a name as Libyan, the problem is that we have no written sources in Cyrenaica parallel to the inscriptions in Libyco-Berber script known in the West. However, it is agreed that the language of the natives of this eastern area was akin to that of the western area, both belonging to the vast family of the Afro-Asiatic (formerly known as Hamito-Semitic) languages. Thus we rely on similarities with the documentation from northern Africa, plus some items from Tripolitania. For the names in those regions, a useful source is the catalogue by Rebuffat (2018), which updates the indices of Chabot (1940).

With a wide prospect, the LGPN collection is based not only on inscriptions from the region and from abroad, but also on literary sources, coins and papyri. However, as far as Libyan names are concerned, the regional inscriptions are our main source. As Libyan names mentioned in literature, we have one instance of Αννικερις (Annikeris), the name of a Cyrenaean who paid for the freedom of Plato then on sale as a slave at Aegina (Diogenes Laertius, De clarorum philosophorum vitis 3, 20). From Herodotus (Histories, 4, 164) we hear of Αλαζɛιρ (Alazeir), a king of Barka and stepfather of king Arkesilas III of Cyrene about 525 BC. Αλατ<τɛιρ> (Alat <teir>), name of an officer in charge of the mint at Barka at the end of the fifth century BC is a plausible variant of the same name.⁸ Both Αννικɛρις (Annikeris) and Αλαδδɛιρ (Aladdeir) are also attested in inscriptions of Cyrene.⁹ Very few amongst the Cyrenaeans mentioned in inscriptions abroad or in Egyptian papyri bear a Libyan name. One deserves mention in Crete: Σɛμηρ Διογένɛυς (Semer son of Diogenes) of Cyrene is mentioned at Phaistos (IC I.xxiii.2, end of the third century BC). Another one is Ιτθαλλαμμων Ἀπɛλλᾶ (Itthallammon son of Apellas) of Ptolemais who acted as ambassador to Rome in 23 BC (IG XIV, 1122). And amongst the many people said to be of Cyrenaean origin mentioned in Egypt, either in inscriptions or in papyri, none bears such names (Berthelot 2016). The only source with more numerous Libyan names comes from Marmarica, the easternmost part of Cyrenaica in the Hellenistic period, which was perhaps part of Egypt at the time of this document, about AD 190: in the so-called cadastre (Norsa and Vitelli 1931) the mentioned names are Greek, Latin, Libyan and rarely Egyptian.

The publication of the online epigraphical corpora enables us to update the lists of already known Libyan names with the help of their indices. The indices ‘Personal names’ allow us to find further occurrences of a name already identified as Libyan, such as Ταβαλβις (Tabalbis): instead of two occurrences registered in LGPN, they amount now to three with IGCyr121400. As for Αριπαχθις (Aripachthis), registered in LGPN with one occurrence from Ptolemais, IRCyr2020 added eight further mentions of this name, corresponding to six or seven persons, all in the western part of Cyrenaica. The oldest instance is in a family tomb in Berenice, at the very beginning of the Roman period (IRCyr2020 B.57 and B.68). The latest instances are graffiti in forts of the defensive belt south of Berenice and east of Taucheira (IRCyr2020 M.45; M.59f). In the city of Taucheira, one example is an ephebic graffito published as IRCyr2020 T.234, from the southern city wall, not to mention a further unpublished example.¹⁰ Another case is that of Ιαρθαμμων, plausibly a theophoric name. Registered with three instances in LGPN, it should now be divided into two names (or two variants of the same one): Ιαρθαμμων (Iarthammon) is attested at Ptolemais (IRCyr2020 P.38 and P.60, first century BC or AD), but at Cyrene in the Hellenistic period we have Αρθαμμων (Arthammon) with two occurrences possibly in the same family (IGCyr 065210.b.23 and partly restored at b.18, about 270 BC). In the less conspicuous group of names in -l, Ιαλ (Ial) now has three further occurrences in the Roman period (IRCyr2020 M.67; P.150; P.331).

