Interpreting Dragomans: Boundaries and Crossings in the Early Modern Mediterranean

Early modern observers rarely failed to comment on the perceived diversity of peoples, customs, and languages of Mediterranean societies. This diversity they sought to capture in travel narratives, costume albums, missionary and diplomatic reports, bilingual dictionaries, and a range of other genres of the “contact zone.” Modern scholars, too, have celebrated the early modern Mediterranean's ostensibly multiple, diverse, and even “pluralist,” “cosmopolitan,” or “multicultural” nature. At the same time, in part thanks to the reawakened interest in Braudel's seminal work and in part as a much-needed corrective to the politically current but analytically bankrupt paradigm of “clash of civilizations,” recent studies have also emphasized the region's “shared,” “connected,” “mixed,” “fluid,” “syncretic,” or “hybrid” sociocultural practices. Of course, these two analytical emphases are far from mutually exclusive, as recently underscored by Peregrine Horden and Nicholas Purcell's comprehensive, longue durée model of diversity-in-connectivity. Yet, neither Horden and Purcell's structuralist “new thalassology,” nor other studies of the early modern Mediterranean have offered a systematic account of how “diversity” and “connectivity” as both the flow of social practices and the categories for speaking about them have been articulated through specific institutions and genres.

civilizations," recent studies have also emphasized the region's "shared," "connected," "mixed," "fluid," "syncretic," or "hybrid" sociocultural practices. 3 Of course, these two analytical emphases are far from mutually exclusive, as recently underscored by Peregrine Horden and Nicholas Purcell's comprehensive, longue durée model of diversity-in-connectivity. 4 Yet, neither Horden and Purcell's structuralist "new thalassology," nor other studies of the early modern Mediterranean have offered a systematic account of how "diversity" and "connectivity" as both the flow of social practices and the categories for speaking about them have been articulated through specific institutions and genres.
To address such processes of articulation, this study looks at the mutual constitution of "diversity" and "connectivity" in one of the paradigmatic institutions of the early modern Mediterranean "contact zone," namely diplomacy. Specifically, it considers the formation and transformation of the Venetian dragomanate, the corps of diplomatic interpreters employed by the Venetians in Istanbul in the sixteenth and seventeenth centuries. Whether born and raised in the Ottoman capital or merely long-term sojourners, dragomans were ubiquitous in the city's Christian suburbs, and appeared as regular guests both at court and in Ottoman officials' homes. They thus served as principal actors in the production and circulation of news in Istanbul, itself a central node of early modern diplomacy, as historians increasingly recognize. 5 Dragomans' interpretive work was crucial in procuring for foreign diplomats a current store of knowledge about Ottoman politics and society. Moreover, like other inter-imperial intermediaries, dragomans did not simply transmit information across disinterested channels, but rather articulated diplomatic knowledge, shaping many of the discourses about the Ottomans that eventually were inscribed in official Venetian diplomatic dispatches and reports from Istanbul. These reports themselves circulated far and wide. Some of them, although officially classified, are known to have been quickly copied and sent off to Rome, while others were translated and anthologized into "manuals of political theory" for European-wide consumption. 6 Yet, despite the wealth of prosopographic studies on the dragomanate and biographical studies of individual dragomans and their families, historians have paid little attention to dragomans' practices of translation and mediation and how these relate to their unique trajectories. 7 This study begins such an exploration by considering how dragomans positioned themselves as mediators adept at crossing political and ethno-linguistic boundaries. 8 To do so, I first situate the emergence of a highly influential cadre of dragomans at the intersection of Venetian patrician practices of descent and kinship-making on the one hand, and Ottoman mechanisms of recruitment into and apprenticeship within elite households on the other. In particular, I discuss the institution of the household of the bailo, the Venetian permanent representative in Istanbul, where the earliest systematic efforts to train new cadres of dragomans were initiated in the mid-sixteenth century, and where dragomans worked for the next two centuries to coordinate relations between various imperial centers. I then show how dragomans' multiple provenances and modes of recruitment, training, and employment belie neat categorization as either "local" or "foreign," "Ottoman" or "Venetian." Indeed, dragomans typify what I call "trans-imperial subjects," actors who straddled and brokered-and thus helped to shape-political, religious, and linguistic boundaries between the early modern Ottoman and Venetian states. 9 Finally, through a comparison of two translations of a letter by Sultan Murad III to Venetian Doge Pasquale Cicogna in 1594, I suggest how divergent trajectories informed two dragomans' sense of familiarity with both Ottoman and Venetian bureaucratic elites, and ultimately shaped their practices of translation and mediation.

PAT R I M O N I A L H O U S E H O L D S A N D T R A N S -I M P E R I A L C O N T A C T Z O N E S
Historians of the early modern Ottoman state have long noted the important roles played by recruitment and training in large elite households in entwining domestic hierarchies with imperial politics, and in institutionalizing and perpetuating ethnic heterogeneity at the empire's core. 10 These patrimonial households-starting with the imperial palace in Istanbul and extending to the households of military-bureaucratic elites in the provinces-served as training grounds for a large body of young cadets who functioned simultaneously as both domestic and civil servants. Initially, candidates for the imperial household were captured primarily through raids beyond the frontier and from among captives and prisoners of war. From the early fifteenth century, additional recruits were obtained through the formalization of the practice of appraisals of dragomans' functions in the Ottoman administration, see Veinstein 2000;Philliou 2001;Ç içek 2002;and Janos 2006. 8 As Palmira Brummett recently suggested, in order to better comprehend the spatiotemporal categories at work in the early modern Mediterranean, greater attention should be paid to "the not-so-liminal crossers of frontiers, those who move from one identity to another reflecting or embodying the porosity of physical, ethnolinguistic, religious, and political borders" (2007: 45). 9 For a more systematic discussion of the concept of trans-imperial subjects, see Rothman, forthcoming. 10 Fleischer 1986: 255-56, passim;Findley 1980;Goffman and Stroop 2004. devşirme, or child levy. This institution ensured the steady supply of slaves for the imperial household from among the non-Muslim rural population of the provinces, especially the Balkans, where a changing percentage of boys and youth were removed from their parental homes and sent to the imperial center. By the late sixteenth century, however, raids, war booty, and the devşirme ceased to be the exclusive source of recruits into the imperial household. To supplement them, specialized personnel with specific skills or technical know-how were sometimes enlisted from among converts, "foreigners" beyond the frontier, and groups in Ottoman society previously deemed "unfit" for service. 11 Regardless of provenance and method of mobilization, patrimonial households had the capacity to transform their inductees. Upon recruitment, cadets underwent a lengthy and rigorous regimen of what Cornell Fleischer has termed "deracination, education, and Ottomanization," which molded them into loyal subjects suited for lifelong service to the dynast in a range of crucial military and administrative roles. 12 This protracted training could easily last over a decade. Recruits were first assigned to Turkish Muslim families to learn the language and become accustomed to hard labor, were then schooled or apprenticed within the imperial household for several years, and only then, upon graduation, entered a variety of positions in the state's expanding military-bureaucratic apparatus. 13 Whereas cadets' marriage was at first strictly limited, by the late sixteenth century it relaxed to the point that true service dynasties began to emerge, particularly among members of the imperial cavalry and other elite office-holders. Here, for the first time, membership in the Ottoman imperial household became a potentially heritable status, with sons of recruits gaining a sense of privilege by descent. These swelled the ranks and ultimately made the devşirme superfluous, leading to its de facto disappearance in the late seventeenth century. 14 This transformation in household recruitment patterns at the turn of the seventeenth century and its implications for conceptions of subjecthood, loyalty, and bureaucratic professionalization have been well-documented not only in the Ottoman capital, but also in the military-bureaucratic elite households of the Ottoman Balkans, Egypt, North Africa, and the Arab provinces. 15 Far less understood are the roles of recruitment into and training within expansive elite households in contemporary Venetian society. To be sure, the historiographical emphasis in the case of Venice has been rather on the exclusivity and endogamy of the metropolitan patrician and citizen classes. 16 Yet, even in hyper-endogamous Venice, studies of elite households have outlined how extended, bilateral kinship orientation was instrumental in consolidating a patrician grip on political institutions, allowing families to weave dense networks of patronage through both the paternal and the maternal lines. 17 Significantly, so far, the shared patrimonial principles and purposes-though not always actual practices-of Ottoman and Venetian elite households have gone for the most part unnoticed. This understudied confluence is especially intriguing in light of the de rigueur, long sojourns at the Porte of many a prominent member of Venice's political and commercial elites. It is well worth asking, therefore, how certain assumptions about loyalty and subjecthood were engendered by Venetian and Ottoman elite kinship and household structures, and how associated roles were inhabited and manipulated by people who were familiar with-indeed, familiars of-both.
