I INTRODUCTION
‘Romanization’ is a debated concept.Footnote 1 In the past few decades, the whole idea of what ‘becoming Roman’ meant to indigenous populations, and the efforts local élites (rather than Roman administrators) put into the process, has been reformulated from a ‘post-colonial’ perspective in different disciplines.Footnote 2 It seems impossible nowadays to conceive of Romanization as a centrally organized process towards political, social, legal or religious unification. The spread of Roman culture, of institutions, artefacts, language, forms of medial representation and more, nevertheless calls for explanation, and ‘Romanization’ remains the most plausible term to encompass the questions that arise. For the Western provinces, its use seems to be well established. But for Asia Minor, the absence of Latin and the continuity of Greek social structures have been adduced to counter the claim that the region underwent a process of ‘Romanization’ comparable to the Western provinces; according to this view, the influence of Roman culture in the cities of Asia can be reduced to matters of taste and style.Footnote 3 As it should be admitted that Romanization is a meaningful concept only if it covers cultural transfers that go beyond the partial adaptation of architectural models or clothing styles, this line of argument raises legitimate questions about the use of the term with respect to the Eastern provinces.
Cultural transfers of a more influential sort can be argued to be found in the organizational structure of a society, or more precisely, the forms of social aggregation that are chosen by its inhabitants. Among these, private associations are a promising test case. From the beginning of the epigraphic record for regions such as Gaul or Spain, collegia are attested, and there is little doubt that they were regarded by all actors involved as an element of the Roman civic order. The Lex Irnitana anticipates that in a Roman municipium, people might want to gather in collegia, and specifies the conditions applicable.Footnote 4 In the imperial era, specialized professional collegia were regarded as an age-old Roman institution — Plutarch and other authors attribute their introduction at Rome to Numa.Footnote 5 Verboven, in a study of collegia in the Western provinces, has aptly stated that in the second century, ‘les collèges tenaient une place centrale au sein du modèle de la cité authentiquement romaine’.Footnote 6 And since associations tended to replicate the Roman civic order on a smaller scale, every single one of these privately-founded groups could act as an agent of ‘Romanization’.Footnote 7
However, what seems evident for the Western provinces — that every form of collegiate organization is a result of Roman influence and follows Roman models — seems a priori false for the Greek East. After all, private associations are widely attested in Greece from the fourth century b.c.e. onwards, Egyptian associations seem to have existed even earlier, and the body of evidence from Hellenistic Asia Minor is constantly growing.Footnote 8 When Gaius discusses the regulations on sodalitates in the Twelve Tables, he cites a Solonian precedent.Footnote 9 While everyone would still agree that the epigraphic record of associations in the Roman era is different from earlier periods, the claim that the Roman provincial administration triggered the evolution of private associations in cities like Ephesos or Smyrna has little to recommend itself. In what may be the most thoughtful treatment of the problem, Cracco Ruggini has in fact reached the opposite conclusion: given the Roman anxiety about the dangers of collegiate organization, and the long-standing tradition of Greek and Anatolian private associations, apparent similarities of Greek and Roman collegia and synodoi in the imperial era cannot be due to Roman influence. Rather, the successful co-operation of civic bodies and associations in Greek cities, based on their ‘compenetrazione di interessi’, served as a model for the Roman state. The latter slowly came to accept the advantages of this sort of associational organization, and adapted it in Italy and the Western provinces; however, due to the different economic and mental framework provided by the Roman social order, this transplantation resulted in the creation of ever less private, more state-controlled organizations, until collegia became hereditary tax units.Footnote 10 The direction of cultural transfer would thus be from East to West, from Asia Minor to Spain, with Rome as a mediator.
This view, if correct, would provide further arguments in favour of a sceptical approach towards ‘Romanization’ in Asia Minor. It cannot, however, stand unchallenged. Some of Cracco Ruggini's arguments are based on traditional assumptions no longer held by the majority of scholars, e.g. the view that Greek cities in Asia Minor preserved a pre-Hellenic, Anatolian social structure with civic subdivisions based on professions.Footnote 11 In addition, recent research has done much to undermine the sharp distinction between state-controlled Western collegia and the ‘contractual’ associations of the East. Like their Eastern counterparts, professional collegia in Gaul or Spain actively sought integration into the public sphere, and employed terminology that served to bolster their claim to represent more than just their individual members’ interests in business opportunities and sociability.Footnote 12 Finally, more attention should be given to both the chronology of the epigraphic evidence and to regional differences. The aim of this paper is to study the connection between associations and Romanization in one particular region, Roman Phrygia, in order to address the relevant issues in more detail than can be done in a general treatment of ‘Asia Minor’.
The choice of Phrygia may at first sight seem unfortunate. While all regional classifications are constructs to some degree, ‘Phrygia’ is a case where this becomes especially relevant. A glance at the geographical situation is sufficient to suggest that the cities of southern Phrygia, like Hierapolis, Laodikeia, or Apameia, had much more in common with cities like Miletus, Magnesia and Priene than they had with the villages and small cities in the highlands of northern and eastern Phrygia.Footnote 13 Observations on the impact of Greek culture in the Hellenistic period (visible early on in the Lycus valley, almost absent in central and northern Phrygia) bolster this impression.Footnote 14 But these differences within the region, and especially the differences with regard to the degree of Hellenization before the Roman conquest, make Phrygia especially interesting for an analysis of the connection between Romanization and associations — after all, the continuity of Greek culture is an important part of the debate. And although practically all epigraphic evidence comes from the imperial era, the possibility of tracing traditions that are neither Greek nor Roman in origin is more realistic in Phrygia than almost anywhere else, given the Phrygian origin of terms such as bennos or doumos. This means in turn that the findings presented here are often specific to this particular region. But as comparative data will be adduced, they can also serve as the basis for some more general insights.
