Part II: The Dissemination of An Imperial Modernity

The Ottoman Province of Trabzon

3. Horizons

Markets and States

Topography and Environment

The Pontic Mountains run from west to east across the upper tier of the peninsula of Asia Minor.^[1] The northern slopes of these mountains comprise a well-watered coastal region with a temperate climate. The southern slopes belong to the more arid, less vegetated, hot-in-summer, cold-in-winter interior highlands of Anatolia. This contrast between coast and plateau intensifies as the mountains rise from west to east (see map 2).

Along all the eastern segment of the coastline, from Ordu to Hopa, there are no large deltas, flat coastal strips are more infrequent and narrower, and the landscape is almost everywhere broken and precipitous. In the foothills near the shoreline, little hamlets with half-timbered houses pepper the foothills, sometimes barely visible among the trees and brush. Around each hamlet there are maize and bean gardens prepared by hand rather than by plough.^[2] Away from the shoreline in the upland valleys, the rainfall gradually decreases, the valleys are more deeply cut, and the mountains become steeper. Here the houses, almost always timbered, are sometimes clustered together on a promontory or in a valley, sometimes dispersed across the face of a mountainside (see fig. 6). The villagers are therefore forced to construct narrow terraces on the sheer mountainsides for their maize and bean gardens, sometimes securing themselves with ropes to tend them.

Owing to the uneven landscape, level cropland and open pasture are at a premium.^[3] To this day neither animal power nor machine power has ever been widely used in agriculture in the region. Along this section of the coast, nothing more than a meager living can be extracted from farming and herding. Farming is usually limited to horticulture. Herding is usually limited to a few cows. Large estates have always been the exception rather than the rule. Nonetheless, some of the area's inhabitants are able to become something other than subsistence farmers and herders, given the rela-tive absence of landlords and overseers. Leaving their garden farms to the care of women, the men of the eastern coastal region can look for work elsewhere in towns and cities of Anatolia.^[4] As a consequence, the residents of this entirely rural area have always included impressive numbers of soldiers, teachers, merchants, craftsmen, sailors, fishermen, peddlers, and laborers.

As one moves from the west to the east, the high Pontic peaks gradually come to define an "island on the land," set apart from the remainder of Asia Minor.^[5] East of the vicinity of the town of Trabzon, the littoral is not so much a greener version of Anatolia, but another kind of world altogether. Streams and rivers running north and south divide the entire eastern region into side-by-side valley-systems separated by densely vegetated foothills, falling and rising hundreds of meters to and from the water courses. Moving from one valley to another is often impractical owing to undergrowth and ravines. Instead, one must descend to the coastline, travel along the narrow shore, and then ascend into the next valley. Not so long ago, such a trek had to be made on foot or by horse. For lack of roadways, not even a cart could make its way through the foothills or across the rocky beaches. So it is that townsmen jokingly claim that the villagers did not learn of the invention of the wheel until the arrival of motor transport.

By contrast, the farming and herding world across the Pontic mountain chain is entirely different in character.^[6] The landscape is open and treeless, sometimes rocky and barren. Winters are cold. Summers are hot. Dry-farming (wheat and barley), stream irrigation (rice and vegetables), and goats and sheep (yogurt and cheese) are the basis of subsistence agriculture. Ploughs and carts are pulled by draft animals. Houses are constructed of mud bricks. Timber, being scarce, is used only for the roof poles, doors, and sashes. Until recently, the only fuel was cattle dung, which was molded into patties (tezek) and dried in the sun. Highly efficient ground ovens (tandıır) are the means for both cooking and heating.

To the north and south of the Pontic chain, rural life is governed by contrasting requirements and possibilities. Crops, flocks, tools, building materials, house designs, kitchen fuel, diet and cuisine, dress and manners, bodies and faces, accents and dialects are different. And some time ago, for a span of centuries, language, religion, and state were also not the same.

Ethnic Fragments and Linguistic Archaisms

Following the Byzantine defeat at Manzikert in 1071, Turkic pastoral peoples began to enter and occupy many sections of Asia Minor, changing the character of its villages and towns. By the thirteenth century, the interior highlands of northeastern Anatolia had been under Turco-Islamic rule for more than a century, and a majority of the population had become Turkish by language and Muslim by religion.^[7] This transformation had not always come about with the displacement of the older Byzantine peoples. In many places, groups of Greek-speakers and Armenian-speakers had gradually assimilated themselves to the newcomers, first losing their languages to acquire Turkish, then losing their religion to become Muslim.^[8] In contrast, the older Byzantine peoples of the eastern littoral were neither Turkicized nor Islamized until a much later date, and then by a different path.

The eastern coastal region, first as the province of Chaldia in the Byzantine Empire (until 1204), then as the Greek Empire of Trebizond (until 1461), had for a long while remained outside the orbit of the Turco-Islamic states of the interior highlands. Finally capitulating to Sultan Mehmet II, this last mainland fragment of Byzantium subsequently reemerged as the province (paşalıık) of Trabzon. For more than a century, most of the older Byzantine peoples remained relatively unaffected by incorporation. Then, during the course of the second century of Ottoman rule, as a consequence of both conversion and immigration, the large majority of the inhabitants became Muslim. Even so, substantial numbers of the Muslims, most of them descendants of the older Byzantine peoples, continued to speak mother tongues other than Turkish. The eastern coastal region therefore stands as an "exception" twice over to what had happened in the interior highlands.^[9] A Muslim majority did not emerge until the seventeenth century, almost four hundred years after the rest of northeastern Anatolia. And when this Muslim majority did emerge at last, its constituents spoke a variety of languages, such as Turkish, Lazi, Greek, and Armenian.

