Onno Van Nijf Termessos Local Knowledge

1 Being Termessian: local knowledge and identity politics in a Pisidian city* Onno van Nijf

Getting to know them: the cemeteries of Termessos Anyone approaching an ancient city would first be confronted with the deceased members of the community. Each city buried its dead conspicuously along the main roads leading into the city. Streets of tombs and cities of the dead often surrounded the cities of the living, and the small city of Termessos, high up in the mountains of Pisidia was no exception.1 In fact due to limitations of space, the dead and the living inhabited areas in even closer proximity than was usual.2 Termessos, which is now the centre of a Turkish national park, was at the turn of the second and third centuries C.E. a thriving, though perhaps unexceptional provincial city. An old Pisidian settlement, it had been drawn into the Greek world only after the conquests of Alexander, and under his successors it slowly turned into a Greek city. It was a staunch supporter of Rome in the late Republic, and it maintained a high degree of independence. In the imperial period it was incorporated into the province of Lycia et Pamphylia. The site was never formally excavated, but it was explored by travellers, and surveyed by teams from Vienna and more recently from Istanbul which has resulted in extensive publications of its more than 1,000

*

This paper is part of a wider project on the study of the epigraphy and society of Roman

Termessos. I shall discuss the honorific spaces of the city and their connection with the political culture in a forthcoming paper: van Nijf (forthcoming). Cf. van Nijf (2000); van Nijf (2003b) I have presented versions of the current paper in Groningen, Paris, Hamburg, Nijmegen, Istanbul and Athens. I would like to thank my hosts at these occasions, as well as the participants in the seminars for their comments. I have greatly benefited from comments or help by Sofia Voutsaki, Rens Tacoma, Christina Kokkinia, and Christina Williamson. 1

For a discussion of this phenomenon see von Hesberg and Zanker (1987).

2

See the map. The cemeteries are immediately north and south of the city centre.

2 inscriptions.3 Termessos may have been unremarkable, but to us it is unique due to this exceptionally rich epigraphic record. The bulk of the inscriptions date to a relatively brief period of time around the turn of the second and third centuries C.E., which represented in many respects the akme of the city’s built history.4 The monumental texts on honorific statues and on funerary monuments served the epigraphic classes as a means of self-representation. I have explored elsewhere the implications of the rich honorific record, but in this paper I want to focus on getting to know the Termessians through the ways in which they represented themselves in the cemeteries of their city.5 Mortuary behaviour in general, and funerary inscriptions in particular, is a promising area to investigate the claims to status and identity in the world of the living.6 Funerary practices can be analysed as strategies of social, political, and cultural self-definition. In this context it is relevant to consider the social importance of the inscribed epitaph.7 Each epitaph was a deliberate and enduring commemoration of whatever features were seen as the dead person's most significant characteristics, the features that defined his (or her) social identity in life as much as in dead. Of course, there is a risk of partiality. Epitaphs could be economical with the truth, or thrive on hyperbole; in death many people became what they never were in life. Tomb inscriptions may, therefore, reflect desired as much as acquired status. But the statements that funerary inscriptions made had to be plausible, at least, and the cemeteries give us a good idea about the kind of cultural norms and values that were generally deemed important in the city.

3

The best discussion of Termessos remains Heberdey in RE (zweite Reihe) V.A.2: 737. The

inscriptions were published by Heberdey in TAM 3.1 and in Iplikçioglu et al. (1991), Iplikçioglu et al. (1992), Iplikçioglu et al. (1994) and Iplikçioglu et al. (2007). 4

RE (zweite Reihe) V.A.2: 739-747 (Heberdey).

5

Cf. above n. 1.

6

For a longer discusion of the methodological and theoretical background to this approach, see

van Nijf (1997) ch. 1. 7

world.

E.g. Meyer (1990) and Meyer (1993) for discussions relevant to the Roman and the Greek

3 The Termessian cemeteries, therefore, may provide us with an insight into the most essential qualities for which Termessians wanted to be remembered, and through which they sought to position themselves in the world of the living. In this chapter I shall discuss a few of the characteristic issues that can be observed based on the funerary material.8 Funerary display in Termessos, i.e. the monumental and epigraphical selfrepresentation, was in the first place a function of wealth and political status in the civic community. Expensive and ostentatious tomb monuments drew attention to the wealth and social standing of the Termessian top families. Their high status was underwritten by, as much as it was reflected in, conspicuous consumption in death. Yet the cemetery was no carbon copy of the city centre, and it is important to be aware of the possibility that the funerary self-representation complemented the monumental language of the city centre in significant ways. Funerary monuments, inscription and all, also seem to speak a language of belonging. By their very nature tombs and epitaphs locate individuals within the context of their family or extended family. The care and attention, and especially the amount of money that the Termessians poured into the funerary monuments of their relatives served to show their piety and proper respect for their ancestors, but also helped them to identify themselves publicly as the heirs of family traditions. As we shall see, the Termessian epigraphy inserted individuals with particular precision within exceptionally detailed genealogies that could extend several generations back. It seems most likely that we can interpret this ‘genealogical bookkeeping’ against the background of a process of oligarchisation, whereby status and influence at the local level came to depend on different forms of symbolic capital, including the claim to stand in a family tradition with strong local roots. And finally, the Termessian cemeteries were a locus where cultural identity politics were played out. The epigraphic evidence presents the Termessians as juggling multiple identities: they were able to present themselves as cultured Greeks, as loyal Roman citizens, but also as the proud descendents of indigenous warriors who had been fiercely independent for much of their history. It is through a study of

8

These issues are only a selection, of course. Other dimensions of Termessian identity include

gender and family relations, the discussion of which I shall have to leave to another occasion.

4 their onomastic habits in particular that we shall see how these different strands of their cultural identities co-existed, and were intertwined at an individual level as much as at the level of the group. Their experiences urge us to formulate a nuanced position in current debates on identity politics in the Roman empire. The Termessian cemeteries were, therefore, a repository of local knowledge: they served the living as a way to identify themselves by reference to the dead. But they also serve us in getting to know the Termessians as closely as we can from the distance, both geographical and chronological, that exists between us and them.

1. The Commemoration of Status Social distinction was certainly high on Termessian minds: the families of notables that ruled the city also dominated its landscape. Throughout the city centre we find buildings that were set up by leading families, for example a stoa (L2) along the central agora that was built by the benefacor Osbaros, or a gymnasium (H) that was built by a husband and wife team.9 However, one of the most striking aspects of the Termessian landscape must have been the omnipresence of honorific statutes that commemorated the members of the city’s elite as priests, magistrates, and benefactors; as loyal subjects of Rome, but also as cultured Greeks, and as dutiful wives or successful athletes. I have argued elsewhere that this monumental language put the elite literally and metaphorically on a pedestal. If you wanted to know who really mattered locally, you only needed to walk the city centre where the monuments provided you with a local ‘who’s who’.10 Although it is possible to make a tally of the various activities and qualities for which the Termessian elite wanted to be commemorated, it was apparently not always necessary to go into detail. Many honorific inscriptions do not list specific actions or achievements. At times it was deemed sufficient to list moral qualities or hint at personal excellence. In such cases we are often dealing with posthumous monuments that were set up by important families who must have successfully petitioned the city authorities to be able to erect a statue for their deceased relatives. The dividing line

9

TAM 3.1, 121, 122.