The same indices include hitherto unknown names evidenced after 1987. For instance, among the group of names ending with -an, Αραταχαν (Aratachan), registered as dubious in LGPN, is now confirmed at Taucheira (IRCyr2020 T.156.iii with J. Reynolds’ clear photograph) and it has probably two further occurrences.¹¹ Libyan names may also have been misdirected. Such is the case of Βαρθυβας (Barthybas). A preliminary remark: Libyan names in -an were either associated with Greek names in -an (genitive -antos) or inflected like the more frequent Greek masculine a-stems of the first declension, with nominative -as and genitive -a (Dobias-Lalou 1999; 2017, 483–86). In this group, we should now register Βαρθυβας, which has often been erroneously considered Jewish. First attested in Cyrene among ephebes of Augustan times (IRCyr2020 C.145 b, i.17 and ii.30), it had been thought Jewish by the first editor (Pugliese Carratelli at †Oliverio 1961, 26), especially since the names Ἐλάζαρ (Elazar) and Ἰησοῦς (Iesous) are mentioned in the same list, and a Jewish etymology was defended by Applebaum (1964, 292; 1979, 177). However, Lüderitz (1983, 7) argued that the name should rather originate from northern Africa, as the only other instance known to him was from Egypt. In fact, this is a dedication to Sarapis, Isis and Heracles by Barthybas on behalf of his parents Polianthes and Bacchis (I.Delta 246.17, Canopus, second century BC). The father's name is typically Cyrenaean, so this man was probably a Cyrenaean enrolled in the Ptolemaic army or a descendant of such. A third and more ancient occurrence is now available in a list of names from Cyrene dating to the last third of the fourth century BC (IGCyr 131500): the father's name plausibly restored as Μɛλάνι[ππος] (Melanippos) testifies once more to the integration of Libyan names into the mainly Greek onomastic stock. Barthybas is plausibly a Libyan name in -an incorporated into the Greek names in -as.

If they do not belong to the main streams in -an, -is, -r or -l, names of Libyan origin are more difficult to identify. Such is the case with the father's name of Σῶσις Βιβιου (Sosis son of Bibios or Bibias), who dedicated a tithe to Apollo in the sanctuary at Cyrene. This inscription was not included in the first edition of IGCyr in the erroneous belief that the father's name would be the Latin Vibius. However, no tithe dedication is known to us after ca. 100 BC and the lettering of the inscription is clearly pre-Roman. It is now published as IGCyr² 137100. A first problem is that the form of the nominative cannot be confirmed with full confidence: as the genitive ending -ου is non dialectal and contrasts with dialectal δɛκάταν (dekatan) at line 3, a nominative Βιβιας (Bibias) is possible along with Βιβιος (Bibios). As to the stem, there exist on the one hand two isolated names in Greek that have been tentatively linked to βιβάζω ‘to cause to mount’ (often about sexual intercourse), although they might also be explained as coming from other languages: Βιβακος (Bibakos) in Sicily from Latin, Βɛιβους (Beibous) in Macedonia from Thracian (Minon 2023, 200). On the other hand, inscriptions from Tripolitania mention names echoing our Bibi-: in the dedication of the mausoleum of Qasr el Doueirat, near Lepcis Magna, the mother's name is Velia Longina Bibai (IRT2021 729, second–third century AD); a man in a Latino-Punic inscription of the Eastern Djebel is Bibas Masierkar (IRT 2021 877a); a man in another inscription at Qasr Doga is Bibi Mythulinim (IRT 2021 873); both Bibas and Bibi are considered Libyco-Berber (Kerr 2010: 191 and 197), although previous commentators had interpreted Bibi as Latin Vibius.¹² Moreover, Bibba at Mactar and BBY at Dugga offer parallels (Rebuffat 2018: 10). It is therefore much more plausible that Bibios/Bibias, the father at Cyrene of Σῶσις (Sosis), who had himself a traditional Greek name in the Hellenistic period, bore a Libyan name adapted into Greek.