As both Anthony Bryer and Xavier de Planhol have pointed out, the high Pontic chain played a decisive role in determining the different course of history in the eastern coastal region.^[10] The arrival of large numbers of Turkic pastoral peoples had guaranteed that the population in northeastern Anatolia would be relatively quickly Turkicized and Islamized. In contrast, the eastern littoral was far less accessible to the semi-nomadic, stock-keeping peoples of the interior highlands. The passes that cut through the mountains consisted of little more than narrow and twisting tracks, buried in deep snows during the winter. Descending into the valleys, these tracks traversed a landscape ideally suited for defensive purposes: virgin forests shrouded in mists at the upper elevations and a dense undergrowth of bushes and vines at the lower elevations.^[11] By these circumstances, the rural societies of the coastal valleys were in a position to limit the numbers of pastoral newcomers who settled in their midst, just as the Greek Empire of Trebizond was in a position to resist military invasion and occupation by the Turco-Islamic states of the interior.

Viewed as a great mountain barrier, the Pontic chain explains why the rural societies of the eastern littoral were slow to change, as well as why the Greek Empire of Trebizond was to endure so long. Otherwise, topography and environment did not consistently function to isolate the coastal region from the outside world.^[12] Even as the high mountains and dense vegetation defined an "island on the land," a kind of refuge area set apart from the interior highlands, its temperate climate and fertile soils were powerful magnets that lured peoples into it. The two opposed qualities of the landscape, defensibility balanced against desirability, led to a pattern of ethnic fragmentation. For whenever outsiders did succeed in penetrating the coastal region, they tended to retain elements of their distinctiveness.^[13]

By the early medieval period, before the arrival of Turkic pastoral peoples in Anatolia, the northern slopes of the eastern Pontic Mountains were occupied by peoples who had colonized the region from different directions. Kartvelian-speakers from the Caucasus, eventually to be called the Lazi, had settled its eastern precincts. Greek-speakers from Sinop, eventually to be called Pontics, had settled the western precincts. Armenian-speakers from the interior highlands, eventually to be called Hemşin, had entered the eastern upper valleys above the Kartvelian-speakers and Greek-speakers.^[14] Thus, the coastal region had inexorably drawn peoples from neighboring territories into its valleys, complicating the ethnic composition of the coastal region.

Almost surely, Turkic peoples appeared in most of the coastal valleys soon after their arrival in the interior highlands, perhaps as early as the eleventh century.^[15] It is even likely that some of these early arrivals assimilated themselves to the existing inhabitants, losing their language and their religion, only to get them back centuries later.^[16] Whatever the case, Turkic pastoral peoples did not initially enter and occupy the coastal region in large numbers, save where the mountains were lower and the landscape less vegetated. Çepni Turcomans, tribally organized pastoral peoples of heterodox Shi' background, first began to settle along the western littoral in the vicinity of Sinop, then reversed direction to move back toward the eastern littoral. By the thirteenth century, the emirates of these peoples governed the coastal region between Ordu and Sinop. And by the fourteenth century, Çepni Turcomans were moving still further eastward, settling the more accessible lower valleys just to the west of the town of Trabzon.^[17]

At the moment of Ottoman incorporation, the overall distribution of ethnic groups in the early province of Trabzon can be roughly described as follows.^[18] Greek-speakers inhabited most of the lower and upper coastal valleys near Trabzon, both to the east and to the west. This inner core of Greek-speakers was flanked by Kartvelian-speakers living in the valleys east of Rize and by Turkic-speakers living in the valleys west of Giresun. Groups of Armenian-speakers inhabited some of the upper valleys above the Kartvelian-speakers in the east.^[19] Groups of Greek-speakers inhabited some of the western upper valleys above the Turkic-speakers in the west.^[20] This pattern of settlement then became further complicated during the period of Ottoman rule. From the later sixteenth century through the seventeenth century, large numbers of Muslim settlers, most of who were Turkish-speaking, but not necessarily Çepni Turcomans, moved into various parts of the eastern coastal region. Their arrival appears to have led to relocations and conversions among some portion of the Christian population, thereby enhancing the mixed and merged character of local communal groupings. Just as some number of Turkic pastoral peoples had probably become Orthodox during the earlier period, many of the older Byzantine peoples most certainly became Muslim during the later.^[21]

Given the traces of many peoples and languages in the coastal region, travelers were consistently perplexed about the nature of its peoples, during the Roman, Byzantine, and Ottoman periods, down to the present. By what name is the population to be called? What language do they speak? With what religion are they affiliated? These questions had no simple answers given the situation of ethnic fragmentation. Groups of different peoples were unevenly distributed across the landscape, sometimes interspersed among one another, and always mixing with one another. In one valley one language would be spoken, but in the neighboring valley another language might be spoken. Furthermore, in the same valley, one language might be spoken at the lower elevations and another at the higher elevations. At the same time, several languages would be spoken in the lowland and highland markets.

But even more puzzling to outsiders than the ethnic and linguistic complexity of the population was the fact that ethnicity and language were not correlated with political identity and religious affiliation during the later Ottoman period.^[22] In the coastal region, as we shall see in later chapters, the older Byzantine peoples became Muslim by participating in an imperial project rather than by assimilating to a Turco-Islamic majority.