II THE SPEARHEADS: RESIDENT ALIENS AND OLD MEN
When looking for privately organized multipliers of Roman values, the obvious starting point is the groups of resident Romans, known in Asia Minor as ‘the Romans doing business’ (hoi pragmateuomenoi Rhōmaioi), ‘the Romans living here’ (hoi katoikountes Rhōmaioi) or the like.Footnote 15 The influx of Italians reached Phrygia with a considerable delay compared to the coastal regions of Asia Minor.Footnote 16 But the impact was profound. In first-century c.e. Apameia, the katoikountes Rhōmaioi were a very influential group that regularly appeared alongside the dēmos and boulē in civic decrees. In this city, resident Romans seem to have had privileged access to the most important magistracies; an inscription from 45/46 c.e. stresses the fact that all five archons were now Roman citizens; four of them came from Italy.Footnote 17 Due to its geographical position, Apameia was the most important transfer site for long-distance trade in Phrygia. This naturally attracted businessmen from abroad; they were responsible for the wealth of a city that otherwise had very little to offer, at least according to the literary sources.Footnote 18 Grouped together in one organization, these Romans could easily play an important rôle in civic politics. By doing so, they perhaps responded at least in part to the needs of the local population, who offered political posts as investment incentives.
The case of Apameia may be special in some respects, but resident Romans are found all over western Phrygia, and we may assume that their political influence in the respective cities depended on similar considerations. The formal traits of this arrangement certainly varied from city to city. In Hierapolis, Romans are not normally mentioned as a decision-making body in civic decrees. But an inscription from the third century does record a joint decision by the council, the people, the gerousia, the synedrion tōn Rhōmaiōn, the neoi and the synodoi (probably referring to the associations of Dionysiac artists).Footnote 19 They honour a person who has served, inter alia, as konbentarchēs of the Romans, a Latinism unattested outside Hierapolis. While the resident Romans did not have the same institutional standing as in Apameia, their affairs were treated as civic affairs in Hierapolis as well. In another inscription for a konbentarchēs, the honours are conveyed by the boulē and patris; the boulē honours him as its boularchos, while the patris refers to his general benevolence, which apparently includes taking over the leadership of the Romans.Footnote 20 This is all the more interesting because the term synedrion could easily be taken to refer to a private association, and the general situation — foreign merchants joining forces — was a constitutive element for many associations in the Hellenistic period. The decrees clearly show that the resident Romans regularly gained political influence beyond what could be expected of a mere private network. But they show us the result rather than the origin of a development that presumably had its roots in the spread of private networks — networks which were explicitly labelled ‘Roman’ by all parties involved.
A similar argument can perhaps be made for a rather different form of corporate organization that had its roots in the late Hellenistic period, but spread rapidly and with a new political focus only in the imperial era, namely the assemblies of old men (gerousiai). In many cities of Asia Minor, the gerousia is mentioned alongside the dēmos and boulē (and at times the resident Romans) in civic decrees. It was tied to the gymnasium, and although its character as an élite club sharply distinguishes it from the traditional Greek conception of age classes,Footnote 21 the gerontes were still treated as one in official parlance (e.g. the gymnasiarch pasēs hēlikias, ‘of all age classes’). The gerousia also appeared, albeit not too frequently, as a personified civic institution both in the form of statues and on coins, as did the dēmos and boulē.Footnote 22 Phrygia could be said to lead the way here: the earliest epigraphic evidence for a statue group comes from a Phrygian city, as does the earliest coin showing gerousia personified.Footnote 23
But besides these seemingly unambiguous signs of the public nature of the gerousia (which would still let it appear as a Romanized institution, but not one that would be of interest in our context), other evidence justifies the conclusion that it was in fact an institution at the border between public and private. Unlike civic boulai, the foundations of gerousiai could be the result of private initiative, as in the case of Lycian Patara, where C. Iulius Demosthenes of Oinoanda, one of the most prominent Lycian benefactors, established the gerousia as a part of his own regional network.Footnote 24 Also in Lycia, the city of Sidyma decided in the time of Commodus to establish a systēma gerontikon, and to choose a prominent citizen to present this idea to the proconsul (who replied that it deserved praise, not permission).Footnote 25 The same situation could perhaps be reconstructed for Phrygian Apameia.Footnote 26 This seems to imply that new gerousiai could be officially recognized as legal associations, presumably on the basis of the Lex Iulia de collegiis.Footnote 27 As regards conditions of membership, entering a gerousia could involve similar procedures to entering an association, such as payment of an entrance fee and preliminary screening of candidates.Footnote 28 An inscription from Akmoneia (64 c.e.) is remarkable precisely because it records an unusual process: a certain Demades is entrusted ‘with the introduction of a name without a fee’; after he has chosen the freedman Karpos, the other members vote that Karpos should ‘partake in the gerousia on completely equal terms’.Footnote 29 Apparently, Demades is rewarded for some benefaction with the right to inscribe one of his friends (or dependants) into the gerousia; the closest parallels in procedure come from the regulations of associations from Delos and Athens.Footnote 30 In Phrygian Sebaste, an inscription records the admission of seventy-one persons into the (newly-founded?) gerousia in 99 c.e. — among them is a whole family of Iulii, including women and children.Footnote 31 Women were not normally represented in politics or the gymnasium,Footnote 32 but they could apparently find their place in the exclusive ‘clubs of elders’ that proliferated under Roman rule.