Such familiarity was cultivated most clearly in the household of the Venetian bailo in Istanbul, an institution whose significance goes well beyond Venetian-Ottoman inter-imperial relations. This household served as a model for numerous other diplomatic residences in the Ottoman capital, not least in its functioning as a center for linguistic training. It did so by recruiting adolescent apprentice dragomans across Venetian and Ottoman territories with the explicit purpose of turning them into loyal and useful Venetian subjects. These apprentices' official Venetian title, giovani di lingua (language youth, a calque of the Turkish dil oglan), already suggests the significance attributed to young age as a precondition for the transformation presumed to result from long apprenticeship. 18 The Venetian dragomanate in Istanbul and its offshoots throughout the Venetian colonies and in Venice proper are a prime example of how the Venetian state adapted prototypically Ottoman mechanisms of subjectmaking and integration through a large elite household. By tracing the emergence and transformation of the Venetian dragomanate we can see how Venetian and Ottoman household patterns and affective ties interacted and sometimes converged in the making of trans-imperial professional cadres. We may also thereby gain a vital perspective on the genealogies of practices of mediation, translation, and scholarly knowledge production in the early modern Mediterranean contact zone, as the body of texts about Islam and the 16 On the consequences of elite endogamy for the Venetian patriciate's inability to reproduce itself as a class, see Sperling 1999. On the fraught relations between Venetian colonial nobilities on Crete and their metropolitan kin, see McKee 2000: 61-66, passim;and O'Connell 2004. 17 On the nexus of patrilineal consciousness, bilateral kinship orientation, and patronage strategies centered on the household in early modern Venice, see Grubb 1996;Romano 1996;Brown 2000;Chojnacki 2000;and Raines 2006. On personal and familial ties to the patriciate as essential in consolidating citizens' grip on the lower echelons of the Venetian bureaucracy, see Pullan 1999: 162-63. On endogamy and patronage in the diplomatic corps, see Zanini 2000. On patrimonialism in early modern European familial states in general, see Adams 1994. 18 On the Venetian term as a calque of the Turkish, see Sturdza 1983: 565.
Ottomans produced by Venetian and other Istanbul-based dragomans was later codified in the discipline of Orientalism. 19

R E C R U I T M E N T
The Venetian dragomanate, which eventually became highly endogamous, was forged haphazardly in the course of the sixteenth century from three very different bases of recruitment: the Venetian citizen class, the colonial nobility of the Adriatic and eastern Mediterranean, and the Latin community of Pera, a suburb of Istanbul. Understanding the differing ties that bound these three groups to the Venetian state and its patrician ruling class is essential for any inquiry into their role in mediating Venetian-Ottoman relations. In Venice, citizens by birth (cittadini originarii) formed a de facto second tier of the metropolitan elite. A clearly self-conscious estate, its men were barred from office-holding and voting, but constituted the government's bureaucratic core, sharing to a large extent patrician understandings of the state. As service in Istanbul was generally considered a stepping stone to more prestigious employment in Venice, numerous citizen families, with a long tradition of supplying secretaries to the ducal chancellery, were willing to send their sons into apprenticeship in the bailo's house. 20 A second group of recruits to the dragomanate came from the Venetian colonial elite in the Adriatic and eastern Mediterranean, increasingly undermined by the Ottoman conquest of Venice's Dalmatian hinterland in the early sixteenth century and of Cyprus in 1571. In the aftermath of these conquests, many feudal families from the Balkans and Eastern Mediterranean sought refuge in other Venetian territories or in Venice itself. Placing a son in Venetian diplomatic service in Istanbul reinforced these threatened nobilities' claims to enduring colonial loyalty. 21 It also offered concrete prospects for social and economic mobility by linking young apprentices with powerful patrons among the Venetian political elite, and by opening up distinct commercial opportunities, in both Venice and Istanbul, later in life. 22 Finally, a third and crucial group of recruits to Venice's dragomanate came from the Latin, that is, Roman-Catholic community of Pera, an affluent, predominantly non-Muslim suburb of Istanbul. Members of this Latin community, the Magnifica comunità as they called themselves, traced their roots to 19 This nexus is not fully charted out yet. For some suggestive studies, see De Groot 2001;Testa and Gautier 2003;and Hamilton and Richard 2004. 20 On the bailo's house, see Bertelè 1932;Coco and Manzonetto 1985;and, especially, Dursteler 2006: 27-40. 21 For cases in point, see Rothman 2006: 232-35. 22 According to Andre Pippidi (1980: 135), Serbians and Albanians of the seigniorial class were particularly prone to accept such employment. It should be noted, at the same time, that many families chose to "hedge their bets" by sending some sons to Venice, and others to Istanbul. See also Wright 2006 for a case in point, and Arbel 2000 on the Cypriot nobility and its relationship to Venice.