10

van Nijf (2000).

5 between honorific and funerary monuments was apparently somewhat blurred. These show that the top families were able to adapt public space for their own private commemoration.

Tomb monuments As the burial plots gradually merged into the built-up area, it is easy to see how the cemeteries were an extension of the public sphere. The same elite that ruled the city of the living also dominated the city of the dead by their conspicuous tombs and monuments. Cormack’s recent discussion highlights the tombs of a few of wealthy individuals who can all be traced to the best known families, such as Apollonios Strabonianos (stemma G17), a proboulos and the son of a civic priest of Zeus Solymeus, who built a tomb for his parents and his son. His wife, Tiberia Claudia Kille (stemma H6), stemmed from another prominent family that descended from a Tiberius Claudius Agrippinus.11 A generation later Claudia Agrippina married into this family. She had a temple tomb built for her husband, the proboulos Ti. Claudius Marcellus (stemma H12), in the northern cemetery.12Another tomb-builder was Aurelia Ge, the daughter of a Hermaios Hoples, who was married to a priest and gymnasiarch, Tiberius Claudius Plato (stemma D1); their descendants also served as priest and proboulos. These were spectacular tombs, built to impress. It has been noted that the inscriptions on the largest monuments were also certainly designed to impress and inform, as they were carefully laid out with large beautifully executed letters.13 Such monuments obviously drew attention to the wealth and command of resources their commissioners had that made possible and justified their leading role in society. However, funerary monuments were not simply an extension of the commemorative practices that we find in the city centre. We should not fall back to a 11

References of the type ‘stemma H6’ are to the stemmata of the major families of Termessos

that were drawn up by Heberdey in TAM 3. For a full discussion of each family, see Heberdey (1929). 12

The decoration on this tomb shows a marked military character (shields, weaponry), which

suggest that this theme was particularly important to Marcellus, or other (male) members of his family: Cormack (2004) 307. 13

Cf. Cormack (2004) 308.

6 simplistic isomorphism: we cannot not assume that all big tombs belonged to the elite- or that all members of the elite were buried in big tombs. Members of prominent families could have been buried in inconspicuous tombs, and there are also indications that individuals further down the social ladder were able to set up fairly conspicuous tombs as well, although they do not seem to have been a match for the families at the very top. Some members of elite families were certainly buried in simple sarcophagi. It should be remembered, moreover, that all tombs and monuments were an indication of relative success with resources to spare: even a small tomb served to distinguish the dead –or his family– from others who could not afford one. Funerary practices bespeak of status, and therefore of status aspirations, but not necessarily in the same manner as the public monuments in the city centre, and it is worth our while to investigate the similarities and differences in a little more detail. It is interesting to note that political success was not usually commemorated on the tomb inscriptions of the Termessian elite. Of course we find individuals who commemorated the offices they had held in life: priests and priestesses, and grammateis are on record. Yet their number is limited, and the Termessian top-elite is noticeable for its reticence in this regard. There are very few members of elite families who had their offices commemorated on their own tombs.14 For the real top families the city centre was a more likely area to advertise their political status. Epitaphs that did commemorate offices and priesthoods, therefore, often belong to a second level within the elite: grammateis of lower boards of officials, chreophylakes, and some priests - or magistrates and councillors who did not belong to the inner core of the Termessian elite.15 To them the funerary area was probably more important as an area of status display than the city centre, where they would be easily overshadowed by the Termessian top-elite.

14

Only Marcus Aurelius Polemon V (TAM 3.1, 730) must have been of relative high status: he

was grammateus of the boule and a keryx. He cannot be connected to a ‘top’ family, however. 15

The offices are listed in TAM 3.1, 342 index V.2.

7 In the few instances that we find references to the careers of local grandees, these were usually put up by family members or dependents.16 It would appear that to these men (and women) status and identity were to be found in making explicit ties with their higher ranking family members patrons. Quite a few epitaphs were set up by men and women who identified themselves explicitly as apeleutheros or apeleuthera of So-and-So. 17 These men and women were apparently not full citizens of Termessos, but they had the status of paroikoi. Freedmen of Roman citizens would be expected to enter into a clientela relationship with their former owners, and this seems to have been the case in Termessos as well. This could be a close relationship: in quite a few cases the freedmen identified his former owner with considerable detail - including the enumeration of his priesthoods or offices. At one level such texts advertised the social control to which the freedmen were subjected, but the freedmen could also make such texts their own by using them to stake out a claim of belonging. It was apparently worth something to be known as the freedman or freedwoman of Platon the priest.18 In some cases the closeness of the relationship was further underlined by the fact that the freedman’s tomb was located in the immediate or close proximity to that of their former masters. Slaves could do something similar: quite a few were able to set up tombs for themselves - and interestingly enough for their natural families- but they usually carefully stated that this was done only with the explicit permission of their masters.19Again we see the double message of subjection and of belonging. As freedmen and slaves are unlikely to have acted here without their (former) master’s consent, we may feel justified in taking these monuments as the media of joint selfrepresentation of master and dependent alike. So, the Termessian elites may not have needed to use the cemetery to convey knowledge about their political and civic 16

Of the inscriptions in TAM that mention priests, only 684, 685, 695 and 787 actually

commemorate a priest directly, in other cases we see that relatives or dependents identify themselves by a reference to a more famous relative or patron: TAM 3.1, 497, 539, 647, 648, 671 and 772. 17

TAM 3.1, p. 351, Index 12.

18

E.g. TAM 3.1, 540 set up by a freedwoman of the priests Aurelius Platonianos Otanes

(stemma F14) and Aurelius Meidianos Platonianos (stemma E11) and Platon (stemma E12). 19

TAM 3.1, 269, 338, 346, 495, 636, 637, 663, 762, 764, 769, 811. The expression used is


φσει το δεσπ του.

