Version classiqueVersion mobile
OpenEdition Books

Regionalism in Hellenistic and Roman Asia Minor

Hugh Elton
Gary Reger

The Impact of Rome on Local Naming Practices in Asia Minor: a Regional Perspective

Andrew P. Gregory

Note de l’éditeur

Cases in parentheses of persons with more than one name

Texte intégral

1The processes of cultural change in Roman Asia Minor appear to have been broadly twofold: (1) a tendency towards a vague kind of homogeneity, with the spread of koinê Greek, recognizably Greco-Roman architecture, Hellenic cults, military service and so forth; and yet at the same time (2) a continuing cultural fragmentation, not just on a regional basis but sub-regional as well. Of course within any given region there is going to be cultural variation (as Levick has so eloquently demonstrated; see chapter 11). Tidy minds will set these cultural variations in binary opposition: town versus rural culture; élite versus popular and so on. Not so in real life. For instance, while the rural inhabitants of Roman Asia Minor might not share or participate in the élite culture of the cities (the world of erudition and philosophy, for instance, or the cult of the gymnasion), many did on occasion enjoy the general popular culture of cities: the great festivals, games, and public banquets. Regional culture must needs be complex and multilayered: Greco-Roman society, though highly stratified, consisted of numerous sub-groups, citizens (local, foreign, and Roman) and non-citizens, slaves, freed and the free, men and women, soldiers and civilians, all of whom had their own cultural and social identities. Within towns, for instance, associations and guilds provided their members with a distinctive sub-culture of their own. Region-wide culture is thus affected by the sum of these varied influences, as well as patterns in language use, and even ecological and environmental peculiarities. Finally, the contact with royal and imperial courts and their representatives have to be added to the picture.

  • 1 Green 1990, 324. This is exactly what Tacitus described among the British as honoris aemulatio (Agr (...)

2My concern is to illustrate the complexity of the cultural scene in northern Lydia and to attempt to isolate some of the factors which affected cultural change and thus identity within a region. One useful way of doing this is to look at onomastic practice and the distribution of personal names in specific geographical areas. Personal names express cultural identity since they reflect family relationships and family sentiment, the predominating religious and/or agricultural interests of native peoples, and occasionally their literary aspirations and imagination. The introduction of a Greek, Makedonian, or Roman name into an indigenous Anatolian family has significance not only according to when and where it is introduced, but also according to the social status of the family concerned. We should not rule out a multiplicity of motives behind the spread of new names and name changes in Roman Asia Minor, but underpinning this investigation is the conviction that for many individuals-certainly the village élites-the changing of a name or the use of a new one should not be readily ascribed to a vague “philhellenism” (or “philromanism”) but to what Peter Green called a “steely determination to get on in the world”1.

  • 2 Cf. Apollonios, Ep., 71.
  • 3 The inscriptions came from the following sources: Malay 1994; Petzl 1990; Çelgin & Malay 1986; Naou (...)

3Writing at the time of Augustus Strabo remarked that under the Romans most of the peoples of Asia Minor lost both their dialects and their names (Strab. 12.4.6)2. The region normally identified as northern Lydia is a good region to test the validity of such comments because of its wealth of epigraphic documentation. In the Middle Hermos valley, tombstones and dedications are frequently dated and this evidence provides an opportunity to assess the impact of Greek and Roman names in town and country within a limited geographical area and over a finite period of time. The dated funerary stelai give a good impression of the overall structure of naming habits in this particular region in the Imperial period. Altogether we have the names of 862 individuals, of whom 762 are dated to before the year 212 CE3.

  • 4 For Persian settlement in Lydia in general, Sekunda 1985.

4The main categories of names appear broadly to be Greek, Roman, Lallnamen, and a motley group of indigenous Anatolian anthroponyms. Striking is the almost complete absence of Persian influence (cf. Thelymithrês), although it is echoed artificially in the names Persis and Persikos4. The Greek names can be further divided into general, theophoric, and Makedonian/military types. The category “Roman” can be sub-divided into Roman proper (genuine holders of civitas and bearers of Roman gentilicia) and “Romanizing” (see below).

Name Type



Greek general



Greek Theophoric









Other Anatolian











Table I. Personal names by type.

5The first question is whether the “indigenous character” of the region, so marked in its cultic activity and reflected in its traditional toponyms, is present in its onomastic practices. In fact at Saittai there are indigenous names but these are largely of the Lallnamen type, common in many parts of central Anatolia, especially Phrygia. Names such as Ammia, Apphia, Papias, Tatias, and all their variants, make up 120 instances or almost 14 percent of the total of 862, as indicated in Table 1. Together with the other indigenous names the Lallnamen make up well over 16 percent of all dated names at Saittai. While the provenance of some of the tombstones is uncertain it is clear that the vast majority of those with Lallnamen are found in village or rural contexts. The use of these names continues uninterruptedly from the first century through the third, though with a gradual decline setting in during the course of the second century onwards (Table 3). This evidence points unambiguously to the fact that, as other scholars have observed, culturally speaking this part of the Hermos valley belonged to Phrygia. The gradual decline in the use of Lallnamen among the peasant population, along with the decline in other indigenous names (such as Mata, Paterês, Deskylos), can be explained by increasing Hellenization, expressed by the commensurate increase in Greek anthroponyms.