All this hardly fits a categorization of gerousiai as civic bodies equivalent to the boulai. But in conjunction with the political importance accorded to the gerousiai by many cities, it does fit the model developed above for the ‘resident Romans’ rather well: a closed circle of Romanized (or simply Roman) people within a given city assumed a corporate identity, thus accumulating influence and network effects. As such groups could hardly be bypassed by civic government, and could in fact use their accumulated prestige to the good of the city, they were incorporated into the decision-making process. Under Roman influence, new forms of corporate organization were created that quickly found their place in the institutional order of Phrygian cities. Romans and ‘old men’ were the most successful groups, but there were certainly others, especially in the early phase of Roman rule, some of them of a rather curious nature. Thus, in 6/7 c.e., the ‘Greek and Roman women’ of Akmoneia decided to honour the high priestess Tatia.Footnote 33 We should not press Thonemann's assumption that the Romanized élite of Akmoneia consciously followed the model of the Augustan ordo matronarum in giving its women the authority to issue their own decrees. But it is clear enough that the presence of a new Italian élite in the more important Phrygian cities (and the influence of Roman institutions such as merchant co-ordination or the senate itself) had a profound impact on the development of new patterns of social organization. Some of the results remained exceptional and did not persist, such as the association of Greek and Roman women. Others, like the associations of resident Romans and the exclusive clubs of elders, profoundly shaped the political and cultural landscape of Phrygian cities. These cases set the stage for our enquiry. They show that what we are looking for was a very real phenomenon in Roman Phrygia.
III CLAIMING A PLACE: PROFESSIONAL ASSOCIATIONS AND ISOMORPHISM
Professional associations have a key rôle in the studies by Verboven and Cracco Ruggini noted above, and they certainly deserve it. The phenomenon as such is not entirely Roman. In Egypt, there is earlier evidence for professional associations, and a new inscription from (probably) Lycian Tlos that records a funerary foundation involving a koinon tōn chalkeōn proves that they were already known in Asia Minor in the second century b.c.e.Footnote 34 Still, in the Hellenistic period, associations that chose a common profession — and not a common cult — as their main identifying statement were extremely rare.Footnote 35 The exception from Tlos may actually be explained through Egyptian influence, as Lycia was under Ptolemaic control in the third century b.c.e.Footnote 36 In the imperial era, professional associations are known from almost all areas; they are often the only form of the associative phenomenon attested in a given city.Footnote 37 The change seems severe and needs to be explained.
In Phrygia, Hierapolis stands out with roughly twenty professional associations attested in published or unpublished inscriptions.Footnote 38 This unusual situation is certainly explained by both the epigraphic habit and the state of excavation, as comparison with the two other textile centres of the Lycus valley shows. In Laodikeia, a city that was no less important for textile production and whose social milieux show clear signs of Romanization,Footnote 39 the evidence is meagre indeed. Two professional associations are known from one insecurely reconstructed inscription.Footnote 40 One or two other synergasiai with unknown specification are attested in fragmentary seat inscriptions from the theatre, and a new inscription reads topos bapheōn, ‘place of the dyers’.Footnote 41 But from Hierapolis, more than 600 inscriptions are known, while the number for Laodikeia is clearly below 100; in addition, most Hierapolitan associations are known from funerary endowments recorded on sarcophagi from the northern necropolis, while in Laodikeia, this genre of texts is absent. For Kolossai, not a single association is securely attested.Footnote 42
Honorific decrees give us an insight into the activities and the status of professional associations within Phrygian cities. It is noteworthy that not a single inscription is concerned simply with honouring a benefactor of the association; instead, all honoured persons represent the Roman order in one way or another. A good example is Lucius Egnatius Quartus, a Roman military commander who was honoured by the synergasia of fullers in Akmoneia as founder of the city and a true patriot; his benefactions towards the association are mentioned last and not specified any further.Footnote 43 The same person was honoured by the official civic institutions of Akmoneia.Footnote 44 The fullers also honoured Titus Flavius Montanus, a friend of the procurator Vibius Lentulus, as praefectus fabrorum, archiereus Asias, sebastophantēs and agōnothetēs.Footnote 45 That he had also done something for the association seems to be little more than a pretext. Perhaps the benefactor did not wish this rather irrelevant dimension of his public life to divert attention from his main achievements. The result, in any case, was that the association inscribed itself into the Roman order by establishing (and publicly demonstrating) contact with one of its more prominent representatives. In Hierapolis, both the purple-dyers and the wool-washers erected almost identical inscriptions in honour of Claudius Zotikos Boas, first stratēgos, archiereus etc.; the associations are not said to have received any benefactions themselves.Footnote 46 Professional associations could, of course, profit from building projects carried out in the city. In Laodikeia, the fullers and the wool workers were involved in an honorific decree for someone who erected a market hall.Footnote 47 But what seems to have mattered most to professional associations was their self-presentation as quasi-official institutions.Footnote 48 In Apameia, they even served as civic agents. A number of honorific decrees from the first and second century, all issued by the dēmos, the boulē and the katoikountes Rhōmaioi for prominent Romans, were to be set up by professional associations, namely, the shopkeepers, businessmen and artisans of two commercial streets.Footnote 49
Professional associations thus took part in the representation of their respective cities vis-à-vis Roman officials and benefactors. They profited from networks established by official institutions, because they acted in ways that deliberately obscured the boundaries between public and private. A further indication is the appropriation of public space. The seat inscriptions of synergasiai from the theatre in Laodikeia have already been mentioned. Another example comes from the small theatre of Aizanoi, where seats were reserved for the phylai, but also for the association of stone-cutters.Footnote 50 That professional associations could be put on display in this way as part of the institutional inventory of a city is known from other locales in both the Eastern and Western provinces.Footnote 51 For Aizanoi, it has been suggested that the seats were periodically sold by the city;Footnote 52 this would mean that the association of stone-cutters had to actively invest in its integration into the civic order. Additional evidence comes, again, from Hierapolis. The theatre was the place where the association of dyers set up a statue of Boulē personified.Footnote 53 And the most prominent association of Hierapolis, the purple-dyers, added a postscript to a dedicatory inscription on the architrave that recorded their contribution to the adornment and enlargement of the theatre.Footnote 54 The funerary inscriptions also point in the same direction. Professional associations received endowments just like the gerousia; they could also be named as the recipients of fines to be paid for the violation of graves, alongside (or instead of) the Roman fiscus.Footnote 55 It is not quite clear how such rules were put into practice, but we see that founders of funerary endowments treated professional associations on a par with official institutions.