8 careers, but they did use it as a medium to represent the social networks on which their status and influence in society was based. Another aspect of the way that funerary commemoration could be connected with status is the use of funerary fines. Apparently many people lived with the fear that their tombs, their eternal resting–place, would be disturbed after their death. Obviously it would be the duty of a family to protect and care for the graves of the deceased members, but everyone knew that families did not last for ever. One way was to invoke the supernatural: hundreds of tombs throughout Roman Asia Minor were protected by means of a curse.20 But another perhaps more secure way was to enlist the help of people from beyond the immediate family by mobilising social groups such as professional associations or other civic bodies, or even the whole community in the maintenance and protection of the mnema.21 Funerary fines were one way to enforce this. As in many other cities, the Termessian tombs often included a clause stipulating that in case of tymborychia a fine should be paid. Temples and semi-public institutions appear as the recipients of these fines, with the temple of Zeus Solymeus being the most frequently named, followed by the imperial fiscus and the demos. Other recipients include other deities, the boule, the gerousia, and a neighbourhood association.22 It stands to reason that such fines could only be set with the approval of the intended recipients, as they would acquire an obligation to act if necessary. These inscriptions are therefore testimony to the power of certain individuals to ensure the post-mortem continuity of their commemoration by mobilising larger groups and public bodies as guardians. We cannot locate all the individuals who took the step to secure their commemoration in this way with precision, but it is clear that a high proportion belonged to elite families. This correlation is even stronger if we take into account the price of the fines. The listed cases in TAM suggest that the families at the very top of Termessian society, often buried in expensive heroa, could more easily mobilise a

20

Conveniently collected by Strubbe (1997), cf. TAM 3.1, Index 14.4 , sv. e.g. σβεια,

νεχω, νοχος, or πεθυνος. 21

For a study of reflexive funerary foundations, that achieved a similar goal by different means,

see: Andreau (1977) and van Nijf (1997) 55-68. 22

TAM 3.1. 354-355 index 14.4, sv. Multae violati sepulcrae.

9 wide range of socially respectable groups; consequently they appear to have felt justified in frequently setting a relatively steep fine. As so often, social status was given a monetary expression.23 The practice of securing external care for the grave may have stemmed from a concern for the dead, but a surely not unintentional side-effect would be to include larger segments of the population in the process of commemoration. Getting as many members of the community as possible involved, be they clients, dependent professional associations, or public bodies, could only serve to raise the status of the living. Elaborate commemoration rituals and strategies for the protection of the tomb both turned private matter into a public event. Funerals were a continuation of politics by other means: in this sense the graveyard was an extension of the city centre, and a vehicle for the self-representation of a few families. In the next section I shall discuss some more ways in which these families represented themselves.

2. Genealogical bookkeeping Termessians were rarely buried alone. The epigraphic record shows that most tombs were collective: they were set up by the builder for himself and members of his (or her) immediate family. Even the simpler tombs tended to be designed for parents and children. The tombs of the Termessian elite fitted this pattern, as even the wealthiest tombs were normally designed for one or two generations.24 The Termessian funerary habits were not solely aimed at securing commemoration at an individual level, but they also emphasised family connections and descent. This was further underlined by an extraordinary aspect of Termessian epigraphic practice, an obsession with what I like to call 'genealogical bookkeeping'. Inscriptions - epitaphs, and honorific texts - take utmost care precisely to locate each individual within family lines and family trees. This usually takes the form of supplying each person with a long list of direct ancestors.25 What was the function of

23

TAM 3.1, 355 index 14.4, s.v. Summae denariorum sive drachmarum hae solvendae sunt.

24

The funerary monuments of Termessos are discussed in Heberdey and Wilberg (1900), and

Cormack (2004), 306-323. 25

The names are listed in TAM 3.1.313-339 Index 1.

10 this enumeration of ancestors in Termessian society? Why was it important that this type genealogical information was locally known? If we want to appreciate this, we need to place Termessian practice against the background of what we know about genealogical habits in the wider Greek world.26

Genealogical thinking Genealogical thinking was popular in ancient Greece from Homer and Hesiod onwards. Aristocratic families had been interested in linking up to mythical ancestors, but it is striking that Greek aristocratic genealogies were not overtly concerned with listing their immediate ancestors. Although -or more likely because- aristocratic status was boosted by genealogical means, the opportunities for the public display of aristocratic ancestry were limited in the classical polis.27 Greek aristocrats did not engage in the kind of staged funerals we find in Rome, where ancestor portraits and funerary speeches worked at every level to make family history common knowledge. Greek aristocrats may have wanted to do the same, but many isonomic Greek cities had sumptuary laws in place to prevent this happening.28 Family memories were kept of course, but direct memories may usually not have gone back for more than a few generations; the names of the more remote ancestors must have been transmitted orally, if at all. Although Athenian citizens were supposed to look after the tombs of their ancestors, Athenian tomb memorials rarely included several generations. In so far as Greek aristocrats were allowed to display their genealogical interest in public (e.g. in the context of epinikia) most of the attention seems to have been focused on mythical ancestors of the heroic age.29 It is interesting to note, however, that in late Hellenistic and Roman times the tendency to make longer ancestor lists seems to have increased. This may have started with Hellenistic Kings who legitimised their rule with the monumental display of impressive Ahnenreihen, as on Nemrud Dag, but the phenomenon seems to have 26 27

Discussed more fully by Jones, this volume. An exception seems to be the genealogical inscription for Herophytos of Chios: SGDI 5656.

Cf. Thomas (1989) 156-9. 28

Garland (1989), and Frisone (2000) for a collection of the main texts.

29

Thomas (1989) 155-173.

11 appealed to the urban elites as well. There is evidence that in various parts of the Hellenistic and Roman world the listing of ancestors became more popular, exactly at a time that increasing oligarchisation made dynastic thinking politically correct again.30 It is striking that the new genealogists of the late Hellenistic polis did not just spend time and effort on reconstructing mythical kin, but also on the intermediate generations who were more fully mapped than before. In various cities we find long inscriptions that refer to distant generations, and the reconstruction of long -civicgenealogies has become much easier.31 Proven lineage to historical figures may have been particularly highly prised - certainly by the Romans as we know of a number of high-ranking Athenians who raised their own contemporary status in this way. Herodes Atticus was famously able to trace his family back to Miltiades and Cimon, but many other Greek aristocrats of his age did the same.32 But even for those with a less illustrious ancestry it paid off to be able to cite a number of recent family members who had done a service to the city. The tendency to list ancestors was therefore generally on the increase, but it seems to have been especially common in southwest Asia Minor. Perhaps the most famous – and most extreme- example of genealogical bookkeeping is to be found in the small Lycian city of Oinoanda, or Little Termessos as it was also known. Here we find one of the longest and most peculiar inscriptions of the ancient world: the genealogical inscription that a certain Licinnia Flavilla had engraved on her tomb. This is an inscription of hundreds of lines listing more than thirty generations of ancestors.33

30

Thomas (1989) 159. Cf. SGDI 4859 for Clearchus of Cyrene; Robert REG 73 (1960) 184

mentions a late Hellenistic epigram by Agathon of Dodona who claimed that his family had been proxenoi for 30 generations since the Trojan War. The veracity of these claims is not the point of course. It is interesting that the claims were made at all. 31

See the examples from Sparta and Epidaurus in Spawforth (1985).