Genoese and Venetian settlement in Byzantium even prior to the Fourth Crusade, and Pera's permanent Genoese community dates from the 1260s. 23 By the sixteenth century, Pera had become the seat of numerous foreign embassies. At that time, the Venetian bailo began recruiting members of several prominent Latin families as dragomans and dragoman apprentices. These men and youth joined his large household, located in the area, thus giving the bailo direct access to Pera's centers of power, since quite a few dragomans concurrently held an official position in the Magnifica comunità. Between 1580 and 1670, Venetian dragomans or their immediate kin served as priors or sub-priors of the Magnifica comunità thirty-two times, for a total of eighty-eight years. Some also served in various capacities in the numerous local Catholic churches. For instance, Giovanni Antonio Grillo, the Venetian grand dragoman, was appointed procurator of St. Francis in 1626. That same year, two members of the Piron dragomans' family, Matteo and Bartolomeo, became procurators of St. Peter and St. George,respectively. 24 From local recruits' perspective, Venetian employment spelled not only a steady income, but also a source of authority within the shrinking, conflict-riddled Latin community. Such employment also served as a form of legal protection, because local employees of foreign powers could enjoy certain immunities and exemptions from Ottoman law. 25 From the perspective of Venetian officialdom, the recruitment of dragomans from among powerful families in Pera also curbed, at least partially and temporarily, French and Papal inroads into the same community, whose strategic significance was well recognized by the various Catholic powers of the period. 26 In recruiting apprentice dragomans, the Venetians gave clear preference to the sons, sons-in-law, and nephews of acting and former dragomans. 27 Thus, for the youth of Pera, kinship ties offered an accelerated entry into dragoman apprenticeships in the bailo's house. When a local dragoman or apprentice passed away, the bailo was immediately petitioned to take another one in. For example, when Ippolito Parada died of the plague only a few months after starting an apprenticeship in 1637, his family promptly asked that his fifteen-year-old younger brother Michele replace him. Bailo Alvise Contarini, who forwarded the request to the Senate, endorsed it and suggested, "Your Serenity could not do a greater work of charity than this." 28 On another occasion, Dragoman Giovanni Battista Navon, whose father Pasquale and brother Tommaso had already served in that office, petitioned to have his son Alessandro admitted into the bailo's service as an apprentice dragoman. Navon did not fail to mention his father-in-law, Marcantonio Borisi, who had been executed by the Ottomans while in Venetian service, and the stipends disbursed to Navon's now-deceased wife and her sisters in recognition of Borisi's merits. Citing the long service of both families, the bailo recommended admitting Alessandro into service so that "excited by this stimulus of public kindness he will have the diligence to occupy himself and be instructed with the faith and devotion typical of his house." 29 This mechanism of recruitment proved very effective, and within a couple of generations the Venetian dragomanate was populated overwhelmingly by members of the Latin community of Pera. Throughout the late sixteenth and seventeenth centuries several of the most distinguished Latin families had at least one son employed as a Venetian apprentice dragoman at almost any given moment. This system of guaranteed employment to certain families also reproduced on a small scale the Venetian strategy of granting citizens by birth a monopoly over specific positions in the state bureaucracy, thus forging alliances and securing goodwill and collaboration. 30 It is here that we begin to see how the institution of the bailo's house welded classical Venetian patterns of endogamy and social reproduction with patterns of exogamous recruitment and training typically associated with the Ottoman imperial household. On the one hand, admission into Venetian service in Istanbul-not unlike other positions in Venice's expansive state apparatus-relied heavily on kinship and descent. On the other, it also entailed the restructuring of these same familial ties. For upon entry into service, young apprentice dragomans were removed from their homes and from the domestic care of their Greekspeaking mothers, and placed into the all-male space of the bailo's house. There, for the next seven years or more, they were entrusted into the paternal care of their dragoman fathers, uncles, older brothers, and, of course, the bailo himself. The latter, although customarily ignorant of Turkish, personally supervised his dragoman apprentices' linguistic progress and reported on it in his weekly dispatches to the Venetian Senate and in his comprehensive relazione upon return from office. 31 The discussion so far has suggested a simple, tripartite division of the dragomanate into Venetian citizens, colonial nobles, and Latins from Pera. This division is in line with the logic of the early modern Venetian state itself, which carefully distinguished between citizens, subjects, and nonsubjects, each possessing a supposedly inherent and fixed degree of loyalty to the Venetian state, and a set measure of willingness to put the latter's interests before one's own. Indeed, the recruitment of Venetian citizens and colonial subjects as apprentice dragomans was intended precisely to counteract the proverbial disloyalty of Pera-born dragomans who were, after all, Ottoman subjects. 32 Yet this very division of the dragomanate into Venetian citizens, colonial nobles, and Latins from Pera was much eroded by the forms of sociability engendered by the bailo's house and the wider city. So much so that by the seventeenth century the boundaries between the three groups became increasingly difficult to maintain. Venetian subjects and citizens sent to be trained in Istanbul could "go native" in ways unforeseen and unappreciated by their employers. Some embraced Islam, quit the service, and sought employment elsewhere in the Ottoman capital. For example, within three years, 1627-1629, Venice lost three of its Venetian-born apprentice dragomans. Camillo Garzoni was convicted of an unnamed crime (possibly leaving Istanbul without the bailo's permission), exiled to Zadar for three years, and barred from public office for life. Another apprentice, Fontana, converted to Islam. A third, Antonio Torre, also became Muslim, leaving behind a long list of creditors. 33 Some apprentice dragomans took local concubines or lovers in clear transgression of expected affective boundaries, while others still were actually absorbed into the Latin community of Pera through marriage alliances, acquiring in the process in-laws from among more senior dragomans.
The high degree of intermarriage among the three groups led by the midseventeenth century to the establishment of veritable dragoman dynasties that were more or less permanently settled in Istanbul despite their diverse roots across Ottoman and Venetian territories. government on the occasion of his matrimony, Vecchia suggested that the sole purpose of the union was to let him "stay in the country and devote myself until the last breath to the service of Your Serenity," alluding to the Ottoman view of a foreign resident's marriage to an Ottoman subject as a clear indication of intent to naturalize. 34 Five years later, following Grillo's execution by Ottoman authorities, Vecchia, now living in his late father-in-law's house, claimed that Grillo "with all tenderness sought to love me with affection exceeding that of a father." 35 In addressing the Venetian authorities, Vecchia therefore downplayed his affective ties to his local spouse, ties that could be seen as conflicting with his undivided loyalty to his Venetian employer and sovereign. Rather, he professed deep and reciprocated affection for his father-in-law. He thus invoked the notion of cohesiveness among dragomans themselves while bracketing the important role played by wives, daughters, and sisters in mediating these relationships. Such cohesion was both capitalized upon by dragomans and seen by their patrons as vital to the success of their enterprise in the Ottoman capital. Yet it also facilitated the quick absorption of dragomans of Venetian citizen and colonial background into the Latin community. 36 As importantly, dragomans' households were crucial nodes in the development of trans-imperial kinship networks that furthered the production and circulation of timely political and linguistic knowledge. For while some Venetian citizens and colonial subjects were working for the bailo and celebrating their loyalty to Venice in one petition after another, their immediate kin-brothers, sons, brothers-in-law, and nephews-were otherwise employed in Istanbul. Some became dragomans for other foreign embassies, others married French, Dutch, or Danish merchants and physicians active in the Ottoman capital, and yet others married into the nobility of the southern and central European Ottoman-Habsburg borderlands.