We hardly know anything about the internal structure of Phrygian professional associations. No laws have survived, and no officials are mentioned.Footnote 56 The one exception to this rule is revealing for our purposes. Professional associations in Hierapolis could have a management board that represented the association on certain occasions. Some inscriptions from the late second and the third century do not mention the ergasia, but the proedria of purple-dyers. This ‘management board’ can be named as the recipient of funerary endowments just like the association itself. Like the ergasia, the proedria can carry the epithet semnotatē; in one case, we find the designation synedrion tēs proedrias tōn porphyrabaphōn.Footnote 57 Management boards are also attested for the ergasia of linen-weavers, where the progegrammenoi of the association shall receive and distribute the sum of an endowment, and for the association of metal-workers, where the designation in an as yet unpublished inscription is proestōtes.Footnote 58 While individual leading positions are, of course, known from Greek associations from the Hellenistic period onwards, these officials did not form a management board that could act separately from the other members. But we know the phenomenon from Italian inscriptions of the second and third century. Associations could be divided into ordines, following the civic model: there could be an internal ordo decurionum, distinguished from the populus and with the authority to make its own decisions, and magistri were often taken as one class inside the association.Footnote 59 Liebenam's assumption that in the West, such management boards were called praesidium, is incorrect,Footnote 60 although the analogy to the Hierapolitan proedria would be tempting. But the separate designation and activities of a management board in at least some professional associations from Hierapolis is an important sign of Roman influence.Footnote 61 It shows that collegia were indeed among the private organizations which, through their mimicry of accepted norms and institutions, ‘naturalised and legitimated the basic postulates of a Roman social order’.Footnote 62
The religious practice of professional associations could also be influenced by Roman models, but only in an indirect way. The well-known funerary endowment of Publius Aelius Glaukon and Aurelia Amia illustrates the problem well: Glaukon left 200 denarii to the management board of the purple-dyers for them to celebrate ‘the feast of unleavened bread’ from the interest; he also left 150 denarii to the association of carpet-weavers, who were supposed to spend one half on the feast of the Kalends and the other on the ‘feast of the fiftieth day’.Footnote 63 The religious calendar of these two associations was determined by the conditions of the endowment. The donor decided which feasts mattered most — Glaukon gives 200 denarii for Passover, but only 75 for the other two feasts, to be celebrated by a less prestigious association.Footnote 64 Collegia in the West celebrated the New Year's feast,Footnote 65 but this is not clear evidence of Romanization of associations (rather than of the donors themselves). Funerary endowments often refer to the typical dates of collegia festivals in the West, like the feast of the Kalends or the Rosalia, but they also often mention the specific desires of the donor.Footnote 66
Professional associations in Phrygian cities were not official institutions like the resident Romans of Apameia. But they did approach this status as closely as they could, and were treated by donors and others as part of the Roman civic order. The process was not limited to Phrygia; examples from Lydia lead to similar conclusions.Footnote 67 In the terms of neo-institutional economics, all this can be described as a process of institutional isomorphism. As defined by DiMaggio and Powell, ‘isomorphism is a constraining process that forces one unit in a population to resemble other units that face the same set of environmental conditions’.Footnote 68 The environmental conditions in our case were determined by the Romanization of Phrygian cities. DiMaggio and Powell distinguish between three mechanisms of isomorphic change: coercive, mimetic or normative. For Roman Phrygia, the mimetic aspect was perhaps the most important one; it is normally a result of insecurity with regard to new environmental conditions. Organizations model themselves after other organizations that are observed as successfully coping with the common environment.Footnote 69
One factor contributing to insecurity is a lack of clear legal regulations. Roman laws on collegia certainly fit this criterion. The recent trend to see them as rather irrelevant for Asia Minor needs to be reassessed.Footnote 70 There is sufficient evidence for knowledge and application of Roman law at least in the larger cities of Asia Minor, and a recent find from Miletus has provided unambiguous proof that the ratification of associations by the emperor described in the Digest was practised in Asia Minor at least in the time of Hadrian.Footnote 71 The well-known case of Bithynian Nikomedeia, where Pliny wanted to create an association of fabri but was forbidden to do so by Trajan, can no longer be dismissed as an exception; together with the official ratification of the neoi of Kyzikos and the gerousia of Sidyma, the relevance of this aspect of Roman law for Asia Minor seems to be established for the second century at least.Footnote 72 We should therefore assume that the Digest is correct in stating that the general ban on collegia, a result of their activities during the late Republican civil war, was applicable to the provinces as well, and that certain collegia, those designated licita, could undergo a process of ratification and receive official recognition of their utilitas publica.Footnote 73
However, the precise criteria are unclear, and a number of other rules (especially the exemption for tenuiores and assemblies religionis causa) add to the general confusion.Footnote 74 In addition, the sheer number of associations known from Roman Asia Minor casts doubt on the idea that they all had to apply for official sanction. But were they really the collegia discussed in the Digest? From the epigraphic evidence, it seems that the ratification process was an option for official groups such as the gerousia, the neoi, or certain professional associations with high relevance for a city; in each known case, the process was officially initiated not by the association itself, but by civic or Roman magistrates.Footnote 75 These groups, with fixed numbers, names and purposes, became the ‘legitimate collegia’, to be treated henceforward ad exemplum rei publicae.Footnote 76 As there is no indication that, for example, all the professional associations of Hierapolis underwent a similar process, the question remains how the continued existence of associations without such official legitimacy fits into this picture.