32

Herodes Atticus: Philostr. 546f.; Plutarch refers to a Brasidas who, under Augustus, was able

to adduce documentary evidence of his descent from the homonymous Spartan general of the fifth century (Reg. et imp. apophth. 207f).. 33

IGR 3.500 + SEG 46.1709 with the analysis by Hall et al. (1996).

12 This monument has been explained away as the innocent pastime of an elderly lady, but it would be naïve to deny its potent political symbolism.34 The tomb carries a subtle message, aligning ancestors in various different configurations. Through her genealogy Licinnia manages to establish herself firmly within a wide and complex family network, extending from the mythical past as far away as Rome; this was the basis of her high standing in the community. The genealogical bookkeepers in Termessos were in step, therefore, with contemporary practices. Their genealogy was less spectacular than that of Licinnia Flavilla, but their status and identity also depended on the ability to cite an extensive list of ancestors. By using genealogical markers the individuals concerned would strongly identify themselves as belonging to a particular family. The emphasis on genealogy is understandable in view of the demographic regime: not only was the life expectancy at birth extremely low - compared by our standards - but death could also strike suddenly at all ages, showing how fragile the fate of individual families could be. A recent study by Rens Tacoma argues in fact that most elite families in Roman Egypt did not enjoy bouleutic status for much more than three generations.35 The local top elite - the inner oligarchy - was formed by exactly those very few families who were able to survive with their possessions intact for several generations. The ideal of eugeneia turned this stroke of demographic luck in to pure ideology.36 It is not surprising then that Termessian elite families put so much stress on descent and family tradition: it was their symbolic capital, so to speak, as good as money in the bank. The fascination that these genealogical bookkeepers had with ancestry and descent therefore had a clear political dimension. Termessian elite families seem to have used their genealogical bookkeeping strategically to create a suitable ancient background which may have supported their claims to status and power.

34

Hall et al. (1996) 143: ‘The examples of Flavilla and Platonis suggest that Asian ladies of

means, perhaps especially those of mature years, spent considerable energy comparing pedigrees and ranking one another by reference to them’. 35

Tacoma (2006).

36

Zuiderhoek (forthcoming).

13 However these commemorative patterns also acquired a particular local twist. Many of these genealogies included local non-Greek members. Most of the genealogies that were reconstructed by Heberdey were headed by an ancestor whose name clearly identified him as a non-Greek, local Solymian.37 Now, most inscriptions that were used to record these names can be dated to the early centuries C.E., a period that is normally associated with a strong fascination for Greek history and Greek ancestry. Elite families everywhere were displaying their Greek credentials in their bid to status and power. Why then, was there this interest in putting non-Greek ancestry so clearly on display? And what does this tell us about the importance of Greek identity in Roman Termessos?

3. Cultural identity in Roman Termessos The rich epigraphic material from Termessos also allows us to explore the issue of cultural identity. This is of course a central topic nowadays, both inside academe and beyond. There has been an upsurge of interest in identity politics in the Roman empire, and the Greek provinces of the Roman empire are at the very centre of current debates. Recent discussions on identity have focused on the relationship between Greeks and Romans, and on the possible tensions between Greek and Roman identities. The focus has been, of course, on the role of the Greek literature of the socalled Second Sophistic in the expression of cultural identity.38 It has been argued, however, that the picture gets further complicated when we consider material culture as well.39 Local elites everywhere throughout the Greek-speaking provinces of the Roman empire faced the daunting task of ‘becoming Roman - while staying Greek’.

Being Greek in Termessos A first question to ask, therefore, is how Greek the Termessians really were. By the turn of the second and third century C.E., Termessos could boast all the accoutrements of a proper Greek city: with proper public buildings, a theatre, several 37

TAM 3.1.296-311, appendix V. Cf. Heberdey (1929) 58-126.

38

E.g. Goldhill ed. (2001).

39

Woolf (1994).

14 gymnasia and an agora it met all the criteria set by Pausanias, when he judged the wretched polis of Panopeus and found its claim to polis status wanting.40 Yet Termessos’ status as a Greek polis was relatively recent. Southern Asia Minor had maintained links with Greek world from the 6th century onwards, but proper Hellenisation only began in the Hellenistic period. In Termessos this was a gradual process that did not peak until the Roman period. Greek inscriptions appear from the second century B.C.E., and it would seem that at that time the city acquired the political institutions that came along with Greek status.41 When Hellenistic scholars were finally able to identify the Termessians with the Solymoi, a fierce tribe who played a walk-on part in the story of Bellerophon in the Iliad, they could be written into Greek history properly. 42 The local deity was transformed into Solymian Zeus, and other Greek or Hellenised deities gradually began to populate their pantheon. Around 200 C.E. this process appears to have been completed when Termessos acquired, like any old Greek city, its own Greek mythical founder, as an inscription recording the first priest of the eponymous heros Termessos seems to suggest. Termessos may not have been the cultural capital of the Roman Asia Minor, but from now on it counted as a proper Greek city.43 The public sphere, then, was thoroughly Hellenised, but did this imply that Termessians also identified themselves primarily as Greeks? One index for this would seem to be the degree of linguistic Hellenisation. Epigraphic evidence suggests that by the 2nd century B.C.E. Greek had become the standard language to express formal political arrangements, but personal commemoration in Greek still seems to have been limited. The earlier tomb-monuments do not appear to have been inscribed, although it is hard to set up a precise chronology. From the early Roman period Greek epigraphic habits seem to have caught on among the population at large, and Greek 40

Pausanias 10.4. The best survey of the history of Termessos is still that by Heberdey in RE

(zweite Reihe) V.A.2: 732-778. 41

Cf. TAM 3.1.2.

42

RE (zweite Reihe) V.A.2: 737; cf. Str. 13.4.16-17.

43

TAM 3.101. The reconstruction is not certain: το ] πρτου π α!"νος [$(ερως) Τερµησ(?)]σο ,

Cf. Heberdey’s remarks at RE (zweite Reihe) V.A.2: 732, and Heberdey (1929) 33-36. Heberdey wants to identify a male figure wearing a chiton who is represented on a coin in the British Museum as the hero.