In the new corpora, the indices ‘People’, reflecting the same method as LGPN, make it possible to search relations between parents and children, namely to investigate how Greek and Libyan names were distributed between generations. If a father with a Libyan name had a son with a Greek name, this might mean that the father was a Libyan man naturalised as a Greek citizen and the Greek name of his son would be an indication of his being integrated. However, a process of naturalisation of native people was probably rare at least in cities of the Hellenistic period, when society seems to have been rather conservative. The use of Libyan names in the highest levels of society was more probably the result of mothers’ influence: sons of Cyrenaean citizens and Libyan women from the area inside the boundaries of the Cyrenaica were themselves citizens, as prescribed by Ptolemy I (see n. 2). Consequently, a pattern with a father bearing a Greek name and a son bearing a Libyan name might be the result of such marriages. During the fourth and third centuries Αννικɛρις (Annikeris) is the name of three men bearing paternal names Ἐτέαρχος (Etearchos) [--β]ωλος ([- -b]olos), plus one lost name. It is also the patronymic of five men (eight occurrences) bearing Greek names, amongst whom proposopographical arguments allow to group Φίλων (Philon), Καλλίμαχος (Kallimachos) and Πɛιθαγόρας (Peithagoras) as brothers (Laronde 1987, 112–13; 118), whereas Ἑλλανίδας (Hellanidas) and Λῦσις (Lysis) are possibly, but not necessarily, brothers. A relationship between the two groups has not been demonstrated. In the first family, one was an eponymous priest of Apollo and his brothers were high-ranking military officers. The two men in the second group contributed to a public subscription, with Lysis paying up to 1000 drachmas. All were prominent people in their home city and the first Annikeris was a wealthy man.

The aforementioned name Tabalbis was used in a less prestigious, but still well-off milieu. One man bearing that name was father of a Νικόστρατος (Nikostratos), acting as damiergos, an important officer, responsible for the management of Apollo's estates in about 335 BC (IGCyr 011400.3). Another one was father of Τɛλέσανδρος (Telesandros), whose fine funerary monument is dated to the third century BC (IGCyr 121400). The third one was father of Ἄριστις (Aristis), a sculptor in the first half of the second century BC (IGCyr 077100).

Instances of mixed Greek and Libyan names in families are more numerous in rural areas. One interesting case is a list of names of unknown purpose found at Al Qubba (ancient Kelida), in the countryside east of Cyrene and dated to the end of the second or the beginning of the first century BC (IGCyr 033100). On three sides of this stele were inscribed at least 84 pairs of name and father's name. Most of them are Greek names, sometimes with an inaccurate spelling which plausibly reflects the actual pronunciation. Sons with a Greek name and Libyan patronymic are Ἀρίσταρχος Ιγισαντος (Aristarchos son of Igisan, 75–76), Κάλλιππος Βακαλος (Kallippos son of Bakal, 160) and even Αριμαν Ιθαννυρα (Ariman son of Ithannyras, 55–56), where the Libyan patronymic is integrated into the Greek a-stems, whereas the Greek name is modelled upon the Libyan type in an. Sons with Libyan name and Greek patronymic are Μιτταχις Ἀρχɛδάμω (Mittachis son of Archedamos, 73–74), Ανυσαν Ἀριστάρχω (Anysan son of Aristarchos, 135), Ιθαννυρας Ἀριστάρχω (Ithannyras son of Aristarchos, 140), Μɛιραχαν Τɛιμάρχω (Meirachan son of Timarchos, 151). The population at Kelida was to some degree a mixed one.

In the western cities, especially Ptolemais and Taucheira, personal names are documented in less formal formats dated to later periods. We have there many graffiti by ephebes; tombs are simpler there than in Cyrene and hosted many people, whose names were inscribed on the rocky façades. The Libyan names attested during the preceding period are still in use, mingled with Greek names and also with Jewish names.