It seems best to categorize them as a class of associations not directly addressed in Roman law — associations that wanted to show their usefulness to the Roman order and profited from it in certain ways, but were not fully integrated into the system of concessions and privileges that characterized the collegia licita. If the endowments of Hierapolis, often directed towards more than one professional association, could be taken to indicate the deceased's membership in several collegia (forbidden for the collegia licita since the days of Antoninus Pius), they would fit this picture well, but it is not clear whether or not this inference can legitimately be drawn.Footnote 77 Such professional associations would have actively sought integration into the civic and the Roman order, but could never be absolutely certain that their usefulness was duly acknowledged, and there even remained the (albeit remote) possibility that unfavourable circumstances might lead a Roman governor to dissolve them. The predictable reaction of such groups to these uncertainties would be mimetic isomorphism: in observing successful models of corporate organization within a Roman civic order, models that were themselves based on the exemplum rei publicae, they became multipliers of that order, duplicating the state, its organizational ambitions and its values.
Insisting on the rôle of Roman law is not to deny that local middle classes had their own reasons to act as they did. While Roman law offered incentives to participate in an imperial world order, its effect on local societies was not — and could not be — planned on a grand scale; it was one factor among others that motivated decisions.Footnote 78 The craving of the middle classes for status and recognition (what has been called ‘ordo-making’) has to be taken into account; the same is true for the increased possibilities of drawing prestige from craft and workmanship in the second and third centuries, visible in epigraphic data from all over the Empire.Footnote 79 But the rapid spread of an institution that is virtually unattested before the Roman period has to be related both to Roman models of civic organization and to Roman law, and institutional isomorphism is a useful concept for taking into account the perspectives of both actors and organizations.
So does Romanization imply the professionalization of the associative order? It may be expected that isomorphic change was much more difficult to manage for cult associations than it was for professional workers. The Roman civic order offered incentives to develop corporate forms of organization, but clearly favoured groups of common public interest. In searching for cult associations, we often have to go beyond south-western Phrygia, into less accessible regions.
IV CULT ASSOCIATIONS AND VILLAGE COMMUNITIES
In contrast to the professional associations discussed above, associations whose members primarily defined themselves as adherents of a certain cult are difficult to find in the cities of western and southern Phrygia. Even the rich finds of Hierapolis hardly include anything of relevance. The significance of one hieros thiasos is in fact unclear; it may have been an association, or a civic institution established for emperor worship.Footnote 80 Other interpretations appear outdated: the sēmiaphoroi of Apollon Archegetes are probably a priestly college, not an association with secret symbols,Footnote 81 and the ergasia thremmatikē was a professional association of shepherds rather than a cult association caring for metaphorical ‘sheep’, Christian or otherwise.Footnote 82 It should be emphasized that speaking of cult associations as a type does not mean going back to a division of associations according to their ‘purposes’, as was common in very early research on Greek associations.Footnote 83 Professional associations were also cult communities. But there are differences in the ways a group described itself in public monuments. In the cities discussed so far, no association is known that explicitly described itself as a cult association.Footnote 84
One possible exception from Laodikeia paves the way for a more thorough investigation. A relief showing Zeus and Hermes was dedicated as a votive offering by the secretary — written γαματεύων — of the psapharoi.Footnote 85 Iconographic and onomastic indications suggest that the monument actually comes from Thiunta, a village north of Hierapolis.Footnote 86 The ‘dry ones’ could be argued to be a cult association for Zeus and Hermes, but the designation remains incomprehensible. It might therefore be the name of one of the many Phrygian village communities. This touches upon an old question: how do we distinguish villages from private associations, and is that distinction always justified? Poland was willing to cautiously accept the idea that villages could ‘wholly adapt the form of cult associations’.Footnote 87 As has often been noted, Phrygia as a whole is characterized by village communities and rural sanctuaries rather than by the great centres of textile production in the south.Footnote 88 Needless to say, no professional associations are to be expected outside the larger cities — economic specialization does not characterize the social history of rural Anatolia. But the epigraphic representation of villages and their religion took over modes and formulae from the cities,Footnote 89 and it is worth asking whether or not some reflection of the desire to ‘get organized’ (and hence isomorphism) can be deduced. The existence of little-understood village gerousiai seems to be a case in point;Footnote 90 cult associations might be another. In order to avoid unnecessary speculation in a field that already abounds with insecurities, the analysis will be limited to associational terminology found in villages.
Phratra, Hetaireia, Synbiōsis
Two stēlai from Thiunta, a village north of Hierapolis, show the twenty-four members of a ‘brotherhood’, phratra.Footnote 91 In both cases, the dēmos of Thiunta has honoured the ‘brotherhood around (peri)’ two chief officials, whose names have changed in the second inscription. The brotherhood has repeatedly distributed oil for eight days and, according to the second inscription, organized a nocturnal ceremony for Zeus. One of the leaders in the first inscription is an agōnothetēs, one of the members is a paraphylax; in the first inscription, one member is recorded not only with his name, but with the additional information that he has bought the place for the stēlē. The term phratra seems to imply more than a mere festive gathering or an organizing committee. In addition, at least two persons in the second inscription can be identified as sons of members in the first inscription, and further links are plausible.Footnote 92 The presence of a paraphylax in a phratra from Thiunta raises interesting questions, as this official should have been operating from Hierapolis; he may have been a local who came to prominence in the nearby city.Footnote 93 In any case, a close connection between the village and the cult group is evident, leading to mutual expectations: the phratra hosts the pannychis, and is honoured by the dēmos.