Society and State in the Pontic Enclave

The rural societies of the eastern coastal region have been adjudicated, taxed, and conscripted by a state system from the medieval period, during which time the town of Trabzon consistently served as the region's administrative, military, religious, and commercial center. As Bryer has noted, the structure of the state system has been remarkably constant over the centuries.^[23] The valley-systems that constituted districts (bandon) in the ninth-century Byzantine province (thema) of Chaldia reappeared as districts in the Greek Empire of Trebizond. Later, they became districts in the sixteenth-century province (paşalıık) of Trabzon, and then reappeared once again as districts in the twentieth-century provinces (vilayet) of the Turkish Republic. This structure points to the strategic position of the regional capital and its links to other major markets and ports.

Walled and fortified from the medieval period, the town of Trabzon was situated on a rocky coastal promontory rising between two ravines. Some of the best anchorages along the eastern coast are in the vicinity, and a route from the town through the mountains is passable during both the summer and the winter. An entrepôt for maritime and overland commerce, the regional capital was inhabited by officials, mercenaries, merchants, craftsmen, travelers, and adventurers of diverse backgrounds from different homelands.^[24] The rural societies flanking the regional capital were in contact with the more cosmopolitan society that appeared there, and some number of their residents lived and worked in its administrative, military, religious, and commercial centers. As we shall see, however, their links with the outside world beyond their valleys was not restricted to this single channel.

An early hint of the special relationship between rural societies and the state system appears in the works of Procopius. The sixth-century Byzantine historian refers to the inhabitants of the coastal region as the "Romans who are called Pontics." In so choosing his words, Procopius designated rural societies of the coastal valleys as a provincial population of the Byzantine Empire. He therefore conceived of a collection of different peoples speaking different languages as representatives of an imperial civilization. Much later, during the time of the Greek Empire of Trebizond, outsiders had considered the inhabitants of the coastal region to be primarily "Laz." Once again, the term in question characterized a diverse collection of peoples as the provincial population of what was then a "regional" imperial civilization.^[25]

In the eighteenth century the same pattern appears again, even though the population in question was no longer either Byzantine or Christian. Outsiders commonly used the term "Laz" to describe what was now a large Muslim majority in the coastal region as a provincial population of another imperial civilization, that of the Ottoman Empire. As before, this provincial population included a diverse collection of peoples. Descendants of Turkic-speakers, Kartvelian-speakers, Greek-speakers, and Armenian-speakers had aligned themselves, however imperfectly, with official Ottoman Islam, and all of them spoke greater or lesser snatches of Ottoman Turkish. Since none of the four groups had previously featured these attributes (Turkic pastoral peoples included), they each separately indicate a strong tendency for the residents of the coastal valleys to position themselves in the larger social, economic, and political world beyond their scattered hamlets.^[26]

While geography worked to preserve ethnic fragments and linguistic archaisms, it also worked to homogenize identities and relationships. The terrain of the eastern coastal region isolated the residents of the coastal valleys from local processes of social assimilation, but it also encouraged, if not necessitated, that they connect themselves with the outside world. Each valley linked oversea routes with overland trade routes. Some had good anchorages along their shorelines. Some provided easy access to the interior highlands through mountain passes. Some were located near major markets of Anatolia. Some combined all these features, so that they rivaled the regional capital at Trabzon. At the same time, each valley was also a total ecological system. Every year individuals, families, and communities moved up and down the valley. Petty traders, craftsmen, and wage laborers moved from the markets of the coastal lowlands to the markets of the interior highlands. Family members moved cattle and flocks in stages through the upper valleys, gradually ascending to the highland pastures in the spring and descending to their lowland hamlets in the fall. Entire villages moved between a winter settlement near the shoreline and a summer settlement in the mountains.

As homelands, the valleys were not what they first appeared to be, motley groups of people living in isolated hamlets dispersed across the landscape. The valleys were instead transit systems whose inhabitants had a large stake in the security of the people, animals, and goods moving through them. All the residents of each valley-system, whether lowlanders or highlanders, therefore had an interest in knowing one another and cooperating with one another. Accordingly, each valley-system comprised an integrated social network. That is to say, all its residents were more or less organized for the purpose of communicating information, organizing cooperative ventures, and providing mutual assistance. In this respect, the peoples of the coastal region, densely settled and agriculturally impoverished, were also oriented toward outside markets of all kinds, for labor, skills, services, and goods. And given the existence of networks of relationships and lines of communication inside their homelands, they were in a position to make their way in the world abroad beyond their coastal valleys.

As a consequence, the rural areas of the coastal valleys have been inhabited for many, many centuries by cash-croppers, soldiers, miners, carpenters, sailors, metal-workers, weavers, rope-makers, dyers, masons, cooks, laborers, and traders, as well as a small minority of extortionists, counterfeiters, smugglers, gunrunners, kidnappers, bandits, pirates, and assassins. The residents of the rural areas have plied all these legitimate and illegitimate activities in their homelands since ancient times, but they have also left their homelands to go abroad as migrants when there was an opportunity to do so. Similarly, just as these rural societies have been oriented to external market conditions, so, too, outsiders were always in a position to become insiders among them. Certain kinds of "specialists" have always infiltrated the coastal region when they had something to offer its resident peoples.^[27] Among them we find animal transporters, religious teachers, military officers, stone-layers, wood-carvers, metalworkers, shipwrights, and so on. In this respect, the coastal region, which is otherwise completely rural in character, bears a resemblance to an urban center.