This dual movement of simultaneously "going local" and forging trans-imperial kinship networks is best illustrated by the case of the Borisi, Brutti, and Tarsia families. These three closely inter-related dragoman dynasties traced their roots to then-current or former Venetian colonial territories, including Albania, Montenegro, and Istria. Yet by the seventeenth century certain branches of all three families were firmly settled in Istanbul. In the course of the century between 1570 and 1670, they produced five generations of dragomans in Venetian service in the Ottoman capital, totaling at least thirteen men. At the same time, these dragomans' immediate relatives were placed all across the Venetian and Ottoman empires, and their daughters and sisters married into at least five other dragoman families in Istanbul, as well as into Venetian, Habsburg, Danish, Polish, and Moldavian aristocracies and merchant elites. By the mid-seventeenth century the Borisi-Brutti-Tarsia extended family had spanned three empires and over half a dozen locales. 37 These complex genealogies underscore the importance of marriage alliances for transforming Istanbul's dragomans into a unified and socially mobile group, regardless of their diverse origins and employers. Moreover, the forging of kinship networks that crisscrossed political, spatial, ethno-linguistic, and estate divisions furthered not only dragomans' internal cohesion and selfconsciousness as a professional cadre, but also their alignment with a truly trans-imperial-rather than specifically Venetian-aristocratic milieu. The resultant uncertain political loyalties of this group were, not surprisingly, at the heart of Venetian concerns over the services rendered by its dragomans, who were all too localized but possibly not sufficiently "Venetianized." From the Ottoman state's point of view, too, dragomans were both "foreign" by virtue of serving foreign embassies, and "local" by virtue of their numerous relations in the Ottoman capital and provinces. 38 And, as mentioned, dragomans themselves often conflated this distinction further by sending different sons to work for the Habsburgs as well as the Venetians, French, English, and Dutch. For example, in the late sixteenth and early seventeenth centuries, members of the Olivieri family worked in both the French and Venetian embassies. A certain Nicorosio Grillo, relative of the Venetian Grand Dragoman Giovanni Antonio, was employed by the Dutch ambassador Haga in 1616. Other local families, including the Navon, Piron, and Parada, similarly had some sons working for the Venetians and others for the French, the British, the Dutch, and the Habsburg embassies. 39 Officially, the Venetians disapproved of having their dragomans' immediate relatives employed by other powers for fear of espionage. However, extended kin and friendship networks also offered dragomans vital access to local and inter-imperial information, and thus often proved quite beneficial for their employers as well. The dragoman Panagiotis Nikousios, for example, who served the embassy of the Holy 37 For further details and a genealogical tree, see Rothman 2006: 255 -59, 464. 38 On the fraught attempts by the Ottoman state to classify dragomans as subject non-Muslims, see Van den Boogert 1997. 39 Hitzel 1995: 53;De Groot 1978: 192. Roman Empire in Istanbul, and later became Ottoman grand dragoman, maintained a decades-long friendship with Venetian dragomans such as Ambrosio Grillo, as well as with their employers the baili, providing much needed information on political maneuvers in the other embassies and in the chambers of the grand vizier himself. 40 The political usefulness of maintaining members of local elites on one's payroll is well attested in the bailo's long-term employment of several entrylevel apprentice dragomans who displayed only minimal linguistic skills but who were well connected in Istanbul. In an extensive report on dragomans' performance, baili Pietro Foscarini and Alvise Contarini cautioned in 1641 against discharging any dragomans or apprentices from service, regardless of poor performance, since they would immediately be recruited by the French and English. 41 In another case, a dispatch by Venetian Secretary (and de facto bailo) Giovanni Battista Ballarino in 1655 confirmed that the aging Giovanni Piron, who had been employed as an apprentice dragoman for twenty years, had finally mastered some languages, and endorsed his petition to be promoted to the level of dragoman. Apparently he was not promoted, and in 1664 Piron, aged seventy-five, still appeared on the payroll as an apprentice. According to Ballarino, for the previous six years Piron had visited the bailo's house only at Easter and Christmas. However, given his brother Antonio's position as dragoman for the English, and Giovanni's close friendship with Giorgio Draperis, the English grand dragoman, Ballarino considered his absence from duty an advantage, since it kept the disgruntled apprentice from disclosing secret information to Venice's commercial rivals. 42
Ottoman frontier. Several Venetian dragomans were sent on diplomatic missions to far-flung provinces of the Ottoman and Safavid empires. At least four of them left detailed reports of their embassies. From 1539 -1542 Michiel Membré, the Cypriot-born future Venetian public dragoman, was sent to the Safavid Shah Tahmāsp; another dragoman, Venetian Secretary Vincenzo degli Alessandri, was dispatched to the same Safavid court in 1570. Pera-born Giovanni Battista Salvago, who in 1624 was appointed Venetian ambassador to the Regencies of North Africa, was followed in 1633 by a fellow Ottoman compatriot and member of a dragomans' dynasty, Ippolito Parada, who traveled to Algiers by order of Bailo Cappello. 43 Other dragomans also traveled extensively, since assistance in Ottoman-Venetian border negotiations and periodic assignments to various posts throughout Venice's Mediterranean and Adriatic colonies were customary before assuming more prestigious positions in either Venice or Istanbul. For example, the Pera-born dragoman's son Stefano di Gioveni, who had served the Venetian consul in Alexandria since 1581, was recalled to Istanbul five years later to replace the deceased dragoman Ambrosio Grillo. 44 Once dragomans were back in the Ottoman capital, unlike today's interpreters, face-to-face simultaneous interpretation between Ottoman and Venetian officials occupied relatively little of their time. More often they translated written documents and, especially, engaged in independent visits to the divan and to the residences of Ottoman officials, where dragomans themselves frequently acted as sole Venetian representatives. Thus, while some Venetian diplomats acknowledged it more openly than others, it is hard to overstate the degree to which their ability to produce and circulate timely news from Istanbul relied on dragomans' local ties and interpretive work. It was, for example, with some degree of pride that Bailo Almorò Nani wrote to the Senate in 1616 of his secret contacts with a Hungarian dragoman, who debriefed him about events in the Habsburg court, as well as in Dalmatia, a region of key importance to the Venetians. Another dispatch by Nani a few months later reported on the missions of his two dragomans, Giuliano Salvago and Barnabà Brutti, to the Ottoman governor of Bosnia and to Ibrahim Aga, the military governor of Buda, respectively. The dragomans' letters, summarized in Nani's own dispatch to the Senate, included detailed 43 Parada was to recover the possessions of ambassador Cornaro in Spain. He became an apprentice dragoman four years later, and died of the plague a few months thereafter: Cinque Savii, Risposte, b. 149, c. 54r (10 May 1633); Senato, Dispacci Costantinopoli, b. 118, c. 611r (17 Oct. 1637 diplomatic, military, commercial, and ethnographic information concerning various Ottoman provinces. 45 Dependency on knowledge produced by dragomans was redoubled by the baili's short term in office of typically only two years, and their ignorance of Turkish. This dependency was often lamented. In his relazione to the Senate of 1576, Bailo Antonio Tiepolo complained at some length about dragomans' mediation, and his inability to establish direct lines of communication with Ottoman ministers, From which follows that because of the great difficulty of interpretation . . . the bailo can never do anything by himself, since he cannot express his own reasons as effectively as necessary. Therefore, in the absence of this efficacy of words, and in the absence also of the virtue of the bailo's skill in reasoning, from which the Pasha would understand proper respect rather than cowardice or fear, the dragoman, who is often impeded by the difficulty of interpreting, and even more by failing to apprehend not only the issues, but also the bailo's mode of impressing these issues, weakens the arguments and exhibits that timidity which is never the bailo's share; for which reason if [the dragoman] is not aided by the bailo in what to say, and by a face full of confidence, and by a steady voice, the Pasha might dare to refuse or make difficult that which would have been most simple in itself. This disadvantage of the bailo, or rather of Your Serenity, is augmented when negotiating in the divan, where it is not customary for the bailo to go; because the dragoman, while Christian, because he is nonetheless a Turkish [i.e., Ottoman] subject, is fearful by his nature, and even more so for having neither the talent nor the experience to negotiate as would be needed in matters of any import. 46 Early modern diplomacy was deeply rooted in humanist notions of eloquence. This helps explain why the bailo's forced reliance on the linguistic mediation of dragomans in his communications with Ottoman officialdom came to be seen as such an insurmountable problem. 47 The challenge was certainly compounded by the perceived gap between the prototypical speech styles of the bailo and the dragoman, which are here mapped very clearly onto their distinct "nature," that is, their status, personhood, and capacity for confident self-presentation.