Also in the area of Hierapolis, the hetaireia of the Arzimneis, again headed by two persons introduced by peri, made a dedication to the gods of the Motaleis.Footnote 94 The monument contains representations of seventeen persons, one of whom (the third in the list) is called a priest. The dēmos Motaleōn is known from other epigraphic evidence; Arzimneis should be regarded as a regional designation as well.Footnote 95 Does this mean that the hetaireia was an association based in a village that honoured the local gods of another village? The process would be easier to envisage if the hetaireia was a subdivision of the Arzimneis, who for diplomatic reasons honoured the gods of a neighbouring village. Something political also seems to be at stake in another monument of this sort, from the area of Dionysopolis: the ‘brotherhood of the Saloudeis and the Melokometai’ unites men from two villages who seem to have joined forces in honouring someone.Footnote 96 In this case, the phratra has been named after the villages, not after its leaders. Two similar monuments have been preserved only partially; they contain a few names, but the designation of the groups has not survived.Footnote 97
The political context casts doubt on the assumption that phratra or hetaireia were used in such monuments to designate private cult associations.Footnote 98 One should also ask how large the population of such villages was. For the Byzantine period, Thonemann has calculated an average population of twenty-five to twenty-seven inhabitants for villages with corporate institutions and a functioning communal life in the Maeander valley.Footnote 99 Numbers in the imperial era were certainly higher, but a group of seventeen to twenty-four men could still represent a significant part of the male population. As there seem to be no brothers or father-son pairs in the group at the same time, twenty-four men may well represent twenty-four families. Similar calculations may lead to a reassessment of an inscription from Kayılı, a village close to Akmoneia, where ‘the synbiōsis’, consisting of nineteen men, ‘splendid and noble Achaeans’, dedicated an altar to Zeus Alsenos so that he might be kind towards the village.Footnote 100
The number of members seems too high for mere private associations, at least within a rural context. We should entertain the possibility that these groups represented the villages as cult communities before their gods. To be sure, in some cases, a verbal distinction is drawn between the group and the village.Footnote 101 While the qualifying genitive (the hetaireia of the Arzimneis) is not decisive (we would also expect ‘the dēmos of’ …), other formulae distinguish either the dedicant from the beneficiary (the synbiōsis vs. the katoikia) or the honorand from the body that conveys the honours (the dēmos of Thiunta vs. the phratra). But this does not run counter to the idea that such groups united male inhabitants — perhaps a select group in larger villages like Thiunta, almost all of them in smaller ones — who acted on behalf of the whole village population, represented their communities before the local gods, and thus legitimately claimed to be ‘the’ cult association of a given village. The peculiar visual representation of the members in several inscriptions, with its blending of individual elements (names, beards, priestly status) with stereotypical ones, unparalleled in other regions, may have had a double function in this regard: it honoured the men thus represented, but at the same time it named the persons responsible. The creation of corporate identities using established terminology would also have facilitated the representation of villages in the context of trans-local rural cults, well attested in Phrygia and neighbouring regions.Footnote 102
Ramsay, who took the Phrygian phratrai as private associations, argued that they were an age-old Anatolian institution.Footnote 103 More recent findings provide us with some comparative data. Cult groups that called themselves phratrai are known from other regions in both Asia Minor and the Near East, and although it is not always clear whether or not they were civic subdivisions, at least in some cases the private nature of the group is probable. In Phrygia, the term phratores is used in a not very informative inscription from (possibly) Dorylaion, and in a difficult text from Orkistos, where it refers to a group of unclear status involved in a dedication.Footnote 104 Another inscription, presumably from the area of Hierapolis, records a memorial for a certain Diodoros set up by ‘his own phratra’, which allows for several interpretations.Footnote 105 Two inscriptions use designations close to the one employed by the dēmos of Thiunta in order to define organizations of unclear status.Footnote 106 But the group of monuments discussed here differs from these in style and content.
It is evidently possible that communal action vis-à-vis the gods took place in pre-Roman institutions. But the choice of terminology as well as the epigraphic commemoration that starts in the second century c.e. seems to betray the influence of Roman models. If villages observed the epigraphic trends of the cities, as we know they did, but could not follow the trend towards ordo-making due to lack of social diversification, terms like hetaireia or phratra were useful not least for their ambiguity: they could refer to private associations,Footnote 107 but did not have to, and they signalled close ties among the members more clearly than a term like koinon. Villages searching for prestigious ways to epigraphically represent (a significant part of) their male population as a cult group could have made far less convincing choices.
