Saint George the Anatolian, Master of frontiers Maria Couroucli To cite this version: Maria Couroucli. Saint George the Anatolian, Master of frontiers. Sharing sacred spaces in the Mediterranean. Christians, Muslims, and Jews at Shrines and Sanctuaries., , pp.118-130, 2012, 978- 0-253-22317-3. ￿halshs-01218190￿ HAL Id: halshs-01218190 https://shs.hal.science/halshs-01218190 Submitted on 20 Oct 2015 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. 1 Saint George the Anatolian, Master of frontiers Maria Couroucli, CNRS in: Albera, D. and M. Couroucli (eds), Sharing Sacred Spaces in the Mediterranean, Indiana University Press, pp. THE VALIANT MAID (MOTHER OF DIGENIS) AND THE SARACEN1 There is fighting in the East and fighting in the West Learning of this, a beautiful woman goes off to war. Dressed as a man, the beautiful warrior takes up her arms. The saddle of her horse is decorated with serpents, The harness fastened with vipers. With one prick of the spur, her horse covers forty leagues, With a second, she is in the thick of the battle. Elle moves on and they get out of her way. She emerges and no one sees her. As she turns about, the straps break And reveal the gold apples hidden beneath the linen. The Saracen sees her from the top of a mount. ‘Do not be cowards, my gallant men, War is feminine, the prize is a bride!’ La belle hears him and runs to seek refuge with Ai Girogi. ‘My lord St George, hide me, virgin that I am. And I will build you a gold door to go in And a silver door to go out. The wooden roof will soon be covered in pearls.’ The marble splits open and she goes in. The Saracen comes to St George. ‘Deliver the virgin unto me, St George the Christian. By your grace, I will be baptised, and my soon too. They will call me Constantine, and my son Yannis.’ The marble splits open and the damsel appears. The existence of ‘shared’ holy places in the Mediterranean, and more especially the post-Ottoman world, relates to one of the general questions raised by both ethnology and history: the tension between the two principles that organize society, namely kinship and territory. Is a social group attached to a particular locality and defined in relation to a territory, or is it primarily organized around ties of kinship, descent and alliance (Leach 1982; Goody 1990; Derouet 1995)? Is it possible that the distinction between ius soli and ius sanguinis, which operates elsewhere, ceases to be pertinent when we are dealing with holy places, or in other words with the sites of important cultural activities that presuppose the existence of culturally significant 1 From Politis (1978 [1914]), ‘Marching Songs’, no 32, trans. Marianne Gokalp. Couroucli 2012 St George the Anatolian 2 representations derived from a specific organization of time and space? And what of claims pertaining to such places in the current political context? The starting point for this article was an ethnographic study of such a place during the annual feast of St George in contemporary Istanbul.2 It very quickly became clear that an understanding of the ethnohistorical context, and especially of these practices in terms of the Ottoman tradition, could add to our ethnographic knowledge. This essay is therefore part of an on-going historical anthropological study of the Anatolian Greeks, and deals with a phenomenon that still remains largely unknown.3 The second aspect of this study concerns the symbolic world with which cultural practices centred on St George are inscribed. The existence of the saint is analysed as the organizing principle behind a space-time that is bound up with the local community and that supplies a representation of the relationship between men and nature, namely the calendar of the year’s local feasts. These feasts often mark the seasons, that is to say the beginning and end of work in the fields and other important activities such as transhumance, navigation and war. Those activities take place in the summer months, and in this region their start and finished coincide with the cycle of the Pleiades, which is itself associated with a mythological set, elements of which are also found in ancient Greek traditions and in the traditions of the peoples of both Anatolia and the Middle East. HOLY PLACES AND LIMINALITY Anthropological analyses of religious rituals often make use of the concepts of communitas and liminality, which, Turner (1969) suggests, can be applied to the study of religious phenomena in an extension of Van Gennep’s work on rites of passage. According to this analysis, individuals who take part in a common ritual form a communitas but place themselves outside society in a state of liminality in which the rules of everyday life are in a sense suspended or inverted. Turner also describes this state as an ‘antistructure’. A communitas emerges when individuals either ignore the rules that organize their society or find themselves outside the established social structure, and organize community bonds around cultural practices. This theory, which has been criticised by many researchers, proves to be much more convincing when it is related to the ethnological data on Christian places of worship in those 2 Earlier versions were presented to the European network REMSH’s conference on ‘Networks, Exchanges and Conflicts in the Mediterranean Area’ (Athens, 26-28 May 2005), the Max Planck Institute’s workshop on ‘Eastern Christianities in Anthropological Perspective’ (Halle, 23-25 September 2005), and at the day conferences on shared shrines organized by the Laboratoire d’Ethnologie within the REMSH network (Nanterre, March 2006). My warmest thanks go to all those who took part for their comments and suggestions, and especially to Dionigi Albera, Jean-Pierre and Marlène Albert, Michèle Baussant, Chris Hann, Roger Just, Antoinette Molinié and Anna Poujeau. 