Jewish names in Cyrenaica

Jews settled in Cyrenaica under the Ptolemies. Flavius Josephus’ statement (Contra Apionem 2.44) that Ptolemy I sent Jews to Cyrenaica is not supported by local sources and any such Jews might have acted only as soldiers in garrisons. Jewish presence is best established for the period beginning when Egypt and Cyrenaica were again politically united under the second Euergetes’ sceptre after troubled times, i.e. after 145 BC. For the period up to the Jewish Revolt in AD 117, inscriptions document both the life of Jews’ own communities in Berenice and Cyrene and their participation in civic life among the other citizens, either as ephebes, being educated as future citizens, or as officers, acting for instance as nomophylakes. If not settled by one of the Ptolemies, the Jews of Cyrenaica may have been connected with the Hellenised Jews who left Palestine after the Seleucid conquest. In Cyrenaica we have hardly any inscriptions in Hebrew. The language used for inscriptions publishing official decisions, as well as the vernacular language of epitaphs and graffiti, is Greek, koine rather than the earlier dialect, and there we find Jewish and Greek names side by side.¹³

One honorific decree of the Jewish community of Berenice in AD 55 (IRCyr2020 B.45) mentions the names of 18 donors who contributed towards repairs to the synagogue. They include the priest, whose name and patronymic Καρτισθένης Ἀρχία (Kartisthenes son of Archias) are Greek, ten men with a higher rank, labelled ἄρχων (archon, something like ‘head’) and seven other people, including two women. The onomastic formulas of those 18 persons provide a total of 30 different names, of which only two are Jewish: Ἰωναθᾶς (Ionathas) and Μαρίων (Marion).¹⁴ One name is Latin: Cornelius. Some Greek names are known to have been favoured by Jews, because they translate Jewish names: in this list we find the most common are Δωσίθɛος (Dositheos, four occurrences) and Ἰάσων (Iason), which, however, had already been a traditional name in the region before the Jews came; perhaps also Θɛύφιλος (Theuphilos).¹⁵ All other names are common Greek names, some of them favourites of the region, such as Εὐφράνωρ (Euphranor), Καρνήδας (Karnedas), Πρᾶτις (Pratis), Πρατομήδης (Pratomedes) and furthermore Ἀμμώνιος (Ammonios), a Graeco-Libyan name, insofar as the god Ammon of Siwah was considered a regional Libyan god in Cyrenaica, different from the Egyptian Amon. The cultural mixture is so strong that we also find a series of names referring to pagan deities, such as Zeus in Zηνόδωρος (Zenodoros) and Zηνίων (Zenion), the Egypto-Libyan Isis in Ἰσίδωρος/Ἰσιδώρα (Isidoros/Isidora), Herakles in Ἡρακλɛίδης (Herakleides) and the Graeco-Egyptian Sarapis in Σɛραπίων (Serapion). It is difficult to decide whether these names were chosen for their sound alone and were not semantically analysed or whether they were considered ‘very Greek’ and thus retained for the sake of social integration. Similar statements are provided about the village of Trikomia in Egypt by Clarysse and Thompson (2006, II.321). At Berenice, two other decrees from the same community (IRCyr2020 B.75 and 76), although containing fewer names, present similar issues.