There were, then, two contradictory sides to the rural societies of this segment of the coastal region. Their homelands served as isolated refuges where archaic traits and dialects were preserved. And yet, their homelands, while isolated refuges, were also connected with the outside world. Topography and environment did not have a single, undivided consequence for social life in the coastal region. It worked to fragment human groups at the level of mother tongue and family life even as it also worked to unify human groups at a level of entrepreneurial, governmental, and educational engagements.

The two opposed tendencies explain why the residents of the coastal region have more or less consistently had the reputation of hicks since the medieval period. They issued from their remote valleys with outlandish accents, manners, and dress, but, unlike other rural peoples, they were nonetheless consistently present and active in provincial centers. What is significant is not that they were dismissed as rustics, but rather that they were always on the scene in the towns and cities of Asia Minor, ready and available to be designated as countrified interlopers. The peoples of the coastal region were always striving to integrate themselves into a larger state society, even as they could rarely achieve urbanity and sophistication, since most of them retained contact with their rural homelands as refuges and hideouts.^[28]

Once any group of people left the interior highlands, crossed the Pontic Mountains, and settled in the coastal region, their work habits and mental outlooks were inexorably transformed. Gradually, the newcomers found it more and more difficult to conceive of the local community as an entity in itself and they came to see themselves in relation to the commercial and governmental systems beyond their valley homelands. They tended, then, both as individuals and communities, to see themselves as participants in universal projects of power and truth. Once in the coastal region, the heterodox became orthodox and the illiterate became literate. They became partisans first of Romanism and Byzantinism, then of Ottomanism and Nationalism. A further comparison of the coastal region with the interior highlands brings to light the way in which local and global factors combined to reinforce a preoccupation with, as I have chosen to phrase it, "the horizon of elsewheres."

Economic Flexibility and Elasticity

Across the Pontic chain in the interior highlands, subsistence farming and herding are demanding occupations, narrowly constrained by circumstances and resources. Winters are long, requiring the maintenance of sturdy houses, stables, and barns. In the absence of trees, fuel for cooking and heating is limited but necessary for any measure of comfort. Plough farming is not possible without large draft animals, which have to be maintained year-round. The growing season is short, so farm work intensifies at specific times, when family groups must work long and hard. During the warmer months, from March through November, there is a time to sow, a time to harvest, a time to gather and store fuel, a time to pasture and corral flocks, and a time to repair roofs and walls. Moreover, each household's activities cannot be organized independently from others. The fields of the villagers are mingled together at the periphery of their nucleated settlements. Fallowing, ploughing, harvesting, pasturing, manuring, and penning are virtually impossible without collective agreements and mutual assistance.

In contrast, the subsistence activities of family groups in the coastal region are not so closely defined by climate, labor power, resources, equipment, and community. In most places the growing season is long, almost year-round near the shoreline. Water is almost never in short supply, and wood for fuel is readily available. Each hamlet is more or less isolated in the midst of its own lands, and, in many places, it stands as a kind of little kingdom on a promontory or hilltop. As a result, the villagers, especially the majority in the lower valleys, are able to organize their subsistence activities more flexibly, coordinating them with other kinds of engagements and activities. The maize and bean gardens can always be made a little larger or a little smaller.^[29] The planting of potatoes and squash can be expanded or contracted during the warmer months, just as the planting of cabbage and leeks can be expanded or contracted during the colder months. Since there is no need for oxen, ploughs, horses, and carts, subsistence activities can also be incrementally adjusted downto a bare minimum. There can always be a few more or a few less animals in the stable beneath the house, or even no animals at all.^[30]

By virtue of the flexibility and elasticity of subsistence activities, those members of the household who preoccupy themselves with subsistence activities are variable in number. Some members of the household can be absent from the hamlet for long periods of time, even during the warmer months, without seriously interrupting the cycle of farming and herding activities. When other sources of income present themselves, some members of the households can leave. It is almost always men who do so. Those remaining, some of the men and all of the women, can scale back their subsistence activities accordingly. This allows for the coordination of subsistence farming and herding with itinerant trades and crafts. Such coordination was an ancient practice in the coastal region, and especially characteristic of the later centuries of the Ottoman Empire.

If a handicraft was seasonally in demand in a town of the interior highlands, some of the men of the coastal region learned it, temporarily migrated, and practiced it. If commercial possibilities arose in some section of Asia Minor, the Caucasus, eastern Europe, or the Middle East, some went there as petty traders and wage laborers.^[31] If state officials were looking for soldiers and sailors, some took this chance to gain some experience abroad. As a consequence, the men of the coastal region were never entirely bound to the peasant way of life.^[32] Their thinking and practice took into account opportunities that might appear over the horizon.^[33] If they had a keen sense of their local origins, they nonetheless conceived themselves in relationship to all kinds of elsewheres, even to the point that subsistence became a secondary rather than primary concern.^[34] Cash-cropping, handicrafts, manufactures, and labor migration therefore tended to degrade the production of basic food stuffs.^[35]