3 The first anthropological study of Anatolian Greeks was carried out by Renée Hirschon in 1960-1970 in a refugee district of Piraeus, near Athens (Hirchson 1989). Couroucli 2012 St George the Anatolian 3 regions of the East where Christians and Muslims have coexisted for centuries.4 Turner appears to know nothing of the practices of Eastern Christians, and still less of inter-confessional situations, yet it is precisely in such situations that the notions of communitas and liminality take on their full meaning (Brown 1981: 42; cf Pujeau’s contribution to the present volume). Turner regards religious cults as transitional phenomena characterized by s state of liminality. Because they are ‘betwixt and between’, they mark an ambiguous period that is conducive to the formation of a communitas, or a non-structured community whose members are all equal. To some extent, this analysis takes us back to the work of Hasluck, who carried out research into what he called ‘ambiguous sanctuaries’ –sanctuaries of Christian origin managed by Muslim brotherhoods, and especially the Bektashis—in the Balkans and Anatolia in the 1910s (Hasluck 1929). Hasluck, like Turner, thought that such ‘shared’ sanctuaries represented a transition from one stage to another: the transformation of a Christian holy place into a Muslim holy place was a transition within a broader process of Islamization in the Ottoman world.5 We will see later that this hypothesis does seem to be corroborated by recent research into the period of the Ottoman conquest of, for example, Thrace; at the same time, it seems difficult to assert that we are dealing with a place that is ‘in transition’ whenever we encounter a ‘shared’ holy place of a ‘mixed’ pilgrimage. What might be called the Byzantino-Ottoman zone is a vast territory in which the populations of various regions experienced many periods of Islamization over a period of more than one thousand years. It therefore seems to me that the phenomenon is more complex than a ‘moment’, or a mere changing of the guard, and that it would be more useful to 4 Turner’s theories have recently been criticised by the many anthropologists who have been unable to confirm his theses about the liminal state or the establishment of the anti-structure (communitas) in the many cases of pilgrimage they have observed all over the world. Cf Morinis (1992), who refers to his studies of Bengal, and to the work of Eickelman in Morocco, Pruess in Thailand, of Messerschmidt and Qharma in Nepal, of Van der Veer in India, of Sallnow in Peru and of Pfaffenburg in Sri Lanka. Dubisch (1995), who has worked in Greece, is another critic who has doubts about the pertinence of the concept of communitas. All insist that pilgrimage is an individual affair, that it is experienced as such, and that there is no element of communitas in the process. In their critical introduction to theories of pilgrimage, Eade and Saalnow (1991) emphasize the anti-Durkheimian nature of Turner’s theory; by placing the emphasis on the anti-structural nature of pilgrimage, they call into question the notions that its function is to reinforce social cohesion. Eade and Sallnow also stress that the phenomenon of pilgrimage is not uniform, which is why any theoretical discourse about ‘pilgrimage’ will inevitably be incomplete: ‘If we can no longer take for granted to meaning of a pilgrimage for its participants, one can no longer take for granted a uniform definition of the phenomenon of “pilgrimage” either’ (1991: 3). Turner himself actually thought that pilgrimages are essentially intra-confessional activities, and paid almost no attention to inter-confessional pilgrimages: ‘With rare and interesting exceptions, the pilgrims of different religions do not visit one another’s shrines, and certainly do not find salvation extra ecclesia’ )1978: 9). 5 Hasluck describes Bektashi cultural activity at these shared shrines as a ‘superimposition’ of religions. In his view, Bektashism spread through the local populations in non-violent ways: ‘Either by process analogous to that known to the ancient world as the “reception” of the new god by the old, or simply by the identification of the two personalities. The “ambiguous’ sanctuary, claimed and frequented by both religions, seems to represent a distinct stage of development –the period of equipoise, as it were—in the transition both from Christianity to Bektahism and, in the rare cases where political and other circumstances are favourable, from Bektashism to Christianity’ (Hasluck 1929: 564). Couroucli 2012 St George the Anatolian 4 examine in detail one instance and to contextualize our ethnographic