15 became the standard language of epigraphic self-expression at a personal level. It has been noted that the Termessians were sufficiently proficient, as even slaves and freedmen by and large seem to have mastered passable Greek.44 We also find indications that there was a certain enthusiasm for Greek high culture: the theatre (O1) and ōdeion (O2) offered places to a fair proportion of the male population. There were two gymnasia (H, I) where young Termessians could pick up the essentials of Greek paideia, and where apart from the usual athletic disciplines, contests in Paean dancing were organised.45 There are even signs of some (sub)literary activity. In their recent corpus Merkelbach and Stauber have included 28 epigrams from Termessos. Nine of these were found on public monuments and may have been worded by professionals. But the remaining 18 were on private monuments, mainly tomb inscriptions, which suggests that their authors were keen to advertise their personal familiarity with, or enthusiasm for, this Greek literary form.46 That some of these texts indeed represented an explicit claim to Greek cultural identity is illustrated, I think, by one funerary epigram, the author of which styles himself –metrically correct- as ‘not the most inconsequential of Greeks.’47 Moreover, there are more texts that show signs of versification, or betray a modest literary ambition,48 but the results were not good enough to be included in SGO. This shows exactly the kind of pitfalls that a poetically inclined would-be Greek could encounter. Epigrams that represented a claim to Greek identity were exposed to

44

Heberdey at RE (zweite Reihe) V.A.2: 737: ‘selbst Sklaven und Freigelassene sprechen

Griechisch im ganzen leidlich korrekt’. 45

I have argued at several occasions that the pursuit of Greek athletic activities represented in

itself a claim to Greek cultural identity: e.g. van Nijf (2001), and van Nijf (2003a). Paian dancing is on record in TAM 3.1, 142, 154, 163. The Termessian agōnes are discussed in Heberdey (1923) and Heberdey (1929) ch. 4. 46

SGO 82-104, nos. 18/01/01-18/01/28. We cannot be sure of course, that the dedicators were

also the authors of the texts. But the quality of the Greek does not suggest that they hired a professional. 47

SGO 18/01/26 = TAM 3.536: Κ,νδιδος, -λλ/νων οχ 0 παρεργ τατος. It should be noted that the

names of husband and wife (Candidus and Severa) suggest that Greekness was not the only aspect of their identity that they were concerned to commemorate. 48

Cf. TAM 3.1, Index 18: Carminum Exordia. Sermo Poeticus.

16 the critical gaze of connoisseurs, grammarians and sophists, who had set themselves up as the arbiters of Greek taste, and who might judge the efforts - and therefore any claim to true Greek identity- wanting. It should perhaps be remembered that the term 'solecism' was first coined to describe the bad Greek of the inhabitants of Soloi, in neighbouring Cilicia.49 This raises the question of how deep the Greek language and Greek paideia were actually rooted in this mountainous city. Other languages were spoken and written in Roman Asia Minor besides Greek and Latin. In Phrygia some 100 inscriptions show that the Phrygian language was going through something of a revival in the first three centuries of our era.50 The language may have continued as a spoken language well into the fifth century C.E. But not all languages have left material traces. The public epigraphy of Iconium is purely in Greek, but when people spoke to their gods local Lycaonian appears to have been the language of choice, as is demonstrated by a well-known passage from the New Testament, where the apostles Paul and Barnabas are mistaken for Zeus and Hermes, and are addressed in Lycaonian.51 We do not know, of course, how widespread the use of this local language was, but we can be fairly sure this episode was only the tip of an iceberg. The continuation of local languages must have influenced local self identification to a considerable degree, and outside the Greek homeland itself it would be unwise to assume that the Roman Empire consisted exclusively of a ‘huge reservoir of monoglot local élites’ as was suggested by Greg Woolf.52 The construction of cultural identity often took place in a complex situation. Greek identity may have been straightforward in the Greek home land, or in the old cities of Asia Minor, but the Greek credentials of the various peoples further East and further inland - were more problematic. The ‘Graeco-Roman Empire’, as Paul Veyne called it, was ethnically and linguistically diverse.53 Greek and Latin were not the only languages spoken in the Roman empire, and the complexity of many local 49

Salmeri (2004).

50

Brixhe (2002), Drew-Bear (2007).

51

Acts 14.8-18.

52

Woolf (1994) 131.

53

Veyne (2005).

17 cultural affiliations and modes of identification seems to increase wherever we zoom in.54 This was emphasised already in the 1960s in two seminal articles by Fergus Millar and Ramsay MacMullen.55 The latter takes a global overview, discussing the evidence for Aramaic, Egyptian (Coptic) and Gallic languages, whereas Millar focused his attention on the situation in Roman North Africa, where epigraphic evidence exists for the use of Punic and Berber alongside Latin. Recently this diversity has been studied mainly from the perspective of diglossia or bilingualism. People would have been used to employing different languages – and language contact was an acknowledged fact.56 This linguistic diversity should not surprise us. Historically bilingualism – or in many cases even multilingualism - was probably the normal situation before the advent of the nation-state .57 What we should like to know, therefore, is what the linguistic situation was in Termessos, and in particular, how strong the local Pisidian traditions still were at the turn of the second and third centuries C.E.?

What’s in a name? Onomastic habits in Termessos Although the epigraphic classes were fully versed in Greek, this does not necessarily imply that Hellenism was their only or even main source of cultural affiliation at a personal level. The use of Greek in public inscriptions was of course politically and culturally correct, but it may have reflected not much more than a desire to present a Hellenising public persona. If we want to find out how Termessians really identified themselves ethnically we may have to dig deeper. One way of assessing how the Termessians identified would be to investigate their onomastic habits. Names are an obvious vehicle for self-identification. Recently Anna Mopurgo-Davis has argued that the intentionality of naming makes a study of personal names particularly revealing of the cultural identity of a community, as it

54

Parca (2001).

55

Millar (1968), MacMullen (1966).

56

Adams et al. eds (2002), Adams (2003).

57

So Janse (2002).

18 ‘tells us something not only about the natural preservation or otherwise of onomastic characters, but also about a set of deliberate choices in name-giving and namepreserving that, in their turn reflect specific attitudes to language but also to community, life, kinship, continuity, etc. in a cultural context characterised by linguistic variety’.58 Modern anthropological and linguistic studies have also found that the study of names is particularly fruitful in multi-cultural contexts.59 Names are a flexible means of self-definition in the sense that they allow for various strategies on the part of the carriers. They will of course often reflect dominant political and cultural power hierarchies, but nomenclature may also serve as a complementary mode of identification. The retention – or re-introduction- of local (or ethnic) names may reflect a desire of a person - or of his or her parents - to mark out a specific identity or identities. If we want to say something about identity, politics and the cultural affiliations of the Termessian population, a study of their onomastic strategies may prove fruitful. Even a preliminary analysis of the onomastic material of Termessos will help us to understand how the inhabitants of this small city projected their particular identities onto themselves and their children. The rich epigraphic material of Termessos provide us with more than 3,000 names, which is a good sample as ancient history goes.60 Even a brief glance at this material suggests that the Termessians chose their names from a diverse pool, including Greek, Roman, and epichoric (ethnic) onomastic traditions.

Greek names We need not be surprised that Greek onomastic habits were widely spread. The great majority of the personal names that we find in the indices of TAM and the supplementary volumes are identifiable as Greek. Ιt is not always easy to distinguish clearly between (e.g.) Greek and Roman names, or between Greek and ‘local’ names.

58

Davies (2000), esp. 24-25.

59

Aceto (2002).

60

Listed in TAM 3.1, 313-339, Index 1.