The assortment of local products continually changed and shifted with market conditions.^[36] During the later Ottoman period, dried and fresh fruits, such as grapes, cherries, and citrus, were cultivated and exported in significant quantities to Istanbul. To the east, locally grown hemp and flax were used to make thread, nets, and cloths.^[37] A fine linen (Trabzon bezi, Rize bezi) that became prized in the Middle East and Europe was also produced.^[38] To the west, grapes from vineyards were distilled into a powerful liqueur (nardenk) and exported to the Crimea for reduction into eau-de-vie.^[39] Wooden boats (kayıık) for transporting cargo to Black Sea ports were constructed in Hopa, Rize, and Araklıı, as well as elsewhere along the coast in secluded coves invisible from the sea.^[40] Copper, lead, and silver were mined, smelted, and cast in different vicinities near Trabzon.^[41] Knives, swords, pistols, and rifles were manufactured no later than the nineteenth century and perhaps considerably earlier.^[42] Paper money was counterfeited no later than the beginning of the nineteenth century, and one can guess that metal counterfeits had been produced from ancient times.^[43] Most of the local products that flourished during later Ottoman period have declined or vanished since the beginning of the last century, only to be replaced with others. But a very few were produced for centuries, and some for much longer. Tea is a crop grown today that was first introduced between Rize and Trabzon during the mid-twentieth century, but hazelnuts have been cultivated and exported from the eastern littoral for two thousand years.^[44]

By virtue of a market orientation, subsistence farming and herding placed no Malthusian limit on how many people could inhabit the eastern coastal region.^[45] So the population of the coastal region, like that of a metropolitan area, could expand indefinitely, limited only by the ingenuity of its inhabitants to produce for a market, turn a profit in commerce, or go abroad to ply a trade or seek out work.^[46] And in fact the population did expand once the coastal region became more firmly integrated into the imperial system. British consul James Brant, traveling eastward from Trabzon in 1835, made the following observation:

I passed in succession the districts of Yemourah, Surmenah, Oph, Rizah and Lazistan. . . . The country is so wooded and mountainous, that it does not produce grain sufficient for the consumption of the population, yet not a spot capable of cultivation appears to be left untilled. Corn fields are to be seen hanging on the precipitous sides of mountains, which no plough could arrive at. The ground is prepared by manual labour, a two pronged fork of a construction peculiar to the country being used for this purpose. Indian corn is the grain usually grown and it is seldom that any other is used for bread by the people. What the country does not supply is procured from Gouriel and Mingrelia.^[47]

As the population doubled, redoubled, and doubled again, grain deficits became endemic to the eastern coastal region.^[48] And when such deficits were exacerbated by poor harvests and market crises, the rural societies of the coastal region suffered extreme hardship and impoverishment.^[49] British consul William Palgrave, who resided in Trabzon from 1868 to 1873, described the miserable condition of the villagers near Trabzon in the following terms:

The inhabitants are, with hardly an exception, wretchedly poor. The plot of ground on which each man cultivates his maize, hemp, and garden stuff, yields little more than enough for his own personal uses and those of his family; the maize–field and garden supply their staple food, and the hemp their clothing: this last coarse and ragged beyond belief. And no wonder, where a single suit has to do duty alike for summer and winter, day and night.^[50]

Cash-cropping and labor migration were alternatives to the meager fare gained from gardens and stables, but they also exacerbated the precariousness of rural life. The market orientation of the villagers decreased the local production of basic foodstuffs and so increased dependency on external economic and political conditions. The villagers were ever more inclined to contemplate the possibilities and opportunities beyond their homelands as the population density climbed. The eastern coastal region therefore represented an impressive reservoir of men who were able and eager to take up any opportunity that might be present itself. At a certain point of crisis in the imperial project, Ottoman officials took note of this fact.

Ottomanization of Trabzonlus, Trabzonization of the Ottomans

In addition to cash-cropping, handicrafts, manufactures, and labor migration, some inhabitants of the coastal valleys had always entered governmental service. Most commonly, it seems, they served as mercenaries for all kinds of power-holders in the coastal region as well as other parts of Asia Minor. As it happened, however, the inhabitants of the eastern coastal region began to expand the extent of their participation in centralized governmental institutions during the third century of Ottoman rule. Afterward, the rural societies might have appeared to the casual observer as dispersed and fragmented peasantries, but they had become, more than ever before, provincial extensions of the imperial system.

During the second half of the seventeenth century, official policies led to an unprecedented increase in the numbers of individuals associated with military and religious institutions in all the core provinces of the Ottoman Empire. In a context of internal instability and external competition, a significant fraction of the population joined the ranks of soldiers and preachers. Sometimes employed and sometimes unemployed, tens of thousands of individuals came to identify themselves with the imperial system. The rise in the numbers of soldiers and preachers, which would have had affected towns and villages everywhere, had a special impact on the rural societies of the coastal region, once again by reason of topography and environment.

Elsewhere among the peasant societies of Anatolia, soldiering and preaching had the potential to wreck the rural economy coming and going. By the coincidence of these activities with the growing season, they drew able-bodied men out of the subsistence economy when they were most needed, then returned them when prospects for productive activities were at a low point. In contrast, the men of the eastern coastal region could depart from and return to their homelands without seriously disrupting the subsistence economy. And given their background of participation in market and state systems, they could immediately understand how state service was an opportunity rather than a catastrophe. Soldiering and preaching were ways of insinuating themselves in the social networks of the state system. And as well, soldiering and preaching could be combined with trade and crafts.^[51] As a consequence, the rural societies of the province of Trabzon did not remain Christian but became largely Muslim, as they were once again, but more than even, re–integrated within a state and market system, now global rather than regional.