observations with the help of the work of historians, if we wish to understand the nature of these strange phenomena. My analysis is inspired by the approach adopted by Peter Brown, the only person to have described the phenomena of the cult of saints throughout the Christian world, in both the east and the west, whilst paying particular attention to their historical contexts (Brown 1981; 1987). The present study concerns St George, who is very popular throughout the Christian east from Egypt to Georgia. Several of the shrines dedicated to the saint in the eastern Mediterranean attract Muslim pilgrims (Voile 2004). My argument is based upon a series of ethnographic observations of the saint’s feast in the sanctuary dedicated to him on Princes’ Island (Prinkipo or Büyükada, meaning ‘large island’ in Turkish), which lies opposite Istanbul. They will be analysed in the light of a study of documents about the saint’s feast days in the Black Sea region at the beginning of the twentieth century. They are drawn from the archives of Greeks who fled Anatolia in 1922.6 SYNGHENEIS AND SYNCHORIANOI: THE PRINCIPLES OF KINSHIP AND TERRITORIALITY An analysis of the ethnographic data on the ways St George’s day is celebrated in the eastern Mediterranean takes us back to two notions that are essential to the analysis of social facts: kinship and territory. Those notions relate to the principles and blood and soil which, whilst they are a priori antinomic, are associated in most societies (Kuper 1994). The complementary nature of the two principles is also evoked with reference to agrarian societies in Europe, where lineage the one hand and residence on the other appear to determine the rights of individuals. Kinship and locality appear, in other words, to constitute the conflicting logics around which land-tenure, for example, is organized (Derouet 1995). Now, as Leach demonstrates, it is not easy to maintain the distinction between the two referents when we are analysing ethnographic data. Leach ‘deconstructs’ the concept of kinship and suggests that this type of relationship should be examined in terms of its material context, rather than as a behaviour that obeys a set of juridical rules. Kinship systems, long regarded as a basic ‘structure’ of social organization in the British tradition, ‘have no “reality” at all except in relation to land and property’ (Leach 1968: 305). We should not, in other words, confuse kinship organizations with representations relating to them. The same type of ‘deconstruction’ might shed light on the study of ‘shared’ or ‘common’ holy places within the Ottoman world if we place the emphasis on the 6 I prefer to use the archaic ‘Prinkipo’ rather than the more common ‘Büyükada’ when referring to the island because it is more evocative of its multicultural past. The term is part of the local vocabulary, and is understood by old Stamboulites of all confessions. In Greek, the grammatically correct term is Prinkipos/Пριγκηποσ and, for the islands as a whole, Prinkiponoisa/Пρικηποσά, (Princes’ Island). The reference is to the Byzantine princes (and subsequently the sons of the sultans) who lived there under house arrest. In modern Turkish, the islands are generally referred to as ‘ the Adlar’ (islands), and tourist guides use both terms. See, for example, http://www.guideofistanbul.net/en/adalar/htlm; Istanbul, Guide bleue, Paris: Hachette 2002; Istanbul, Paris: Guide Gallimard 2002. Couroucli 2012 St George the Anatolian 5 distinction between the analysis of social organization in relations to the religious communities (millet) that were established with the Ottoman juridical and administrative system, and the analysis of the religious practices and representations of local communities.7 This raises the question of the fit between juridical affiliation and religious practice, which underlies our investigation into ‘shared holy places’. The presence of Muslin pilgrims at the shrine of a Christian saint seems to contradict a certain conception of the millet that sees the cultural practices of individuals as being ‘in phase with’ their membership of a given religious community, which determines their juridical status. We will come back to this point. Now we cannot analyse this phenomenon without calling into question our understanding of the local categories involved: practising a ‘religion’, belonging to a ‘community’, ‘sharing’ an identity or holy places, and being from a ‘country’ [pays], in the sense of a locality, are concepts that must be contextualized within the historical and sociological space- time that concerns us here. A study of the stories collected in the archives of Greek refugees from Anatolia reveals the presence of a strong local identity, and a feeling of belonging to a village community that crossed the frontiers between religious communities in the early twentieth century. That feeling of identity and belonging is the basis for the bonds of solidarity that are forged people who come from the same village, or who are regarded as synchoriani, from the same chorio (village). ‘When a Muslim took the decision to expatriate himself, he was given protection and help by his expatriate Christian synchorianos.’8 These strong ties have to be related