6Before we turn to the general Greek anthroponyms let us look at theophoric personal names. Clearly detectable, as one would expect, is the influence of religious cults on local naming patterns. The breakdown according to cult is represented in Table 2.


Number of cases

% of All Theophorics





36 (1)














Mother goddess




9 (1)













Table II. Theophoric names at Saittai.

  • 5 Petzl 1995; Keil 1923.
  • 6 See Petzl 1994.

7There is a close link between theophoric personal names and local cults, and a less distinct but nonetheless tangible link between similar names and cults attested at the regional level5. There are 43 names based on Apollo and 20 on Mên, including two from the earliest dated inscription on Saittan territory (65/64 BC). Altogether theophoric names attest for 181 of the dated names in the area of Saittai, about twenty-one percent of the total names in the catalogue (Table 1). If we combine the indigenous, non-Greek names with these theophorics together they make up well over a third of all dated names in the area. Yet while the indigenous and Hellenized cults of the region had a considerable impact on naming habits in town and country, the catalogue of personal names tells us little about changes in religious sentiment. Although theophoric names are attested well into the third century, there is a steady decline in the use of theophorics at Saittai, from 38.5% before the year 70 to only 9% in the period after 212 (Table 3). How are we to explain this shift? Perhaps as no more than a change in taste. Even if the theophoric names could be said to disappear in the later third century, no distinctly Chrisianizing names appear in their place. A decline in interest in local cults may not be the answer: we know that the Confession stelai, so characteristic of the region, continued to be set up through the third century6.

Table III. Personal names by type and period

  • 7 Strubbe 1984-1986; Spawforth & Walker 1985 and 1986.

8Very popular Greek names, apart from the theophorics, were Alexandros (33), and the forty-seven names containing the element nike, which can be broadly grouped together as “Makedonian/military.” Such names, as well as conveying notions of personal success, seem to reflect in a somewhat trite fashion the martial spirit of ancestors who may have been ‘Makedonian’ veterans and colonists. What these names indicate, beyond the simple habit of repeating family names across generations, is that certain segments of both the town and country populations felt that they had some claim to a “Macedonian” ancestry, or connections to military settlers of former times. Yet the figures in Table 3 make it quite clear that the use of Makedonian names was a creation of Imperial times, for there is only one dated case before AD 70. The dramatic increase in Macedonian names in the Hermos valley in the Roman Imperial period can be seen as part of the general interest shown by Anatolians of this period in their Greek cultural heritage, and reflected for instance in the development of the Pan-Hellenion or the spread of foundation myths7.

9The remaining Greek names include those which connote good news, success and good fortune, benefit or advantage, hope, reward, and the easy life. The few that echo literary names may, in rural contexts, reflect the attempt of the village élite to distinguish themselves and their dependents from the ordinary folk. Above all, Table 3 shows that these kind of general Greek names gradually increased in number and came to dominate in the Hermos valley in the Imperial period at the expense of theophorics and Lallnamen.

  • 8 Hatzfeld 1919, 120, 165; Broughton 1938, 881.

10If the dated personal names at Saittai indicate a basic blend of a limited Greco-Makedonian influence, with Greek theophorics adopted for the traditional deities of the region, and a strong persistence of indigenous Lallnamen in the villages, what do they tell us of the impact of Rome in the region? Roman citizens before the year 212 are only sixteen in number (representing just about two percent of the total of 762 individuals), and only one Roman citizen is found in a village context before 212: C. Oulpios Nereus who was enfranchised under Trajan or whose ancestors were enfranchised by Trajan’s father as proconsul. The evidence points strongly to an Italian presence at Saittai in the Late Republic and Early Empire8. Yet the figures in Table 3 suggest that despite this presence of Roman officials and businessmen Roman citizenship spread slowly at Saittai.

  • 9 Cf. Gregory 1997a, 554; Ramsay 1898, 338; Schulze 1904, 508. On Roman citizenship in Asia Minor see (...)

11More common then Roman citizens with the tria nomina are folks who possess what I term “Romanizing” personal names, by which I mean Latin names assumed by non-Romans. At Saittai before AD 212 these names represent about seven percent of all dated names (54 out of total of 762). Roman influence comes strikingly early: the earliest is the name Syllas (= Sulla) actually found in a village context in 18/19. Despite the painfully slow spread of the civitas and thus of the legal use of the tria nomina, there is substantial evidence for the widespread usurpation of ‘Romanizing’ names9. The practice was common in many provinces of the empire, and was observed in certain areas such as Sicily already in Late Republican times.