Women's labor in the gardens and stables was the precondition for men's participation in the horizon of elsewheres. It was the men who soldiered and preached and, in doing so, traveled to other towns, resided in dormitories, socialized in coffeehouses, set up shops and ateliers in markets, ran caravans into eastern Anatolia, and sailed transport ships along the coast. If women remained involved in the market economy, that is to say, with cash-cropping, handicrafts, and manufactures, they did so only within their mountain fastnesses. And by the logic of such circumstances, women's work in the fields and stables became the confirmation that the men of the household were something more than peasants. Women were therefore obliged to work in the fields and stables whether or not the men were absent from the homestead.

In this way, local engagement in the imperial project came to shape gender relations in the coastal valleys of the eastern littoral. Women were closely identified with subsistence tasks, which were necessary but unrewarding and undignified. Men were identified with governmental and commercial engagements beyond the hamlet, which were potentially more rewarding and certainly more prestigious and dignified. By the pressure to claim social standing, each household was organized in such a way that the men were freed from subsistence activities, regardless of their actual position or achievement. While men were absent from the homestead during the day (in principle if not in fact), women took care of the children, the house, the stable, and the fields.^[52] It was just as important for a man to be seen as free from farming chores as it was for him to be involved in some other rewarding enterprise.

The issue was as much a question of propriety as economic necessity or possibility. When they were able to do so, regardless of their actual occupation, men wore the "official" attire of businessmen, professionals, and bureaucrats. Women, regardless of their circumstances, wore the peasant costumes of striped aprons (önlük) and shawls (çarşaf) or blouses and baggy pants (şalvar), dress suitable for heaving, carrying, digging, and clearing. Wicker baskets, a distinctive feature of the subsistence economy in the coastal region, were another indication of the differing relationship of men and women to subsistence activities.

Women with baskets—some as small as a rucksack and some as large as a person—took the place of animal power and wheeled vehicles.^[53] Until recently, women were seldom seen away from their homesteads without one, gathering fodder for the stables, harvesting crops from the fields, or transporting produce and supplies. Women were the means by which men could become something other than subsistence farmers and herders, that is to say, the means by which men could claim imperial identities and affiliations. This explains why there was competition among men for the control of women as well as why this competition was expressed during marriage festivities in terms of the assembly of large numbers of men engaging in impressive displays of firearms. One could not be a participant in the sovereign power of the state system without women's labor, hence the manifestations of numbers and firearms were necessarily linked with claims over women. Villagers who saw themselves as potential participants in the sovereign power of the state system were also villagers for whom numbers and firearms symbolized their claim to women.

In the same way, imperial engagement explains why the eastern Black Sea people are known by story and proverb as possessive and jealous. Competition for women's labor arose because men were required to be absent from the homestead, not just during the day, but also during major parts of the year. Locally, these circumstances were associated with the notion that women should be able to defend themselves with weapons in the absence of men.^[54] But more tellingly, they reinforced the moralization of men's claims over women, such that a principle of manhood required respect for another man's claims over women. This moralization appears in the orientation of religious thinking and practice in the eastern coastal region. All those elements of official Islam that touch on the morality of gender relations are emphasized, if not exaggerated, in the eastern coastal region.

So it is that the eastern coastal region, by participation in imperial military and religious institutions, became famous for a "traditional" assertion of male rights over women. Andsince its soldiers and preachers became instrumental in the ottomanization of provincial society, so too they transmitted the moralization of gender relations to other parts of Anatolia. The male residents of the eastern Black Sea coast became soldiers and preachers only by the advantage of local circumstances. And as they did so, they carried into the state society their own special concerns that had arisen from the circumstances of their participation. As they were transformed by taking their place in the lower echelons of the imperial system, so too they transformed the lower echelons of the imperial system. Ottomanization of Trabzonlus led inexorably to Trabzonization of the Ottomans.

A State Society Before Contemporary Modernity

Travelers and visitors have repeatedly perceived the landscape of the eastern littoral as a stabilizing foundation for social life. Xenophon in the fifth century B.C.E., Procopius in the fifth century C.E., and then French consul Fontanier in 1827 all describe the same scene of isolated hamlets scattered across the hillsides. Topography and environment, it would seem, have always regenerated the same way of life, despite the arrival and departure of various peoples. A closer look reveals, however, that this continuity of observation is misleading.

Domestic architecture is less uniform here than anywhere else in Asia Minor. Sümerkan notes that the techniques of house construction in the coastal region change every fifteen to twenty kilometers, an indication of the diverse origins and skills of its inhabitants.^[55] And if some domestic implements and structures are common to all the peoples of the littoral, reappearing in every coastal valley, such devices are everywhere named by different terms from different languages. Sümerkan notes, for example, that the maize granary was designated by as many as fifteen different terms.^[56] And yet all this variability at the level of everyday practical activities is strangely accompanied by uniformity at the higher level of social relations. Sümerkan notes that the terms for domestic architecture are Arabic for activity rooms, Ottoman for the sitting room, and Persian for the balcony.^[57] In other words, local fragmentation and difference is accompanied as well by the disruptions and intrusions of world commerce and imperial government.