Mystai
The search for cult associations has thus uncovered an inherent — and perhaps consciously created — ambiguity in the epigraphic evidence. We may hope to reach safer ground when discussing another phenomenon that is important in Phrygia, but also known from other regions of Asia Minor, namely, the spread of ‘initiates’ (mystai) in the imperial era.Footnote 108 In Phrygia, the first peculiarity to note is that there are hardly any mystai in the larger cities where one would expect to find them. One hierophantēs of Dionysos Kathegemon is now known from Hierapolis, but there is no indication that he belonged to an association.Footnote 109 Inscriptions from other regions are more likely to have been set up by associations. In a village next to Akmoneia, the ‘mystai of the first, holy thiasos’ dedicated their meeting place to Dionysos Kathegemon, and a group of mystai from Apollonia honoured an otherwise unknown person.Footnote 110 In a village 10 km north-east of Sebaste, ‘the initiates’ set up a bust for Dionysos Kathegemon and published a list of members (second/third century c.e.); only the names of the two priests and their respective sons have survived, but they indicate leadership by local élites.Footnote 111 Further evidence comes from rather remote places in the Upper Tembris valley, between Nakoleia and Dorylaion. A number of quite similar, four-sided altars from the second and third century were erected for Zeus Dionysos by groups of initiates, but also by individuals.Footnote 112 It has been argued that these monuments attest ‘an old local cult, which was taken up, and perhaps enlarged, in Roman times’.Footnote 113 While the old age of a cult for Zeus Dionysos in this region cannot be proven, the organization of worship is decidedly Roman, if the use of associational terminology characteristic of Romanized regions is regarded as an indication. Altars are set up by a group of mystai that had a priest, a hierophantēs and a speirarchēs, by neoi mystai who underline either the recent creation of the group or the age of the members, and, in one case, by a family that honoured its deceased father, assisted by a hiera speirē.Footnote 114
It is in this context that we again encounter designations with a strong local flair. The mystai Koroseanoi neobakchoi dedicated an altar to Zeus Dionysos ‘for themselves and the village’; other dedications were made by Ptolemēnoi mystai and Krēouerēnoi mystai.Footnote 115 The attributes clearly refer to villages otherwise unknown. Given the frequency of references to ‘initiates’ in Asia Minor and the inflation of cultic language with vocabulary derived from the mysteries,Footnote 116 there is little reason to believe that all these μύσται had undergone formal procedures of initiation.Footnote 117 The spread of μύσται-terminology in Asia Minor was probably influenced by the privileged rôle accorded to the mysteries in the representation of ‘imperial religion’.Footnote 118 These groups could be the only associations of their — otherwise epigraphically unattested — villages. But in light of the evidence discussed above, it is tempting to assume that this is another case of villages describing themselves (or rather a representative part of the male population) as a cult community by applying corporate terminology that was known to be acceptable in Roman contexts.
This hypothesis may also shed some light on yet another enigmatic corporate designation, from Amorion in eastern Phrygia. A certain Antipater made endowments to ensure the commemoration of his deceased daughter Kyrilla.Footnote 119 The recipient is a group called phylēs Dios mystai, ‘initiates of the tribe of Zeus’. This could be an association for Zeus, but no cult of Zeus is provided for in the inscription, although it appears that the group came into being for the purpose of the endowment.Footnote 120 Might we be dealing with mystai from a civic phylē?Footnote 121 Since both public and private groups could be involved in private grave care, the case must remain open. But this might be another example of the overlap between kinship terminology, institutions of villages, and corporate designations taken over from the (Graeco-)Roman civic order.
Bennos, Doumos
The conclusions reached so far touch upon the question of how indigenous, pre-Roman traditions were transformed under Roman rule. As all available data stem from the Roman period, reconstructions of ancient Phrygian institutions are obviously liable to ungrounded speculation. The supposed origins of professional associations in ancient Anatolian societal subdivisions has been mentioned above; more specific arguments include the idea that the washers of Roman Hierapolis are to be connected to Hittite temple slaves, supposedly organized in associations more than a millennium earlier.Footnote 122 The use of Neo-Phrygian terms, albeit problematic in itself, may serve as a more trustworthy guide to indigenous concepts. In the case of associations, we are again led to the north and east of Phrygia.
Several inscriptions from Phrygia and a few from Bithynia mention to bennos and Zeus Bennios. Due to the work of Drew-Bear and Naour, it is now generally acknowledged that bennos was a cult association, to be derived from indo-european bendnos (‘covenant, union, association’).Footnote 123 This interpretation has led to a satisfactory understanding of almost all the relevant monuments. An altar from the area of Aizanoi may serve as an example: ‘Tryphon, son of Meniskos, (has set up the altar) for Zeus and the benneitai’.Footnote 124 The text is perfectly intelligible if bennos is an association; benneitai are its members. In a village near Nakoleia, two brothers dedicate a crown to the bennos of Zeus Bronton.Footnote 125 A text from Krateia in eastern Bithynia also shows that a bennos could be the source of benefactions, appropriately returned through the erection of an altar dedicated to Zeus Bennios.Footnote 126 And in a metrical text from Midas City, benneuein designates a positive state, possibly harmony and unity among citizens.Footnote 127 Problems nevertheless remain. The first is one that we have already encountered above. Again in the area of Nakoleia, a certain Markos, son of Markos, dedicated a crown to Zeus Bronton and the bennos Sereanon. The Sereanoi are a village west of Nakoleia, not an association, so the question of villages posing as associations is raised yet again.Footnote 128 A similar, but earlier case is a dedication made by a private person to the bennos Soēnōn for the victory of Trajan.Footnote 129 The second problem is raised by the endowment of a certain Skymnos from the very east of Phrygia, on the border with Pisidia. He left 100 denarii to the village, ‘in order that from the interest a bennos is made in honour of Zeus Kalakagathios, for the fruits’.Footnote 130 In this case, bennos seems to be a regular feast, to be financed ek tokou like the annual coronation of graves in Hierapolis — not an association.Footnote 131 It therefore seems that the meaning of the word was rather broad; it could designate a cult community, but also a feast or festive assembly.Footnote 132 The argument that a bennarchēs must always be the leader of an organization is not sufficient to rule out this possibility.Footnote 133
The evidence does not allow for a history of the bennos to be written, as all relevant inscriptions seem to be roughly contemporary (second/third century c.e.). But it is perhaps possible to develop a plausible historical scenario in which the meaning ‘festival’ or ‘festive gathering’ is the original, indigenous, pre-Roman one. In this sense, bennos was a Phrygian term to designate cult communities, without a clear distinction between the participants of the feast and the feast itself. When in the Roman period models of corporate organization that were based on membership became widespread, some inhabitants of northern Phrygia adapted their indigenous terminology to match new social structures; βεννεῖται could now designate formal organizations. The process could be described as a re-interpretation of Phrygian traditions under Roman influence. But the original meaning was not forgotten, and so the village of the Sereanoi could describe itself as a cult community by using the term bennos, and Skymnos could expect to be understood when he ordered that a regular bennos should be performed out of the interest on his endowment.