Yet, if Tiepolo contrasted the bailo's "proper respect" with the dragoman's "cowardice" and "fear," there is no doubt that dragomans' steadfast position as intermediaries was due in no small measure to their ability to inhabit a deferential role vis-à-vis both Ottoman and Venetian elite interlocutors. This ability was cultivated through their lifelong service in Istanbul and extensive contacts in the city. Such "localization" also made dragomans much more familiar with Ottoman diplomatic protocol and with court affairs than their Venetian employers could ever be.
Ultimately, it was precisely dragomans' sense of appropriate selfpresentation and familiarity with and claim to expertise on customary diplomatic practice-not just their knowledge of Turkish or Ottoman juridical status-which lent authority to their pronouncements on local "custom." Bailo Andrea Badoer's report of an encounter between dragoman Mateca Salvago and an Ottoman provincial governor in 1573 illuminates this power of dragomans to speak authoritatively in the name of tradition. Upon arrival at the governor's residence, Salvago was asked whether he had brought the governor his gift. Salvago responded in the positive, and added that, as usual, the gift consisted of silken cloth and other fine textiles. The governor exclaimed: "But where are my one thousand sequins?" Salvago, according to Badoer's report, asserted that monetary gifts were not customary. Moreover, when the governor protested that textiles were not customary gifts either, and demanded to see the bailo himself, Salvago retorted that this would be "superfluous," since even the bailo could not give him money as a gift, adding that the governor "could not say what was not true." 48 If some baili complained that Ottoman subjects made timid and complacent dragomans, Salvago's dealing with the governor implies quite the opposite.
Evidently, their deep familiarity with customary diplomatic practice allowed dragomans to effectively represent Venice to Ottoman officials, and to continue to monopolize this position against possible competitors or alternative channels of communication. 49 Some Venetian diplomats certainly recognized this. In 1553, Bailo Bernardin Navagero praised Gianesin Salvago for his long and loyal service, and the respect he had earned in the divan. Not only did Salvago "understand very well the humors of that nation [i.e., the Ottomans]," but "he is most obliged to the Pashas, and especially to Rüstem, with whom he has become very close, and [shows] such familiarity that he speaks without respectful address, and laughs with him." 50

T R A N S L AT I O N A N D M E D I AT I O N
Translation studies scholars have long acknowledged the interpretive act inherent in any translation. Premised as it is on culturally-specific assumptions about translational faithfulness and the very translatability of foreign concepts and terms, translation is always the product of specific power relations, and as such the site of struggle over meaning. 51 It is thus well worth asking how their 48 Capi del Consiglio dei Dieci, Lettere di Ambasciatori, Costantinopoli, b. 4, c. 102r (20 July 1573). On garments as customary gifts in foreign diplomats' visits to Ottoman officials, see, for example, Inquisitori di Stato, b. 418, unpaginated (10 Jan. 1663 m.v.). On the circulation of Italian silken cloth among Ottoman elite households, see Mackie 2001. The sixteenth-century Ottoman sequin coin was roughly equivalent to the Venetian gold ducat, with which it agreed in weight (3.5 grams). 49 Unlike most other foreign ambassadors to the Porte, Venetian baili only rarely relied on Jewish, Armenian, or Greek dragomans. 50 Albèri 1839: 104. Rüstem Pasha was a grand vizier and son-in-law of Sultan Süleyman the Lawgiver (1520-1566). 51 Robinson 1997;Pym 1998;Tymoczko and Gentzler 2002. translational practices reflect Venetian dragomans' diverse backgrounds and training, and the intersecting logics of endogamous descent on one hand and transformative apprenticeship in the bailo's house on the other. To explore these issues, I compare two translations of a letter sent by Sultan Murad III to Doge Pasquale Cicogna in 1594. The letter concerned a raid on a Venetian galley by North African corsairs in Ottoman territorial waters in the Adriatic earlier that year; it was sent in response to formal Venetian protests. 52 Of the two translations of the letter compared here, one was prepared in Istanbul by Venetian-born dragoman Girolamo Alberti, and enclosed with a dispatch sent to Venice by Bailo Marco Venier in early June. A second translation was produced in Venice about a month later by Giacomo de Nores, who was apparently unaware of Alberti's text. 53 Our knowledge of dragomans' translation practices is still incipient, and therefore this comparison cannot be conclusive. However, sufficient differences are discernible, and enough is known about the backgrounds of the two translators, to suggest that divergent notions of loyalty and the translator's position were at play in formulating the two translations. 54 The two translators differed markedly in their life trajectories and in their connections to Venetian elite milieus. Girolamo Alberti, a scion of a wellestablished family of Venetian citizens by birth, was born circa 1561. He entered the School of St. Mark as a boy, following in the footsteps of his grandfather, uncle, father, and brother, who had all served as secretaries in the Venetian chancellery. 55 In 1582, his father, Secretary Gasparo Alberti, requested to have his firstborn, then twenty-one, sent to Istanbul as an apprentice dragoman. 56 The Senate approved, and Girolamo was to stay in the bailo's house for seventeen years, in the course of which he sometimes served as the sole 52 On corsairing in the Adriatic and eastern Mediterranean and its diplomatic and commercial implications for Venetian-Ottoman relations, see Brummett 1994: 89-121. 53 Both translations, along with the original Ottoman text, are in Documenti Turchi, b. 9, fasc. 1057-59. My deepest gratitude to Vera Costantini for taking photos of the Ottoman letter with the kind assistance of ASV State Archivist Michela dal Borgo. For more information on the archival series of Documenti Turchi in general, and on the specific events that led to the diplomatic exchange between the Sultan and the Doge, see Pedani Fabris 1994. 54 In what follows I was greatly assisted by two separate transcriptions and translations of the original Ottoman document prepared by Tijana Krstić and by Gulay Yarikkaya, to whom I am immensely grateful. I remain solely responsible for the argument here. 55 Established in 1446, the School of St. Mark admitted annually twelve students in their early teens "so that they might go to school to learn 'grammar, rhetoric and other subjects useful for the Chancellery and how to write well.' " By the mid-sixteenth century, the School's state-funded lectureships in poetry, oratory, history, grammar, rhetoric, and especially Greek and Latin philology helped seal its humanist reputation. Ross 1976: 526, 532 -35. See Grendler 1989: 61-70 on public schooling in sixteenth-century Venice more generally.