The truth of the matter is that topography and environment worked to destabilize, not stabilize, the character of social life. On the one hand, outsiders were always attracted to this fertile and temperate region, thereby disrupting the consolidation of parochial habits and custom. On the other hand, once these outsiders had become insiders, they were pressed to orient themselves to the commercial and governmental systems beyond their homelands. And as the villagers oriented themselves to the market and state systems of which they were a part, so local identities and relationships never came into balance but always remained off-center, like the world system of which they were a part. The surface appearance of an unchang-ing way of life, dispersed homesteads in a garden landscape, therefore concealed a history of ethnic fragmentation and imperial reorientation.

The most perceptive observers of the old province of Trabzon noticed this peculiarity of the eastern coastal region, even though they did not understand exactly why or how it had come about.^[58] J. Decourdemanche, who visited the coastal region during the 1850s, described the Muslim population as a collection of diverse peoples who had coalesced to form a regional society in the imperial system. "The inhabitants of Trabzon consist of many races . . . who came closer to each other in religion and speech [in the past so that] today they are one caste which has social and political influence.^[59] Alfred Biliotti, serving as the British consul in 1873, also described the Muslim population as diverse in background but nonetheless united in their imperial sentiments and participation. "However much the Muslims who live here may be from the point of view of root or structure from different races, it can be thought that they are in the same fashion close to one another in regard to the Ottoman Empire, religious belief, and political ties.^[60] When these observations were made, the steamship and telegraph had only recently arrived in this part of Asia Minor. It would be almost a century more before motor transport and electronic communications would become important factors in the social life of the coastal region. Nonetheless, in the absence of technological modernity, a collection of peoples living in a rural landscape more or less without towns had become a state-oriented society.

Today, in the eastern Black Sea provinces of Turkey, from Artvin to Ordu, the traces of ethnicity, Lazi, Armenian, Greek, and Turkic, are easy to discover in language, stories, customs, and dress. And yet, in contrast to all these differences, the inhabitants of the rural societies still lack a strong interest in their parochial backgrounds and traditions.^[61] With the exception of a few recent authors and books, there is no developed culture of ethnicity in the eastern coastal region.^[62] Instead, social manners and relations are more or less homogeneous at a certain level, a direct reflection of a local engagement in wider market and state systems, now national as once before imperial.

These conditions of social life among eastern Black Sea peoples are indirectly recognized in Anatolian folklore. The diet of cornbread and yogurt is the simple and modest fare of village life in the coastal valleys. According to villagers and townsmen in the rest of Anatolia, these food items are said to be the cause of the restless energy and quick temper for which the "Laz" are famous. Cornbread and yogurt are indeed responsible for their liveliness, but only indirectly by their correlation with cash-cropping and labor migration. To be obliged to survive on cornbread and yogurt was also to have an intense interest in the horizon of elsewheres.

Notes

4. Empire

Gaze, Discipline, Rule

The Problem

In the last chapter, I noted that the inhabitants of the eastern coastal region eventually came to identify with and participate in the institutions of the Ottoman Empire. But if these rural peoples were inclined to align themselves with the imperial system, this does not mean that the ruling institution itself would have permitted, let alone encouraged, such an accommodation. Indeed, the very idea of a rural people becoming ottomanist in orientation contradicts the prevailing historiography of the Ottoman Empire. Most commentaries have emphasized an unbridgeable divide between its ruling (askeri) class of state officials and its ordinary subjects (reaya), both Muslim and Christian. How then could a population of gardeners residing in remote mountain hamlets find themselves a place in the imperial system?

To answer this question, I first review how the Ottomans incorporated the eastern coastal region soon after the conquest of Constantinople, at a time when they were perfecting the classical ruling institution. This done, I analyze the architecture and ceremony of the governmental complexes that they built in the new imperial capital. This analysis features a double objective: to lay bare the distinctive configuration of sovereign power in the imperial system, and to expose channels of popular identification and participation that would lead into it.

Ottoman Centralism and Exclusivity

From the early sixteenth century, western European observers began to perceive the Ottoman Empire as a remarkable example of the centralism and exclusivity of sovereign power. What they noticed were the features of a new imperial system that Mehmet II had developed following his conquest of Constantinople in 1453.^[1] Niccolò Machiavelli's comparison of the French and Ottoman governments in The Prince (1515) exemplifies the contemporary assessment:

The entire monarchy of the Turk is governed by one lord, the others are his servants; and, dividing his kingdom into sanjaks [sub-provinces], he sends there different administrators, and shifts and changes them as he chooses. But the King of France is placed in the midst of an ancient body of lords, acknowledged by their own subjects, and beloved by them; they have their own prerogatives, nor can the king take these away except at his peril. Therefore, he who considers both of these states will recognize great difficulties in seizing the state of the Turk, but, once it is conquered, great ease in holding it. The causes of the difficulties in seizing the kingdom of the Turk are that the usurper cannot be called in by the princes of the kingdom, nor can he hope to be assisted in his designs by the revolt of those whom the lord has around him. This arises from the reasons given above; for his ministers, being all slaves and bondmen, can only be corrupted with great difficulty, and one can expect little advantage from them when they have been corrupted, as they cannot carry the people with them, for the reasons assigned.^[2]

The passage points to a stereotype of Ottoman government that had come to prevail in Christian Europe. The sultan ruled through a body of officials having the legal status of household slaves. Recruited from the children of Christian families and trained from adolescence within the confines of the sultan's palace, they had no independent social identities or loyalties.^[3] At the same time, these slave officials (kul) were unchallenged by any system of estates. There was no aristocracy composed of lords who ruled their own peoples and territories, and no bourgeoisie composed of merchants or bankers who had been granted the privilege of governing their own towns and cities. The "Grand Turk," as the sultan was sometimes styled in Christian Europe, seemed to enjoy a measure of sovereign power unmatched by any other monarch in early modern Europe.