to the entire population’s involvement in the annual commemorations of local saints that mark the transition from one season to the next. The fact that an entire population recognizes the sanctity of a holy place is, in other words, merely one aspect of the coexistence within the same territory of different religious communities, or an expression of the ties between individuals and their local community. The local Anatolian population is a social reality that is not frozen in time: whether it be monoreligious or plurireligious, it is a product of centuries of population movements and religious conversions. When we come to look at the question of ‘sharing’ as a tradition, we will see the importance of ethno-historical studies, and 7 On Greek Orthodox millet see Anagnostopoulou’s recent study (1997). 8 CEAM, OD, 56/1959, Rapport de mission Tsalikoglu, p. 79. The archives in the Centre d’Etudes d’Asie Mineure (CEAM) were collected during the exchange of populations between Greek and Turkey that followed the war between the two countries in 1919-1922, and which ended with the defeat of the Greek army. Both nation-states then pursued a policy of homogenization within their frontiers: some were ‘expatriated’ and others were ‘expelled’. 1,200,000 ‘Greeks’, or member of the Ottoman’s Empire’s Greek Orthodox minority, were repatriated to Greece, whilst 350,000 Muslims were sent to Anatolia. The departure of the Greeks marked the end of a process that began with the Armenian genocide of 1915. In 1913, one in five inhabitants of what is now Turkey was Greek; by the end of 1923, the proportion was no more than one in forty (Keyder 2002: 43). The historical context was that of the end of the Ottoman Empire and the beginning of the formation of the modern Turkish nation- state. Collecting the ethnographic data that make up CEAM’s archive was a scientific project carried out using the ethnographic methods of the day (1940-1970). The project was initiated by Melpo Merlier, an ethnomusicologist who was married to Octave Merlier, Director of the Institut français d’Athènes. Couroucli 2012 St George the Anatolian 6 especially those of Hasluck, who studied heterodox Islamic practices in the Balkans and Anatolia in the early twentieth century. For the moment, we will simply note that Hasluck, whose analyses have inspired many recent studies of religious minorities in the Ottoman Empire, believed that heterodox religious practices should be related to the periods, which vary in length, over which the populations concerned were converted from Christianity to Islam (Hasluck 1913-1914; Shankland 2004). COUSINS (SYNGHENIS) AND BLOOD RELATIONS (OMOEMI); TWO DISTINCT CATEGORIES In the Ottoman Empire, religious communities coexisted in the context of a juridical and administrative system that implied several levels of segregation, that found mixed marriage inconceivable and that prescribed rights and duties that varied in accordance with the individual’s ethnoreligious affiliations. According to the principles of Ottoman law, communitarian endogamy was an absolute rule, and mixed marriage resulted in the wife’s conversion to her husband’s religion and her exclusion from her community of origin. Under these conditions, marrying outside one’s community meant defying all social logic; this could put an end to cycles of alliance, and represented a ‘loss’ for both the family groups concerned. The criteria applied during the population exchanges of 1923 followed the same rules: all those inhabitants of Anatolia who were included in Greek Orthodox parish registers when they were baptised were expelled to Greece; only those Christian women who had married Muslims had the right to remain by virtue of their conversion –they no longer belonged to the Rum millet or the Greek Orthodox ‘nation’.9 It is well known that ties of biological kinship do not necessarily result in social bonds. In this case, the effect of the regulation of ‘impossible exogamy’ meant that ties of kinship were recorded in two registers with the local community. On the one hand, they left administrative traces in the parish registers of Christian churches, and in the registers of the cadis who represented the administration at the local level. On the other hand, they left traces in the collective memory of relations between men and women. In the Greek world, both registers were visible in the traditional distinction between the terms synchorianos (‘from the same chorio or village) and synghenis (of the same genos or lineage): being from the same villages did not necessarily mean being of the same family. But in the case of mixed localities in Anatolia, a further distinction existed between individuals of the same parentage (synghenis) and individuals who shared the same blood (omoemi). The distinction took account of local social realities: it was possible for individuals to be biological 9 The passports of the Rum who were ‘exchanged’ were stamped (in French): ‘Il ne peut retourner’ (‘cannot return’). The refugees from Anatolia reached Greece in two waves. The first wave followed the fighting between the Greek and Turkish armies. It involved coastal populations who left in a hurry and often in tragic circumstances, including those