From the Hellenistic period onwards, most inscriptions at Cyrene which are labelled as lists of ephebes are actually parts of dedications to Hermes and Herakles, the two gods of the gymnasium, made by the ephebes who honoured their teachers and trainers. In a list of 80 names dating to AD 3/4 (IRCyr2020 C.145) three young men bearing typically Jewish names are mentioned amongst the bulk of Greek names. They are one Ἰούλιος Ἰησοῦτος (Ioulios son of Iesous) and two brothers, Ἐλάζαρ (Elazar) and Ἀγαθοκλῆς (Agathokles), sons of Ἐλάζαρ (Elazar). Latin, Greek and Jewish names were intertwined in those families. Many graffiti were added by various hands upon other sides of the stele by later users of the gymnasium, such as Χαιρέας Ἰούδα (Chaireas son of Ioudas, 3.13), who, given the lettering and the place chosen, seems to have done this shortly after the erection of the monument. Other similar dedications by ephebes are known from the second part of the second century AD and, of course, no longer feature Jewish names. A marble base with a different scope may also be dated to the Augustan period thanks to its lettering. Here, one Τίμαρχος Zɛυξιμάχω (Timarchos son of Zeuximachos) dedicated to Hermes and Heracles a list of names, presumably of ephebes (IRCyr2020 C.431–32).¹⁶ From the 18 onomastic formulas still readable, we get 24 different names. Only one, Ἰησοῦς, is unmistakably Jewish. Additionally, Ἰάσων, Φίλων and Εἰρηναῖος are often, but not necessarily, associated with Jews. In contrast to the presence of Jewish ephebes at Berenice and Cyrene, a list of ephebes from Ptolemais, which is dated a few years earlier, contains a good series of Libyan names, but no Jewish name.¹⁷ It was published with thorough comments on onomastics by Reynolds and Masson (1976), now IRCyr2020 P.79. At Cyrene, the presence of Jews in civic life is also attested in a list of nomophylakes in office during the two years AD 59/60 and 60/61: Ἐλάζαρ Ἰάσονος (Elazar son of Iason) is one of them (IRCyr2020 C.98.8).

At Taucheira, many graffiti were scratched on the walls of the gymnasium and on blocks reused for the late city wall. No essentially Jewish name may be read there, while instances of Greek names used by Jews are very few and are not necessarily related to Jewish people: Θɛόδοτος (Theodotos), Δωσίθɛος (Dositheos) (IRCyr2020 T.100; 134.v; 146.iii; 148.iv), Ἰάσων (Iason) (T.254; 710.i). The names are Greek and Latin. The paucity of Jewish names in an ephebic context is in contrast with the well-attested Jewish families buried in the same city. More generally speaking, Jewish names are recorded in funerary inscriptions all over the region, in cemeteries of towns and smaller country settlements. In such cases we again find the two groups of names already mentioned: names transliterated from Hebrew and Greek names favoured in the Jewish milieu. In the second case, further arguments are necessary for the classification. For instance, in the many family tombs at Taucheira, a menorah cut on the rocky wall is a decisive argument. The presence of several epitaphs with ambiguous names in the same tomb or in neighbouring tombs may also help to decide. In that city, numerous Jews were buried during the first century AD and the beginning of the second. For instance, on the façade of a tomb in quarry East XV, seven epitaphs were inscribed.¹⁸ Six examples clearly refer to the same family, all bearing the patronymic Ἰοάνους (son of Ioanes), except for one Ἰωσήπου (son of Iosepos). Some children themselves had Jewish names: Μάρθα (Martha), Ἰουδίων (Ioudion), Ἰοάνης (Ioanes). Three girls had Greek names: Φιλοξένα (Philoxena), Τρύφαινα (Tryphaina) and Καλλώ (Kallo). We cannot guess what sort of relationship existed with one Φίλιππος Φιλίππου (Philippos son of Philippos) whose epitaph, very similar in shape and layout to the other ones, was inscribed amongst them.

Conclusion

Jewish personal names in Cyrenaica reflect a very different situation from that of Libyan names. Jewish names in some cases resisted the Hellenisation of people who belonged to various layers of society, until the Jewish revolt of AD 117 put an end to their presence in Cyrenaica.¹⁹ Libyan names had been adopted much earlier, in the first period of the Greek presence in Libya, and became part of an essentially Greek culture, so that they did not disappear until the arrival of the Arabs in the 640s. A very useful document on this point is the P.Marm.R. (Norsa and Vitelli 1931). In this document, the estates are identified with the tenants’ names, where Greek and Roman names are the most numerous. However, some Libyan names are still to be found, such as Ανυσσαν, Ιαφθας, Ιθαννυρας, Μιτταχις, Υραθθις.