19 Many ‘local’ names were adapted to Greek flexion, whereas other names may sound unfamiliar, but can in fact be linked to a local or regional preference in the use of Greek. In his important study Les 'oms Indigènes Louis Robert has applied his critical acumen to many names that his predecessors had down as ‘native’ to show that these were in fact (local) Greek.61 Most of the Greek names, however, were ‘run of the mill’ Greek names that were common throughout Asia Minor, such as Apollonios, Diogenes, and Hermaios, whereas other names were less common, or even particular to Termessos.62 Such names may reflect the fact that at a basic level Hellenisation was (socially) widespread, but at the same times this leaves open a gamut of different attitudes to Hellenisation in this community.

Greek ‘designer names’ It is, therefore, more interesting to note that there was somewhat of a fashion for Greek ‘designer names’ that made very explicit references to the classical Greek literary and cultural heritage. Among the more striking examples we find names as Apelles, Atalante, Achilleus, Europe, Homeros, Iason, Kadmos, Kallipateira, Kleon, Perikles, Pangkrateia, Phililogos, Platon, Solon and Sokrates to name only a few striking examples. Standard Greek names may have been used by the Termessians to simply indicate a broad Greek identity, but with such highly classicising names it is more likely that they were positioning themselves with some emphasis within a Greek cultural tradition. This is if anything explicit in a name like Philologos (appearing three times in Termessos), which was often used to convey high cultural aspirations of the bearers, or rather of their parents.63 But other ‘high-brow’ names will have had a similar effect. It is particularly interesting to note that some of these names were found in a much higher concentration in Termessos than elsewhere in the Greek

61

Robert (1963), Index 7, s.v. Termessos en Pisidie.

62

Louis Robert again has warned us not to draw this conclusion too quickly, as the epigraphic

record of Termessos exceeds that of many other neighbouring cities Robert (1963) 205. 63

Robert (1989) 23. n. 10.

20 world. In fact, names such as Kadmos, Perikles, and Platon are found relatively frequently in Termessos, but they do not seem to have been very common among native Greek populations in the Greek home-land.64 It may be suggested, therefore, that these names were selected with the deliberate purpose to flag familiarity with the high-brow traditions of Greek paideia, but also to give a local spin to this choice. Such names would have been particularly important to members of the Termessian elite, who used their mastery and internalisation of Greek paideia as a support for their social dominance.65 It is not surprising then that Greek names appear with some frequency in the stemmata of the Termessian top-families.66 But the social spread of Greek culture does not seem to have been limited to the elite: not all the individuals who show their allegiance to Greek literary high culture can be traced to elite families. This goes to show that Greek cultural identity was indeed relevant to a relatively large section of the population.

Onomastic Romanisation However, ‘being Greek’ was not the only thing that mattered to the Termessians. Termessians were Roman too, and we should like to know how this affected their identity. There are no Latin inscriptions in Termessos, and it is not likely that Latin was commonly spoken, although we may assume that individuals with a Roman (military) career would have had knowledge of the language.67 Leading Termessians often appear to have been Roman citizens, but Roman citizenship was also spread among the rest of the population as well. As in other parts of the empire, civitas romana also implied the adoption of Roman onomastic habits. Termessians with Roman citizenship were of course entitled to sport the tria nomina. This would be more or less expected – if not obligatory - on public monuments.68 It has been

64

We find these names hardly represented in the web-indices of the LGP'.

65

E.g. Schmitz (1997), various articles in Borg ed. (2004).

66

Platon (20) and Perikles (10) are particularly common among the elite familes, whose

stemmata Heberdey has constructed: TAM 3.1, Appendix 5. 67

E.g. TAM 3.1.52.

68

Holtheide (1983).

21 suggested that the use of Roman names on funerary inscriptions was used to stipulate tomb ownership or other arrangements that were protected by Roman law, and this served likewise as a deliberate marker of Roman status.69 If this is so, these tombowners flagged their Roman identity in two ways. The use of Roman names in the Eastern provinces has received some treatment. A recent study by Solin briefly lists a number Latin cognomina that were used in Athens, Central Greece and Lydia and compared these with similar lists drawn up by Kajanto for Rome, but as this study represents an early stage of a research project, no major conclusions can be drawn.70 The onomastic material from Greece has been collected by A. Rizakis and his team at the Centre for Greek and Roman Antiquity at the Greek 'ational Research Foundation, but this rich material is only now beginning to be explored.71 The subject has been broached for western Asia Minor by Holtheide from the perspective of the expansion of Roman citizenship.72 He emphasises that in the first three centuries Roman citizenship was increasingly important for the provin cial, and with a slight delay, for the local elites as well. Before the Constitiutio Antoniniana, the spread among the lower classes would have been much slower, and was limited mainly to soldiers and successful athletes, both of whom can be seen as agents of Romanisation.73 Only towards the end of the second century can we see numbers of citizens rising steeply, until of course the Constitutio Antoniniana established the same legal situation for all the free male inhabitants of the Roman Empire. As most of our inscriptions date from the end of the second and the beginning of third centuries C.E. the issue of legal privilege would be less important. In this chapter I am less interested, however, in the legal implications of Roman names, and more in the way that they were used to convey Roman identity. How did the non-native Latin speaking inhabitants of Termessos use Roman names to 69

Cf. Meyer (1990).

70

Solin (2001).

71

E.g. Rizakis (1996), Rizakis and Zoumbaki (2001), Rizakis et al. (2004).

72

Holtheide (1983).

73

Holtheide (1983) 132. For a discussion of athletes and performers as the agents of a

globalising Roman imperial culture, see van Nijf (2006).

22 flag cultural adoption of Romanitas? Among the Roman names in Termessos, Aurelii abound of course, but a number of Tiberii Claudii, certainly among the elite families, indicates that Roman onomastic habits were adopted as early as the first century C.E.74 However, onomastic Romanisation did not necessarily imply that Roman citizens had to give up their own cultural traditions or even their own name. Greek names (or other local names) were routinely integrated into the tria nomina. Most Termessians with Roman citizenship had in fact a Greek cognomen, but Roman cognomina are not trailing far behind, and we also find cognomina with a more local character. All Roman names may be seen as an aspect of the personal identification of a subject (or of his or her parents) with the Roman empire, but some onomastic practices must have been intended as a more deliberate marker of personal ‘Romanisation’ than others, for instance when the referent was particularly closely identified with the Roman centre. In this category we may place names as Agrippa, Victoria, Italicus, Caetolinus, Corbulo, Quietus, Varus, Seneca and Faustina. In many cases we find names of Roman origin that had been adapted to the Greek onomastic system. 75 This suggests that Roman onomastic practices went beyond the simple adoption of Roman names to comply with Roman legal requirements. Being Roman was for many simply a part of life, but in many cases we may even seem to be dealing with a strong desire to flag Roman identity.