translator of Ottoman administrative and diplomatic texts into Italian. 57 Already in 1589 Bailo Giovanni Moro sang Alberti's praise, noting, "Not only does he understand well that which is said in Turkish, and translate into Italian the great number of writings which I continually send to Your Serenity, and which I require daily, but being the only one in this position, he works at almost all hours, and his ready efforts give me full satisfaction." 58 Alberti's command of Venetian chancellery practice and fluency in reading and translating official Ottoman documents were further recognized in 1600, when his request to return to Venice to attend to family matters was approved. 59 Alberti, then, was a poster child of dutiful civil service by a Venetian citizen. By 1594, moreover, he had lived and worked in the bailo's house for twelve years, and had had ample opportunity to master the intricacies of Venetian-Ottoman diplomatic translation practice.
Such a trajectory stands in sharp contrast to that of Giacomo de Nores. Born around 1569 in Nicosia, the capital of then-Venetian Cyprus, de Nores was the descendant of two of the island's oldest and most distinguished noble families, the de Nores and the Podocataro, whose roots on the island extended back to the Crusades. 60 Shortly after his birth, in summer 1570, his eponymous paternal grandfather, the Count of Tripoli, was killed while defending Nicosia from the besieging Ottomans. The following year Giacomo's maternal grandfather Livio Podocataro, the viscount of Nicosia, lost all his possessions with the Ottoman conquest of the remainder of the island. Consequently, several dozen members of both the de Nores and the Podocataro families, including toddler Giacomo himself, were taken captive and sent to Istanbul, Chios, Rhodes, Algiers, and elsewhere in the Ottoman Empire. Most of them were eventually ransomed and departed for Venice, Spain, Rome, or other Christian territories. But a few, including one of Giacomo's aunts, converted to Islam and stayed in Ottoman territory. Her two daughters later became sultanas to Mehmet III. 61 Giacomo himself spent his childhood and youth as a slave in the household of Ottoman bombardier Turan Bali of Scutari (Ü sküdar, across from Istanbul on the Asian shores of the Bosphorus). In 1581 he traveled with his master to the 57 Senato, Dispacci Costantinopoli, b. 28, cc. 481r-482r (11 Feb. 1588 m.v.). For examples of Alberti's translations of Ottoman texts in the course of his long career in Istanbul, see Senato, Dispacci Costantinopoli, b. 28, cc. 397r, 445r-v (5 Jan. 1588 m.v.);b. 50, cc. 20r, 23r-25r (4 Sept. 1599). 58 Senato, Dispacci Costantinopoli, b. 28, c. 481r (11 Feb. 1588 m.v.). See also Capi del Consiglio dei Dieci, Lettere di Ambasciatori, Costantinopoli, b. 6, c. 98r (22 Aug. 1588). 59 Senato, Dispacci Costantinopoli, b. 50, c. 265r (10 Jan. 1599 m.v.). A few years later, Alberti became secretary to the Venetian provveditor general in Crete. For additional details of his long career in Istanbul and Crete, see Capi del Consiglio dei Dieci, Lettere di Ambasciatori, Costantinopoli, b. 6, cc. 127r-v (17 Nov. 1591); Bailo a Costantinopoli, b. 275, fasc. 1, cc. 41r-v, 147r-v (23 Aug. 1605, and25 Jan. 1607 m.v.). 60 The de Nores were among the islands' Frankish feudatories, while the Podocataros were Greek-Cypriot Catholics. On the Cypriot nobility under Venetian rule, see Arbel 1995. 61 Rudt de Collenberg 1982 Safavid frontier, where apparently he learned some Persian. Six years later he was ransomed by a Cypriot merchant, Dimitri Gonneme, who paid Turan Bali 260 ducats on behalf of de Nores' mother, Maria Podocataro. 62 At the time of his manumission and arrival in Venice in 1587 de Nores was thus a youth of seventeen or eighteen, with no experience in Venetian service, and with limited, if any, command of Italian. 63 Yet starting in 1589, he was employed as a public dragoman-an official interpreter for the Venetian Board of Trade. This position entailed close interactions with, and oral interpretation for sojourning Ottoman merchants and their Venetian brokers. His written translations in that period would have consisted primarily of notarial and commercial rather than diplomatic records. Indeed, what little fluency he was to achieve in the conventions of diplomatic translation, or, for that matter, in any official register of Venetian or Italian written language, was apparently acquired on the job, in Venice. The two dragomans' divergent trajectories are reflected in their translations of the sultan's letter. While the overall structure and content of the two texts are similar, significant variations in lexicon and person-marking suggest the two dragomans' differing understandings of what constitutes faithful translation and what typifies and thus defines the difference between Ottoman and Venetian authority, agency, and voice. Furthermore, as the analysis below shows, Alberti's translation follows what were probably Venetian diplomatic conventions for rendering Ottoman official terminology in Italian, conventions that had been developed in the bailo's house over decades. Systematic training as a dragoman apprentice would have provided Alberti with a clear set of procedures for translating Ottoman diplomatic vocabulary. He evidently attempted to voice the source as transparently as possible, fully assimilating the sultan's perspective to his own.
On the other hand, de Nores' more piecemeal training "on the job" in Venice would have exposed him far less to the intricacies of Venetian-Ottoman diplomacy. His bread-and-butter employment was not in translating diplomatic correspondence but rather in negotiating commercial disputes among merchants and brokers. Not surprisingly, de Nores used what were probably less conventional solutions than Alberti. More importantly, he was not as consistent at maintaining the sultan's perspective, and used several devices to signal his role as mediator of knowledge. He thereby ended up distinguishing his own perspective from that of the sultan.
In general, throughout his translation de Nores conveys his familiarity with Ottoman political structures, while also showing great sensitivity to the 62 Notarile, Atti, b. 32, cc. 41r-42v (17 Feb. 1591 m.v.). The deed for the ransom, signed by Podocataro and Gonneme in 1581, is discussed in Corazzol (1994: 776), although the author does not identify de Nores as a future dragoman. 63 I thank Maria-Pia Pedani for emphasizing these issues in a personal communication.
