The “ghost” represents vanished culturally diverse societies and suppressed and lost histories – forces us to come to terms with legacies of colonial violence and slavery

Indeed the fate of minority property has been perhaps the major impediment in the study of intercommunal relations in the late 19th and early 20th Eastern Mediterranean. As Maryline Crivello points out, this has led to the formulation of exclusive national narratives reinforced by insecurities related to the recently established –and therefore still contested- homogeneity of Mediterranean nations and the legitimacy of their states.103 In other words, any explicit reference to the illegitimate ways in which property belonging to other ethnic minorities is perceived by officially-sanctioned histories as a threat to the legitimacy of the nation-state itself. But “[c]ontrary to the ordered memory of the archive,” as Iain Chambers writes, “to the filed away names, events, and explanations, there is the persistent murmur of a world whose insistent ‘noise’ betrays the order that seeks to regulate, and in the end, remove it.”104 More and more researchers have indeed chosen to pay attention to this “murmur,” override the various forms of cultural or sometimes legal censorship to explore the dissonance, throughout the region, between nationalist narratives and unofficial or marginal memories. Recent work on the confiscation of Armenian property by the Young Turks and the Turkish Republic or that of Muslim and later Jewish property in Salonica/Thessaloniki by the Greek state highlight the often denied destructive nature of nation-state formation around the Mediterranean following the gradual collapse of the Ottoman Empire.105 Rebecca Bryant’s research on post-intercommunal conflict Cyprus also reveals the often very material foundations of what she calls the “partition of memory” between Greek and Turkish Cypriots. Property seized, Bryant notes, is never fully appropriated by its new owners; it is a site without a past or rather with a hidden past that resists any attempt at obfuscation. And so, “these remains of history become history’s remainders, pieces in a past that do not fit, and hence reminders of a past that is unfinished and so capable of molding the present in new and unexpected ways.”107 Indeed, by directly confronting the selective amnesias of official histories, such work highlights the political difficulty in writing about intercommunal relations in the modern period. Whether scholarly or meant for a broader readership, these studies also underscore the corrupting effects of denial on tolerance to diversity, supposed to be the bedrock of contemporary democratic polities. Yet this paper suggests that the violence that put an end to erstwhile multicultural settings -and its consolidation through property transfers- is not the only reason behind “partitions of memory” and the production of soliloquous narratives in the Eastern Mediterranean. Indeed a careful reading of our sources seems to suggest that interactions across communal lines were tenuous and formal long before the trauma of separation. To return to our Rodesli testimonies, there seems to be a genuine element of surprise at the fact that certain Greeks applauded the 1944 deportations or, on the contrary, at the readiness of Muslim families to provide shelter to persecuted Jews. Indeed, if the deportations revealed anything about intercommunal relations in Rhodes, it is that the island’s different communities seemed to have known little about one another. To the extent that Rhodes may be considered as an emblematic case in the Eastern Mediterranean, this observation does call for a corrective to that quintessentially Mediterranean trope, “cosmopolitanism,” which is often considered to have characterized pre-national Eastern Mediterranean regions. Of course this notion of a pre-national cosmopolitanism, coterminous with a post-Braudelian understanding of the Mediterranean, has been undergoing sustained critique for the past thirty years.109 Mostly research has been concerned with an uncritical usage of the notion of cosmopolitanism, impervious to the power relations that enabled an economically and politically privileged elite to be “cosmopolitan,” that is, to entertain a semicolonial, cross-cultural sociability at the expense of a local, poorer, and as a consequence culturally much less fluid population.110 What our sources do suggest, is how speculative it is to assign any a priori quality to pre-national cross-cultural interactions. Establishing that different religious groups coexisted for a long period of time in a specific location is not sufficient to allow us to infer anything about the nature of their mutual relations. And so we find ourselves facing the central question running through this paper: how do we use soliloquous sources to imagine past societies defined by their cultural diversity without falling in the trap of invoking “multiculturalism” or “cosmopolitanism”? 111Interestingly, in seeking to grasp an intercommunal past without reverting to idealized references to cosmopolitanism, many students of the Eastern Mediterranean invoke the “ghosts” of past –deported or annihilated- communities in any given setting, from mainland Greece to Anatolia, through Cyprus and Israel and Palestine. This figure of the ghost, which literally pervades the literature on the region, is meant to evoke the violence of the destruction of a given community and the lingering signs of its erstwhile existence in the built environment, oral testimonies and traditions and in documents. The ghost becomes both a metaphor for, and the witness of, vanished culturally diverse societies.112 Overall however, the “ghost” in this literature remains a rhetorical figure rather than a proper analytical concept. A major reference in helping us shoring up the analytical purchase of the figure of the ghost is Jacques Derrida’s Specters of Marx. In this series of lectures aimed to devise a hauntology as an approach and a method able to address what Derrida considers to be some of the innate flaws of ontology and phenomenology, the author writes that all hegemonic discourses –historical and otherwise- necessitate in their formulation the concealment of signs that contradict their unifying principle.118 But because these repressed traces can never be suppressed, they permanently haunt and threaten univocal narratives.119 To break out of this logic leading to ever more coarse devices used against the stubborn presence of ghosts, it is necessary, according to Derrida, to learn to live with them, in their companionship, how to “talk with him, with her, how to let them speak or how to give them back speech,”120 to allow them to “work” and make us work.121 For the purposes of this paper, this means acknowledging that the sheer existence of apparently incommensurable sources dealing with the same events force us to forsake, in our own accounts, any unifying principle –such as characterizes “collective memories”- and accept to “gather under a single roof the apparently disordered plurivocity”122 of different recollections. As a reading and writing technique this would require a bold departure from the time-honoured “synthesis”: the goal would thus no longer be to use one discourse as a corrective to the other with a view to finding a presumably more plausible version of the past. Indeed proceeding in such a way would not resolve the basic problem of competitive memory inherent in the very production of the sources; the brittle consensual memory established in this manner would not resist long before temptations to restore a “hierarchy of suffering” that seems to be congenital in much of the historiography of the Eastern Mediterranean. Instead, the various sources at hand should be treated on equal –and “disjointed”- terms, an approach respectful of their difference.