‘Ethnic’ names The final category that we want to investigate is that of the local names. Termessos had a long Pisidian history, and their local language, Solymian which was a Pisidian dialect, had a long and strong tradition. It should not come as a surprise that there is a great number of local – epichoric – names to be found in the Termessian material, even though some names were adjusted to Greek flexion.76 Among them we find Armasta and Armaos (cf. Hermaios), Bekkobais, Gamodis, Kakasbos, Kendeas. Kinnonis, Mamotasis, Masas, Moles (cf. Molianos), Morsis, Motosourgis, Nannelis, 74

Stemmata D, H, and M were headed by Tiberii Claudii. TAM 3.1, Appendix 5.

75

TAM 3.1.339 Index 2.

76

The classic survey of local onomastic traditions in Asia Minor is Zgusta (1964). A brief

survey in Neumann (1992).

23 Oa (cf. daughter of Platon), Oples (cf, gen. Oplounos), Otanis (cf. Otanianos), Piaterabis, and Trokondas. The normal expectation of the process whereby subject nations of the Roman empire accommodated their naming patterns to imperial onomastic habits, is that vernacular speakers would first have adopted Roman names alongside their native names, and then gradually dropped their old names. As John Adams has remarked: ‘it may be assumed that changing names went hand in hand with changing languages.’77 Another, at first sight reasonable assumption would be that class was a determinant of the pace of this onomastic/linguistic change. The higher classes – the order of local councillors – would on this model be the first to be Romanised, and also the first to shed their local names. The other classes would only gradually follow suit, and be the last to adopt ‘foreign’ names.78 The Termessian data suggest, however, that the picture may be more complex than this. Due to the chronological concentration of the epigraphic material it is difficult to get a sense of the development over time, but it is obvious that by the turn of the 2nd and 3rd centuries ethnic names had by no measure died out. So, despite a pretty complete political Romanisation of the community (by 212 all male in habitants would have been Roman citizens) and a gradual - but by all accounts strong – cultural Hellenisation, its local onomastic traditions still retained currency even at this late stage. The class approach does not apply either, since the bearers of these names were not limited to the lower classes. As we saw above, elite families were equally keen to advertise their local roots as well. Moreover, we have no idea at what rate Termessian names actually declined, or whether there had been any fluctuations or fashions over time.

Supernomina This picture is even further complicated by the fact that the different onomastic traditions were mixed to a high degree. One striking manifestation of this 77

Adams (2003) 290.

78

Holtheide (1983).

24 is the frequent use of supernomina.79 The Termessian record shows indeed a wide variety, including a reference to a non-Termessian origin, as in the case of an Alexandreus, or to a past as a slave (Etoimos). Heberdey lists 97 supernomina, of which more than half (50) are clearly Greek, 15 are Roman, and 29 local.80 A small selection suffices to show the varied choice available to the Termessians and the effects that were achieved.

•

Τιβ(ερ2α) Κλ(αυδ2α) Κ2λλη 3 κα5 Καπετωλε6να (268)

•

Μ;ρ(κος) Αρ(/λιος) =ρτµων 0 κα5 Ερ(µα6ος) (60)

•

-ρ(µα6ος) @ρστου 0 κα5 Aλβιος (139)

•

Μ;(ρκος) Α(ρ/λιος) Οπλεσιαν ς 0 κα5 =ρχιγνης (171)

•

Τροκονδας εG Αττεους 0 κα5 =ριστ νεικος (188)

•

Αρ(/λιος) -ρµα6ος Αρ(τειµου) 0 κα5 Τιλλ ροβος (441)

•

Αρ(/λιος) Κορκαινας Αρ(τειµου) Λουκρ2ωνος 0 κα5 Γρε6πος (561 - 662)

•

ΠNιγNερλωνιςN 0 κα5 Κ,στωρ =πελλ[ο] N (722)

•

Αρηλ2α ΝεικηφοριανPN ΜNορσανδα 3 κα5 ΠNλατων5ς (623)

•

Οα Μο(λεους) βG το κα5 Γα2ου (670)

Supernomina are usually explained as the result of the incompatibility of Roman and Greek naming systems, allowing Romanised individuals to keep using non-Roman names, that were added to the Tria Nomina by the use of expressions such as qui et, or in Greek variations of 0 κα5. Alternative expression include
πιλεγ µενος (called besides), whereas προχρηµατ2ζων may indicate a previous name. 79

TAM 3.1.341, Index 3.

80

TAM 3.1.341 Index 3.

25 But the supernomen could also be employed accommodate other additions to the traditional naming pattern such as ethnics, nicknames or signa.81 It is fair to say that interest in these naming patterns has been dominated by a cognitive and philological approach to this linguistic phenomenom: what is the meaning of a particular name? And what does its occurrence tell us about linguistic developments?82 But if we shift our perspective to ‘a context in which meaning is socially constructed by the use of language(s) within a specific culture, multiple names for individuals make a difference.’83 Michael Aceto, whom I quote here, has studied the phenomenon of multiple names within the context of the English-language creoles in Hispanophone areas of the Caribbean. As in Termessos, these creoles employ multiple names for individuals. In addition to their official Spanish name often sports an ‘ethnic name’ that is used locally for reference and address, and that ‘defines who members of this community are in terms of culture and ancestry’. The ethnic name represents, therefore, an active cultural choice - not a necessity, and certainly not a left-over from previous fashions. Leaving aside obvious differences between these societies, the concept of ‘ethnic name’ may be useful in Termessos too. By using these ethnic names for themselves and for their children alongside Greek and Roman names, the Termessians behaved like onomastic code-switchers, who could use different names for the same person in response to different social and cultural contexts. Individuals had the choice to emphasise distinct elements of their identity in different circumstances. Someone who may be known simply as Trokondas in one situation, was known elsewhere as Aurelius Klaros Trokondas, 84 and an Aurelia Artemis was also known as Mauenna, the daughter of Dioteimos Maximus.85 This element of choice is brought to the fore even more if we consider the effect of these naming patterns on the representation of families. There is a 81

Kajanto (1967).

82

E.g. Lassère (1988), but see Colvin (2004) who presents a fascinating study of the Lycian

material. 83

Aceto (2002).

84

TAM 3.1.903.

85

TAM 3.1.309.