Venetian position. Given his many years in Ottoman service and his familial connections outside Venice, he no doubt was keen to diffuse any concerns about his loyalty. Certainly this was his aim in a petition for promotion, which he submitted to the Venetian government only a few months after producing this translation. The petition dwelled at length on his aristocratic forebears and the blood they had spilled in defense of the lost colony of Cyprus. By emphasizing his noble status and distinguished ancestry, and thus reaffirming his ties to Venice, he also reminded his patrician interlocutors of their commitment to his well being as a dispossessed colonial subject. At the same time, de Nores' petition did not shy away from capitalizing on his Ottoman sojourn. References to his long Ottoman captivity might have not only induced sympathy for his plight but also lent credibility to his claim to deep knowledge of Ottoman language and society. After discussing his personal merit, the petition reverts to the first person plural to juxtapose "our" customs with "theirs," thus emphasizing the petitioner's role as an intermediary, "since it is no less useful for that task [of interpreter] to have experience of the habits of the Turks, their inclinations and their manners of negotiation, which are very different from ours, Your Serenity can easily be convinced, that being, I might say, born among these people, and to my bad fortune raised and educated [there], having been involved in their affairs for many years, and traveled in many and diverse provinces and lands here and there. . . ." 64 By using the inclusive first person "our" while narrating his tale of a youth spent in enemy lands, de Nores emphasizes his own distance from the Ottomans, who are treated in the third person. By positioning himself squarely within a Venetian moral community, his long sojourn in Ottoman territory thus becomes an asset rather than a liability; it foregrounds rather than undermines Ottoman alterity. His disenfranchisement by the Ottomans comes in this way to strengthen his claim to special sensibilities and helps underscore his antipathy to his former captors.
A similar perspective on the Ottomans-one that suggests deep familiarity while at the same time projecting a distant, metropolitan Venetian point of view-is also evinced in de Nores' translation of the sultan's letter, setting it apart from Alberti's translation. In fact, de Nores' version more explicitly positions the translator as a "cultural broker," whose mediation is required to make the source legible to a Venetian audience. Several devices help to achieve such an effect: calquing and glossing Ottoman terms and concepts, using colloquial Venetian dialect as opposed to the Tuscan standard common in Venetian chancellery writings, and a switch in person-marking at crucial points to separate the sultan's perspective from that of the translator. Part of the difference between the two translations is no doubt due to Alberti's more systematic schooling 64 Senato Mar, filza 128, unpaginated (13 Dec. 1594), my emphasis. in institutions that integrated prevailing Humanist ideas about rhetoric and translation, such as the emphasis on literalism and the translator's "invisibility," including the imperative to avoid interpretive intervention as much as possible. 65 On a secondary level, the very fact that de Nores provides glosses and repeatedly interjects himself as a kind of "cultural broker" into the text marks him as an outsider to the norms and expectations of contemporary diplomatic translation, as practiced in the bailo's house and in the Venetian chancellery. Whether this is by accident or design, it is not possible to establish at this juncture, but the differences would have been clear to any Venetian administrator adept at reading diplomatic dispatches.
Several elements contribute to the difference in perspective of the two translations. First, de Nores includes framing devices in the first two sections of the letter, which are absent from Alberti's translation. These explicitly introduce the source text: "per una supplica presentata hora all'alta mia sedia" (through a petition just presented to my elevated seat); "suggiungendo appresso in essa supplica, che" (it is further added in that petition that). Such frames accentuate that the letter came from "elsewhere," further distance de Nores from the sultan's perspective, and constitute him as a channel rather than a source.
Second, on several occasions de Nores uses terminology that simply calques the phrasing of the original Ottoman text, while Alberti uses what must be taken as idiomatic, and more conventional, phraseology. De Nores translates the Ottoman appellation of the sultan's abode as "la felice Porta" (the felicitous Porte), and "la felice mia ressidenza" (my felicitous residence), following almost word for word the original terms used in the sultan's letter to the doge, asitane-i sa'adete (threshold of felicity or felicitous threshold) and destgahımız (our chief seat). Alberti prefers more classicizing honorifics, and perhaps a more formal Italian register, such as "l'Ecc[els]a Porta," (the sublime Porte), and "la mia Imperial et Cesarea Maestà" (my imperial and caesarean majesty). 66 Similar differences between calquing and idiomatic translation are visible in the two translators' renderings of Ottoman officials' titles, where the translated forms must commensurate administrative hierarchies. In translating Sinan Pasha's title of kapudanpaşa, the Ottoman lord admiral, both de Nores and Alberti use a term current in the Venetian navy: "Cap[itan]o del Mare" (Alberti, using the Tuscan form) or "Capitano da Mare" (de Nores, using a Venetian dialect form). Both translators thus point to the equivalence of Sinan Pasha's title with the Venetian office of lord admiral. But whereas Alberti later glosses Sinan's title simply as "consigliero" (councilor), de Nores also includes Sinan's Ottoman title of "vizir" (vizier, minister), and his specific jurisdiction in the matter of corsairing as "custode d'Alger," the custodian of Algiers (again, following the original letter's Cezâyir emînleri ile kapudânımız olan . . . vezîrimiz Sinân paşa). De Nores thus signals both his awareness of Sinan's elevated position, as second only to the grand vizier in the Ottoman hierarchy, as well as his understanding of Ottoman provincial government structure more generally. To describe the lesser Ottoman officials addressed by the sultan, Alberti uses the form "Sig[no]ri del Mare," (lords of the sea), following the Venetian construction of official titles with the preposition, "del" in the Tuscan form. De Nores adheres to a similar structure, "sanzachi dà mare." His version, however, uses "sanzachi," a Venetianized plural form of the Ottoman sancak (province, often used in Venetian sources to refer to the person of the provincial governor, or sancakbegi), plus the Venetian dialect form of the preposition, "da." These various examples all foreground de Nores' understanding of the act of translation as requiring significant cultural glossing.
Indeed, de Nores is more likely than Alberti to provide his readers with glosses for certain aspects of Ottoman "custom," styling himself as an authority on things Ottoman addressing what he took as an uninitiated metropolitan Venetian reader. For instance, he supplements the original letter's lunar Hijri date with its Gregorian equivalent ("14 della luna di saban, ciò è alli 4 di Maggio"), whereas Alberti leaves the date un-glossed. 67 Alberti similarly uses the original Ottoman term 'arz-a recognizable genre of Ottoman diplomatics-to refer to the official report about the corsairs' attack sent by the district governor, Piri Beg; de Nores transforms it into "notitia et aviso" (notice), a calque of the original 'arz u takrîr.