who were evacuated after the burning of Smyrna in 1922. The second, and larger, wave involved populations from the interior, who left in more organized fashion in 1924. The Cappadocian populations mentioned in Tsalikoglu’s report were part of the second wave (Anagnostakis and Balta 2004). Couroucli 2012 St George the Anatolian 7 kinsmen without necessarily being related in the social sense.10 One example of this third referent –which appears to have been necessary in regions where Christians and Muslims lived side by side—can be found in Tsaliloglu’s report, and more specifically in the passage describing his visit to his village of origin in Cappadocia in 1959. The author describes how a young Muslim man addressed him by calling him a ‘cousin’ without realizing that has was merely ‘of the same blood’ He introduced himself: I am the grandson of Katir Baba, and the son of Osman Cavoul Katir. We are cousins (he meant blood relations, οµάιµουές). On our father’s side we are descended from the Pinyatoglus, and on our mother’s side from the Karakasoglus in our village […] Katir Baba was born in 1825, in the time of the janissaries. His tragic story is known to all Christians. [Descended from two great families], he lost his father when he was little and his mother married a Turk from our village and converted. She kept the boy with her. He grew up as a Muslim and was circumcised. But he knew about his Christian origins. It was not possible for him to revert to Christianity on pain of death […]. The rich Pinyatoglus regarded him as one of their own. They helped him financially. When they came back to the village from trips abroad, they remembered him. They brought him presents […] The Pinyatoglus had made the Muslims of the day an offer: they would keep the son of the converted mother and who buy him by paying his weight in gold. The janissary did not accept their offer, saying that when he grew up, he could earn more gold than the Pinyatoglus were offering. Young Katir (therefore) called me his cousin. We are not cousins (Συγγευέιάυ δευ έχοµεν). The poor boy probably meant to say that we were blood relatives (όµάιµουέυ) … blood can never, never, turn into water. Today, more than ten of the forty Muslim families in Cincidere are of Greek origin. The third generation does not hide the fact (Tsalikoglu: 125-128). SYNCRETIC11 PRACTICES IN ISTANBUL: THE FEAST OF ST GEORGE 10 Tsaliogklu’s text gives an example from twentieth-century Cappadocia. This is comparable with the Cretan tradition in which the head of the family’s religious affiliation with the ‘dominant religion’ (the community of ‘Latin’ Catholics during the Venetian period, and the Muslim community after the island’s conquest) did not lead to the conversion of the rest of the family. According to Greene (2000: 108), the Cretan system functioned because of a ‘public religion’ (Catholic Christianity, Islam) transmitted through the agnatic line coexisted alongside a ‘private religion’ transmitted by the women of the family. In this society, people who found themselves astride the two communities stood a better chance of success (Greene 2000: 204). Having said that, the documents on which historians work (contemporary chronicles, local archives) tell us nothing about the cases that are of most interest to us for our present purpose. What about festive meals? Who shares which meal, and with whom? Which members of the family are invited to celebrate the head of the family’s name day when the family is bi- religious? But they certainly suggest that –in the broadest sense— Christians and Muslims were more likely to ‘share’ –in the broadest sense- at religious feasts (Greene 2000: 106-108). 11 Syncretism is used here to means mixed with Muslim traditions, and to describe practices that bring together elements from several religious traditions. On the etymology and various meanings of the word see Shaw (1994). Couroucli 2012 St George the Anatolian 8 April 23 has been a public holiday in Turkey since 1929. It is ‘National Sovereignty and Children’s Day’ (23 Nisan Ulusal Egemenlik ve Çocuk Bayrami), and commemorates Kemal Ataturk’s seizure of power in 1920. According to the Christian calendar, this is St George’s Day, which is a traditional festival for Turkey’s Rum.12 In the countryside, it is celebrated by masses in chapels or monasteries, followed by open air meals or picnics. In today’s Istanbul, impressive crowds of pilgrims –some 100,000 people—gather for the festival around the Greek Orthodox monastery at the highest point of the island of Prinkipo. The fact that almost all the pilgrims are, in cultural terms, Muslim means that the practices observed there are syncretic. The large number of pilgrims also raises the issue of the religious nature of the event: are we talking about the emergence of a communitas that exists for the duration of these symbolic practices, which are bound up with old sacred places? (Hertz 1970; Turner 199; Morinis 1992) How do the ‘communities’ concerned define their identity? How do tensions and passions whose roots lie in a shared history relate to contemporary issues pertaining to identity, recent modifications to representations, and national or nationalist stereotypes (Kechriotis 