This hypothetical scenario may be strengthened by pointing to the parallel development of the Syrian marzeah/marzeha. In Ugaritic texts and the Old Testament, but also in inscriptions from the fourth and third century b.c.e., the term may mean both a group of people and a feast.Footnote 134 Under Hellenistic and Roman influence, it came to designate organizations with defined membership, as in Nabatean inscriptions of the first century b.c.e.Footnote 135 The evidence from Roman Palmyra suggests that both meanings could co-exist.Footnote 136 Romanization could lead to a certain formalization of indigenous religious life, but the results were often hybrids, as should perhaps be expected.
The persistence of local peculiarities in less Romanized areas such as northern Phrygia shows the limits of this process. Even here, Roman law was present in one way or another,Footnote 137 but the incentives it provided must have been less relevant for both sides than they were in the coastal cities of Asia Minor, or the assize centres of central and southern Phrygia. One last example may illustrate the difference: this concerns another indigenous term, doumos. That the word is Phrygian in origin is now generally accepted. It clearly designates ‘cult associations’, often related to the cult of Cybele in inscriptions from Lydia, Serdica and Thessalonike.Footnote 138 In Phrygia itself, the term is attested only twice.Footnote 139 In a very fragmentary dedication from a rural sanctuary at Phyteia, the meaning may be ‘household’;Footnote 140 more interesting for us is a bilingual inscription from Dorylaion:Footnote 141
ε---γ/τεντουμενος | νιοισιος ναδροτος | ειτου Μιτραϕατα | κε Μας Τεμρογε|ιος κε Πουντας | Βας κε ενσταρνα | [vac.] δουμθ κε οι ουθ|βαν αδδακετ ορου|αν. Παρεθέμην τὸ | μνημεῖον τοῖς προ|γεγραμμένοις θε|οῖς κὲ τῇ κώμῃ·| ταῦθ’ ὁ πατὴρ Ἀσκληπιός.
[Greek part]: I have set up this memorial before the above-mentioned gods and the village. The father Asklepios has done this.
It seems clear that δουμθ is a dative, which makes κώμη the Greek equivalent.Footnote 142 So even a Phrygian term that is used elsewhere to designate cult associations is used in Phrygia itself only to qualify a village as a religious corporate actor.Footnote 143 The evidence for both bennos and doumos thus strengthens the observations made above: instead of the social fragmentation that Romanization caused in the cities of western and southern Phrygia, the main parts of the region, while not unaffected by the trend towards corporate representation, stand out for their use of terms and concepts that indicate social cohesion. If villages employed terminology from the sphere of associations, their aim was not to establish a segregative social order. The spread of terminology and concepts of corporate organization was nevertheless related to Roman influence. It reflects the tendency to assume a corporate identity that could be displayed on stone. But it was not a cultural transfer that profoundly changed the realities of village life. The villages and their élites followed the urban attempts to ‘get organized’, but at their own pace, and according to their own priorities.
V CONCLUSION
Romanization was a process that affected the institutional inventory of cities. I have argued here that the field of ‘private’ corporate organization is an important indicator for the changes brought about by Roman rule. Apart from the ‘resident Romans’ and corporations of old men, evidence from the cities in western and southern Phrygia clearly shows the growing importance of groups modelled on the Roman collegia. Professional associations, almost unattested in Asia Minor before the Roman period, became an important part of civic life, often acting as junior partners of the official political bodies. The influence of Roman ideals of social organization seems to explain this process better than other models.
For rural Anatolia, different results are to be expected due to the different economic and social framework. The evidence does, however, show an awareness of current urban forms of Romanization both within and outside Phrygia, and a desire to adapt them to the very different conditions of rural life.Footnote 144 The view that a frustrated Phrygian population used local religious traditions to shut itself away, and to express rejection of the dominant Graeco-Roman culture,Footnote 145 does not adequately address the influences of that very culture that have been noted here. It is rather an example of a ‘celebration of the local’ triggered by the imposition of an imperial superstructure, visible in many different manifestations throughout the Empire.Footnote 146
Recent debates both on associations and Romanization have often focused on desires for status and representation, which leaves responsibility for the process solely to the local actors. These categories have successfully illuminated the data, and the model envisaged here takes these developments into account. It is nevertheless necessary to look for legal incentives as well, not least because recent findings have added to our knowledge about Roman laws on associations. We should certainly avoid the pitfall of taking the fragments of legal expertise preserved in the Digest as timeless and normative, but it is also not satisfactory to deny any influence of Roman law on the development of associations in the East. We should take into account the fact that Roman law provided not only guides for restrictive action, but also incentives that local groups could respond to on their own terms.
This finally raises the question of whether or not there was a masterplan. Were the developments discussed here the planned result of a centrally administered policy of Romanization? The answer to this simple form of the question will have to be an emphatic no. Still, the idea that Roman administrators were not uninterested in the effects of institutional isomorphism, namely isomorphic pressure that would lead to a certain structural unification, seems to me to have some explanatory value — not least because in less diverse regions than Phrygia, the model is likely to work even better. Thus, the professionalization of the associative order in the Ionian cities, especially Ephesos or Smyrna, is fairly evident. Creating adaptive pressures without having to exert them directly is a reasonable political strategy. While Roman governors certainly were unable (and probably did not try) to precisely predict the outcome of such a strategy, a case for centrally-managed impulses for Romanization can and should be made.