On other occasions, de Nores betrayed his imperfect understanding of Ottoman diplomatic and political concepts. He calls the 'ahdnâme (privileges) granted by the Ottoman sultan to the Venetians, which were ostensibly violated by the corsairs' attack, "conventione della pace, et promessa, che è fra ambi le parti" (a covenant of peace, and agreement between the two parties). Such a notion of bilateralism was quite foreign to contemporary Ottoman diplomacy. Indeed, the original letter refers to the corsairs' attack as a violation of an 'ahdnâme-i hümâyûn, that is, an imperial letter of oath, and as "contrary to the peace and amity" (sulh u salâh). 68 This much more unilateral Ottoman understanding of the 'ahdnâme is well rendered by Alberti, who calls the document "giurati Imp[eria]li Capitoli" (sworn Imperial Articles). Similarly, the 67 The use of explanatory glosses with textual markers such as "that is to say" in order to compensate for presumed gaps in readers' knowledge of the context described was a common strategy among Renaissance vernacular translators of Latin antiquity. See Denton 1998: 70- sultan's domains, which the original letter calls memâlik-i mahrûse, are rendered by de Nores as "custoditi nostri paesi" (our well-protected lands), and by Alberti as "mio Custodito dominio" (my well-protected domain). Although de Nores is both grammatically and lexically closer to the original, it is Alberti's use of "dominio" as opposed to "paesi" which suggests greater familiarity with the legal concepts underwriting Venetian-Ottoman diplomatic relations. De Nores' lexical choices also reveal, whether he realized it or not, traces of an Ottoman perspective. For instance, he calls the three corsair vessels that attacked the Venetian ship "Muslim galleys" ("Galere Mussulmane"), following the original designation as müslümân kadırgası, as opposed to Alberti's "Galee turchesche." To be sure, late-sixteenth-century Venetian readers would have taken "turchesche" to mean "Muslim" as much as "Turkish," making Alberti's translation technically correct, but also complicit in a Venetian (and more broadly European) perspective which conflated Turkish ethnicity, Muslim religion, and Ottoman juridical status. 69 By using the less conventional "Mussulmane" rather than the much more common but ambiguous "turchesche," de Nores avoids this conflation, and instead reproduces the original letter's implicit assumption that the corsairs were not necessarily ethnically Turkish, and that the Porte exerted only limited control over the North African corsairs. 70 On the whole, though, de Nores' translation suggests an effort to extricate the translator from any complicity in the sultan's perspective and to position himself in a supposedly more "neutral" intermediary space. In translating beg-the title that the sultan's letter uses for the Venetian administrator assaulted by the corsairs-Alberti renders it as "Bei," thus upholding his supposed "invisibility" as a translator by sticking to probable convention. De Nores, on the other hand, "re-translates" it back to its presumed Venetian form, "Rettore." He thus avoids using a Turkism as a title for a Venetian official, even though, as noted above, he does not mind using Turkisms when translating the titles of Ottoman officials. De Nores thus marks the boundary between Venetian and Ottoman domains with his lexical choices.
A similar unease about voicing the sultan's perspective is betrayed by several shifts in person-marking toward the end of de Nores' translation. As mentioned, the first two sections of the sultan's letter provide an account, first of the capture of the Venetian galley, and second of the actions already taken by the Porte to identify and punish the attackers and to reprimand the Ottoman provincial officials. Throughout these first two sections both translators use the first person to refer to the sultan, and the second person for the Venetians. The third and last section of the letter marks a subtle but significant shift, as it moves to address the Venetian doge directly and urge him to ensure that in future, rather than taking matters into their own hands (and risk destabilizing the peace), the Venetians should appeal to the Ottoman court. It is in this section of the text, where the sultan seems to make explicit requests of the Venetians, that de Nores seems to get "nervous" about using person-marking that treats his Venetian patrician superiors as "you." Both Alberti and de Nores convey the shift in footing by resorting to the second person plural imperative mode to refer to the Venetian addressees (de Nores: "siate certi," "levandovi dalla mente"; Alberti: "non habiate à dubitar"). Yet, at crucial moments de Nores switches to a Venetian (or, at least, a less overtly Ottoman) perspective. First, on two separate occasions he seems to avoid the use of pronouns that would clearly mark the speaker as Ottoman. What Alberti translates, respectively, as "la buona pace, che è frà di noi" (the good peace, which obtains between us)-here, using the only inclusive first person in the letter-and as "amici di nostri amici" (friends of our friends), de Nores translates as "amicitia, et pace che è fra ambi le parti" (friendship, and peace which obtains between the two parties), and as "amici de gl'amici di questa eccelsa Porta" (friends of the friends of this Sublime Porte). Second, what Alberti renders as "non prestarete aiuto à nostri nemici" (do not lend help to our enemies) becomes, in de Nores' translation, "non darete alli suoi nemici alcuna sorte d'agiuto" (do not give their enemies any sort of help; my emphasis). By suddenly referring to the Ottomans in the third, rather than the first person, de Nores significantly changes perspective, and interrupts the conflation of the sultan's voice with that of the translator, implicit in Alberti's "transparency." Such a dramatic shift suggests the insecurities of a bureaucratic mediator whose foreignness was signaled not only by his Ottoman upbringing but also by his lack of ties to an established metropolitan family. As a late learner of Italian, and non-initiate into the conventions of diplomatic translation, de Nores used a register closer to spoken Venetian, which he would have employed daily in his oral interpretation in front of his patrician employers. All of this probably reinforced his sense of alterity. Alberti, on the other hand, from a recognized Venetian citizen family, and trained in the highest registers of chancellery Tuscan from youth, had no similar cause for concern. Resident in the bailo's house in Istanbul, he was adept at written translation, used a formal style, and inserted few cultural explanations into the translated text. De Nores' glosses for Ottoman custom as well as the framings he gave the translation and the shift in person marking were attempts to distinguish himself from the Ottomans in the act of translation. In so doing, he also showed his awareness of the risks inherent in any kind of mediation. study explored how trans-imperial trajectories informed the constitution of Venice's dragomanate in Istanbul, and how members of this cadre practiced mediation. Among the features I have emphasized are dragomans' explicitly heterogeneous (and increasingly understood as "mixed") provenance from among the Venetian citizen class, colonial nobility, and the Latin community of Pera, their lifelong employment in Venetian service following in the footsteps of fathers, uncles, and in-laws, their decades of residence and apprenticeship in the bailo's house, and their extensive sojourns in Venice and frequent trips back and forth between Venice, Istanbul, and colonies throughout the Adriatic and eastern Mediterranean. I have argued that many of these features closely resemble patterns of recruitment into and training within contemporaneous Ottoman elite households. Indeed, Venice's consciously heterogeneous dragoman recruitment patterns and its institution of lengthy apprenticeship within the bailo's household consolidated roughly at the same time, and perhaps in emulation, of late-sixteenth-century Ottoman emphases on specialized training and heritable status in the making of elite cadres. Yet, significantly, it was not any single feature of dragomans' background or training, or simply their bilingual skills, which constituted them as specialists of Venetian-Ottoman inter-imperial relations. Rather, it was precisely the welding of Ottoman and Venetian practices of subject-making that substantiated dragomans' claim to intimacy with elite culture in both metropoles, and that fostered their emergent perspective on the Venetian and Ottoman empires, their boundaries, and the modalities of bridging them.
Ultimately, the Venetian state, and especially its diplomats in Istanbul, recognized dragomans' in-between position as vital to preserving Venetian interests, but also as posing a perennial threat. Venetian diplomats attempted to control and contain the risk inherent in their reliance on dragomans by emphasizing loyalty and trustworthiness, rather than the command of languages, as the defining traits of a successful dragoman. Furthermore, whereas the potential transformation of Venetian into Ottoman subject was a constant concern, the reverse process of cultivated "Venetianness" was actively encouraged. The alchemy of comportment believed to result from long residence and employment in the bailo's house was such that in the course of the seventeenth century several dragomans with only tenuous claims to Venetian citizenship were able to eventually relocate to Venice and obtain government positions there. 71 In this sense, too, dragomans epitomize trans-imperial subjects, whose career paths and kinship and patronage networks crossed political and social boundaries between empires, between Venetian citizens and colonial subjects, as well as between different estates within metropolitan Venetian society. 71 For a case in point, see Rothman 2006: 362-68.