2002; Calotychos 2003; Pagaroufali 2005; Theodossopoulos 2006)? ‘Ritual time’ and ‘sacred place’ are fluid realities, and I propose to look at them by examining three calendar configurations, one corresponding to ‘communitarian’ time, and others to ‘shared’ or syncretic time. The St George’s Day celebrations of 23 April are a well established local tradition that goes back to the multicultural past of Ottoman society. At the beginning of the twentieth century, Christians, most of them living in Istanbul and towns on the western coast, made up 20% of the population. The population of the city of Istanbul has now risen to some ten million and is 99% Muslim. This situation contrasts sharply with the early twentieth century, when the city had a population of scarcely one million people who spoke several languages and belonged to several religious communities. This was la belle époque and it has often been described as ‘cosmopolitan’ (Driessen 2005; Ors 2006): rich urban families of all religions shared the same way of life in a city where half the population consisted of Greek (Rum), Jewish or Armenian ‘minorities’ (Alexandris 1983; Berktay 1998). Princes’ Island, which one hour away from the centre of Istanbul by steamboat, is where, amongst other places, where the Stamboulite bourgeoisie spent their holidays in the nineteenth century.13 Most of the inhabitants were from ‘minorities’, and the majority were Greek Orthodox, which explains why there are two Orthodox parish churches, a cemetery and a monastery dedicated to St George on Prinkipo. It was already both a popular place of pilgrimage and a playground visited by wealthy people from all walks of life on Sundays. The monastery, known as St George Koudounas (‘decorated 12 The term Rum (Turkish for roman) refers to Turkey’s Greek Orthodox Christians, or the Christian heir to the Byzantine Empire (the Eastern Roman Empire). They refer to themselves as Romii, which derives from Romaios (‘Romans’ in Greek) 13 Pinkipo is the largest of a string of islands to the south-east of Istanbul. The others are Heybeliada (Chalki), Burgazada (Antigôni) and Kinaliada (Prôti). Couroucli 2012 St George the Anatolian 9 with bells’) because of the little bells that adorn the armour worn by the saint depicted in the church’s main icon, was both a tourist attraction and a place of pilgrimage: At least once a year, every family went together to the mountain to venerate the saint, to sprinkle themselves with holy water and to fill the bottles they kept for difficult moments. Whilst ordinary people went for the patronal feast, more ‘well to do’ families often went on excursions to St George’s and held parties in the open air. On special occasions, tables were set up on the square behind the monastery, and it was the higoumene (who acted as host) who took charge of the proceedings. The society pages of the Greek newspapers of the day bear witness to the fact … there were always crowds at St George’s on Sundays (Millas 1988). When the Christians have left, Prinkipo is still the preserve of Istanbul’s high society. This is where the rich have their villas and their local sports clubs. Here, the modern elite enjoy a way of life that is very different from that of ordinary Stamboulites. Even the pilgrims who visit the island on 23 April are townspeople who were born in Istanbul and they do not share the same way of life as the populations from the countryside that have recently settled there. We are in a city with an old tradition of syncretism where the chapels and holy places of Christians have always been visited by Muslims. For the Rum, or Istanbul’s Greek Orthodox Christians, this ‘sharing’ is part of the normal life of the capital. A woman attending Sunday mass at the Church of the Trinity near Taksim Square in April 2004 confirmed that this is still the case: ‘A lot of them go to St George’s on Prinkipo because the saint gives them everything they ask for: houses, jobs, good health. He gives everything.’ According to her, more and more Muslims were going because the saint had a ‘good reputation’ and because the priests were welcoming. According to the priest who was welcoming the pilgrims to the church when I first visited St George’s sanctuary in 1992: ‘The Ottomans (Оθωµάυοι) come with faith. If I refuse to say prayers for them (for their health, houses and jobs), I would be the one who was committing a sin, a blasphemy […] They are the same prayers that we say for the Romii (Rum), prayers from [the book] that protects against the evil eye [Vaskania]. When they come with faith, we cannot turn them away. They [the pilgrims] have faith.’ The usual visitors to these holy places are Turkish-speaking women of Muslim tradition who belong to the city’s educated middle classes; they do not wear headscarves, or at least not Islamic headscarves. Having been born in Istanbul, they are familiar with the sacred geography the city, in which there are other sanctuaries and miraculous springs. These are either Christian of Muslim holy places, and people visit them in search of a cure for their illnesses or malaise (Yerasimos 1992). The pilgrims share this local feeling of belonging or autochtony with the Rum, who were Couroucli 2012 St George the Anatolian