26 surprisingly high degree of intra-family variation. Local names and Greek names could alternate: even families that presented themselves as strong Philhellenes resorted from time to time to local names. A Plato could name his daughter Oa,86 a Socrates his son Oplon, but his grandson was called Platon;87 Corbulo could be the son of a Tryphon, he was married to Pytheas.88A Homeros called his son Hermaos, which is a name with Luwian roots , and the family stuck with this name for some generations.89 Sapron’s son was called Aurelius Zotikos and his wife Kalèmera alias Primigenia; they called their daughter Aurelia Agoraste. Sapron’s first wife, also in the same tomb, was called Anna but was also known as Orestiane, and there was also a foster child called Doris – the Dorian girl. 90 A man called Hermaios, the son of Trokondas, alias Kousion, was the freedman of a Thoantianos, the son of Hermaios and grandson of Arteimos. His wife was called Nannelis, but she was also known as Aspasia. They shared a tomb with a woman Gailla, who was the daughter of Marcus Aurelios Euporos.91 Such examples could easily be multiplied. It is quite obvious that Termessian families had a wide choice, which they seem to have exploited to the full. Naming was clearly not a passive reflection of a pre-existing linguistic and cultural identity, but rather an active factor in the construction - and representation - of personal and family identity in Roman Termessos. Termessians were able to dip into a diverse onomastic pool allowing them to emphasise their familiarity with Greek paideia, their loyalty to Rome, or their strong local roots, just as the situation demanded. ‘Ethnic names’ were deliberately used in addition to Greek and Roman names, to gloss any preconceived ideas about their cultural identity by referring to an alternative cultural affiliation, but it is not possible to state that these names were actually preferred by their users. Whatever their linguistic identity, onomastically the Termessians behave like code-switchers. Individuals and families were apparently able to accommodate 86

TAM 3.1,778.

87

TAM 3.1.184.

88

TAM 3.1.557.

89

TAM 3.1.446.

90

TAM 3.1.509.

91

TAM 3.1.467.

27 different cultural traditions easily in their onomastic system. Onomastic patterns in Termessos had their own dynamics, which shaped as much as they followed the cultural transformation in this mountain site

Becoming Roman – staying Termessian Cultural identity in Roman Termessos appears to have been a complex issue. The onomastic evidence for Termessos suggests that nobody needed to be pinned down to one single cultural tradition. This clearly shows, I think, that in Termessos identities were multiple and in a constant state of flux. Every Termessian could present himself as a composite of cultural affiliations and attachments, and although there may have been a certain hierarchy among the elements that made up individual identities, that hierarchy was not immutable, and it could change with time and context. Their cultural identity was manifold and could not really be compartmentalised. My final question is how we should rate the significance of the cultural phenomenon that we have just observed. Termessians were Greek, as well as Roman, and Pisidian to boot, and this defies any easy attempt at categorising them. It is tempting to explain away this blend of attributes as a local peculiarity of a remote and backward mountain town. It is important, therefore to emphasise that multiple identities, and indeed a strong fascination with their own past may not have been unique to the Termessians. Throughout the Roman empire we find in the course of the first three centuries C.E. evidence for an increasing interest in the local past. Although anti-Roman feelings may occasionally have played a part, it is striking that throughout the Mediterranean ‘global’ and local themes were more often mixed to produce a new blend of a provincial Roman society.92

92

The most dramatic example of tension between Roman and local identities is of course the

Jewish revolt. It should no be forgotten, however, that in the diaspora Jews were often more or less fully integrated in local society. On the relationship between Jews and Romans, one should now consult the masterly analysis in Goodman (2007). In Termessos one Jewish woman is known, she was called Artemeis, which was a name that referred both to the Greek and Pisidian cultural spheres. Her father was called M. Aurelios Keês, son of Hermaos II. Again a mixture of Greek, Roman, and local traditions (TAM 3.1.448).

28 Cemeteries were among the prime locations where such negotiations between the local and the imperial were acted out. David Mattingly has recently shown in a study on family tombs in North African Ghirza how Roman iconography and style were appropriated to serve an indigenous agenda.93 Egyptian mummy portraits represent members of the local elites as Roman citizens, or Greek athletes, while adhering to a marked Egyptian style. And tomb types and other funerary practices in Roman Lycia maintained the funerary styles of the Lycians of the 4th century B.C.E.94 It is an irony of ancient globalisation under the aegis of Rome that an obvious orientation towards the imperial centre also appears to have fostered a growing interest in the local.95 It may be suggested that the most striking example of this empire-wide cultural trend was the Second Sophistic, which seems to have turned Greeks - and would-be Greeks - into nostalgic classicists. It has been common to explain this fashion solely as the result of the interaction of strong Greek cultural tradition with the then realities of Roman domination, but perhaps this was but the most vociferous example of a trend that we can observe throughout the empire of provincials returning to really ‘local’ cultural roots in order to create for themselves a place in the imperial present. At any rate, the situation in Termessos may have been less exotic than it seemed: on the contrary Termessians were sharing in what we might call an empire-wide ‘age of nostalgia.’

Conclusion I have looked at the cemeteries of Termessos as a source of local knowledge and I have tried to explore some of the ways that identity politics were played out in 93

Mattingly (2003).

94

My observations here are based more on a personal impression of the material, and a cursory

look at some studies, than on systematic research. To address this issue in any depth would require the context of an interdisciplinary research project. Until then, see for a discussion of the Lycian material Hülden (2006), esp. 216-217, and for the Egyptian material, Riggs (2002). 95

For a different and inspirational discussion of Roman ways of reclaiming the past for the

present, see the work of Susan Alcock, and esp. Alcock (2002). Globalisation tends to be used in the economic sphere, but it can also be applied to the cultural field. This has been convincingly argued by Chris Bayly in his study of what he calls ‘archaic globalisation’, Bayly (2002). I have adopted this concept (as ‘ancient globalisation’) in van Nijf (2006).

29 this remote city, high up in the Pisidian mountains. I have identified a few key-issues around which the Termessians constructed their identities. Status or distinction was a common theme, and I have argued that the funerary display helped to underwrite the social hierarchy of the community, but I have also argued that the emphasis was different than in the city centre. Wealth and social networks were much more in evidence than the individual’s place within formal political hierarchies. Termessian identity was closely bound up with family identity, certainly for the members of elite families whose status depended i.a. on their success in conveying an image of themselves as standing in a long family tradition. A peculiar feature of the Termessian epigraphy, an obsession for what I have called genealogical bookkeeping, visible both on the epitaphs and on honorific inscriptions of the city centre, can also be explained against the background of political oligarchisation and social hierarchisation which was a feature of civic life throughout the Roman empire. Finally, I have argued that the Termessians were able to identify themselves in various and complex ways as belonging to different cultural traditions. We cannot ask them, but I suspect that they would not have been able to answer unequivocally what their overarching identity was: Roman, Greek or Anatolian? But they might not have cared: as they were Termessians, they were all of these things at the same time, and many other things besides.96

96

It is relevant to quote in this context Amin Malouf, a French writer of Christian Lebanese

extraction: «Depuis que j'ai quitté le Liban pour m'installer en France, que de fois m'a-t-on demandé, avec les meilleures intentions du monde, si je me sentais ‘plutôt français’ ou ‘plutôt libanais’. Je réponds invariablement : ‘L'un et l'autre !’ Non par quelque souci d'équilibre ou d'équité, mais parce qu'en répondant différemment, je mentirais. Ce qui fait que je suis moi-même et pas un autre, c'est que je suis ainsi à la lisière de deux pays, de deux ou trois langues, de plusieurs traditions culturelles. C'est cela mon identité...». Malouf (1998).