  • 10 Hatzfeld 1919, 330.
  • 11 Knapp 1978, 190; Hahn 1906, 64, 158.

12For some Anatolians, as for other provincials, Roman citizens, whether Italians or not, and whether businessmen, officials or soldiers, must have appeared as an influential group of outsiders10. Thus the effect, if not the purpose, of these ‘Romanizing’ names among local populations is obvious. They advertised a friendly disposition towards Roman power and, within the local context, they suggested that certain individuals had some tangible connection to that power, whether it was real (as clients of Romans, for instance) or not11.

13In looking at the Romanizing names it is worth making distinctions where possible between the status of the individuals who possess these names before 212 and speculating on the ways in which they may have acquired their names. While some Romanizing names were given to lower class workers or freedmen in civic contexts, the most salient fact of all is the frequency with which Romanizing names appear in the rural context. The exact status of these villagers in their own communities is sometimes uncertain, but their names surely reflect deliberate attempts by certain families to advertise a special status, and, I think, to reinforce their own position in local society, vis-à-vis a new social structure and new power structure imposed from outside in the shape of Romans and Italians. I am inclined to see the members of village élites–who were directly and regularly exposed to Roman administration–playing the crucial rôle of introducing this naming habit into and maintaining it in their communities.

  • 12 Petzl 1994, no 35 (210/211).

14Whether a descriptive Roman cognomen actually meant something concrete to the Greek speaking locals is uncertain. Did the parents of the two girls named Gemella know the name’s meaning in Latin? Here is not the place to discuss the spread or use of the Latin language in the East but it seems likely from this sprinkle of evidence that some Anatolians were familiar with some basic Latin vocabulary, though for every Anatolian who knew a word or two of Latin there was one of two or more in fact who knew none. In time, of course, the use of Romanizing names by other villagers came about, perhaps through deliberate imitation of the village élites and others, or through increased popularity or daily use. But the important factor must have been direct and regular contact between certain sectors of the local population and the representatives of an external élite, the Roman publicani, negotiatores, officials, and soldiers. Moreover, Romanizing names were never adopted wholesale, at the expense of other family names, which continued to be passed on from generation to generation. Traditional theophorics could sit comfortable side by side with imported names. The names of the brothers Phlabianos and Mêtrophanês, sons of Philippikos, at the Lydian village of the Tazênoi in the early third century, are in some ways typical of the blend in the Imperial period12.

  • 13 This paper is based on work presented in my doctoral dissertation, Gregory 1997c. Thanks are due to (...)

15The onomastics of Saittai indicate that the Middle Hermos valley belonged indisputably to a Phrygian cultural milieu. Phrygian identity steadily declined in the Imperial period; presumably it was already waning in late Hellenistic times. Greek names reflecting came to replace traditional Lallnamen and theophorics. Roman influence was felt most sharply among the civic and rural élites, who were the first to adopt Romanizing names in the period between Augustus and the mid-second century. Thereafter such names become much more common, just as instances of full Roman citizens increased in number. The adoption of Romanizing names in the 150 years between Augustus and Hadrian was most likely the result of local élite families having prolonged contact with Roman officials and businessmen. The explanation for the different “types” of regional Greco-Roman societies–as we should really be describing them–which emerged in Anatolia must be sought in the varied experiences and the different histories which various regions possessed regarding exposure to Greek and Roman civic and military culture, and in the social and political needs of local élites and their relations with other groups, but also in the varied local sentiment, popular attitudes and deliberate choices made in the respective communities13.


1 Green 1990, 324. This is exactly what Tacitus described among the British as honoris aemulatio (Agr., 21).

2 Cf. Apollonios, Ep., 71.

3 The inscriptions came from the following sources: Malay 1994; Petzl 1990; Çelgin & Malay 1986; Naour 1985; Malay & Petzl 1985, 65; Tanyeri 1985; Naour 1984; Naour 1983; Malay 1982; Malay & Gül 1981; Naour 1981; TAM, 5, 1; Bakır-Barthel & Müller 1979.

4 For Persian settlement in Lydia in general, Sekunda 1985.

5 Petzl 1995; Keil 1923.

6 See Petzl 1994.

7 Strubbe 1984-1986; Spawforth & Walker 1985 and 1986.

8 Hatzfeld 1919, 120, 165; Broughton 1938, 881.

9 Cf. Gregory 1997a, 554; Ramsay 1898, 338; Schulze 1904, 508. On Roman citizenship in Asia Minor see Holtheide 1983.

10 Hatzfeld 1919, 330.

11 Knapp 1978, 190; Hahn 1906, 64, 158.

12 Petzl 1994, no 35 (210/211).

13 This paper is based on work presented in my doctoral dissertation, Gregory 1997c. Thanks are due to the editors of the volume and speakers at the conference for suggestions and comments. I remain solely responsible for errors.

Table des illustrations

Légende Table III. Personal names by type and period
Fichier image/jpeg, 143k


Independent scholar

© Ausonius Éditions, 2007

Conditions d’utilisation :