Machiavelli's analysis reduced the ruling institution to a simple and static formula, thereby concealing both its complexity and instability. Still, his formula directs our attention to a distinctive feature of a new governmental system that gained ground during the early classical period. As the Ottomans launched a world imperial project during the later fifteenth century, they reinforced the centralism and exclusivity of the ruling institution. The recruitment and training of "slave" children to serve as high state officials was just one of the measures they adopted. By means of a range of policies, the Ottomans came to rely on a special class of military, administrative, and judicial officials who lacked affiliation with the governed. This raises the question of how the Ottomans incorporated a region whose peoples had such a large stake in market and state participation.

Mehmet II had annexed the Greek Empire of Trebizond (1461) just as he was beginning to devise and apply the new imperial system. Süleyman I had later reorganized the province (paşalıık) of Trabzon as a new administrative entity (1519) at the high point of classical institutions. So the substantial Christian population of the coastal region had become subjects just as the Ottomans were perfecting the centralism and exclusivity of the ruling institution. Higher state officials were more than ever composed of slave officials, and other entry points into the ranks of officialdom were regulated more than ever. Thus the shock of conquest was compounded by the shock of subjection. The old rural societies of the coastal region and the new imperial system were exactly mismatched. The inhabitants of the province of Trabzon had become part of a governmental system based on principles that stood in direct opposition to compelling local interests.

As we saw in the last chapter, the mismatch was transitory rather than permanent. As the domains of the ruling institution reached their maximum limits, the conduct of warfare was shifting away from the use of cavalry toward the use of infantry with firearms.^[4] Under these circumstances, the Ottomans came to require larger numbers of men with a wider range of skills, even before the close of the classical period. They therefore took steps to widen the circle of participation in imperial military and religious institutions during the seventeenth century, in effect compromising the principles of centralism and exclusivity. As they did so, problems of imperial competition at the military frontier were joined by problems of internal instability in the core Ottoman provinces. Provincial governors had begun to defy the central government, asserting themselves by collecting illegal taxes and maintaining their own private armies. In response, the palace widened the privileges and prerogatives of provincial elites at the district level in hopes of curbing the powers of provincial governors.^[5] As a result, principles of centralism and exclusivity were compromised still further. By the beginning of the eighteenth century, the distribution of sovereign power had moved outward and downward into the imperial system, weakening both central and provincial government. Higher state officials found themselves unable to rule save with the acquiescence and assistance of provincial elites.^[6] The post-classical period was thereby characterized by a progressive decentralization of sovereign power.

All the core Ottoman provinces were affected by the changes I have just summarized, but the province of Trabzon is an especially revealing example of the phases of decentralization. By the close of the seventeenth century, many of the districts where the population had been almost entirely Christian at the moment of Ottoman incorporation had become almost entirely Muslim. Furthermore, large numbers of the men among these new Muslim populations had associated themselves with local branches of imperial military and religious institutions. By the close of the eighteenth century, local participation in imperial military and religious institutions had resulted in an entirely new relationship of state and society. Provincial elites at the head of armed followings asserted their prerogatives in the imperial system, sometimes defying, even threatening, higher state officials, both those in Trabzon as well as in Istanbul.

The instance of the eastern coastal region therefore poses questions of general significance for the understanding of the ottomanization of provincial society in other parts of Asia Minor and the Balkans. How did a population composed of different ethnic groups attached to different religions come to participate in imperial military and religious institutions? And in doing so, how did this diverse and mixed population become state-oriented, official Muslims, given that the imperial regime was based on radical principles of centralism and exclusivity?

Ottoman Incorporation of Trabzon

In his account of the ruling institution during the classical period, İİnalcıık explains how the palace, the seat of centralized government, organized and supervised the core Ottoman provinces in Asia Minor and the Balkans.^[7] A hierarchy of military officers was appointed to govern a hierarchy of administrative units. There was a governor of each province (beylerbeyi), a few sub-governors of its several sub-provinces (sancakbeyi), and a large number of subordinate officers (sipahi) assigned to groups of villages (tıımar) in the sub-districts of each sub-province.^[8] The palace appointed the governors to their positions for a limited term, rotating them from province to province. The palace also approved the governor's appointments of subordinate officers, also for a limited term, rotating their assignments from time to time.^[9] The powers of this hierarchy of military officers was further subject to certain checks and balances. The chief treasury official (hazine kethüdasıı) was responsible for seeing that the fiscal affairs of the province were in order. The chief court official (kadıı) issued judgments in accordance with administrative and religious law. The provincial governor could dismiss both these officials, but only upon notification of the palace.

As we have seen, the military, treasury, and judicial officials were members of a special class without ties to the lands and peoples they governed. Only some members of this special class would actually have been trained in the institutions of the imperial capital, but of those who were not from Istanbul, only some would have been born and raised in the eastern coastal region. So the military, treasury, and judicial officials of the province of Trabzon had no multistranded ties with either the lands or the peoples for whom they were responsible. In this regard, they were unlike the lords and vassals of western Europe during the Middle Ages. They generally lacked castles or estates in the provinces to which they were posted, just as they lacked supporters among their inhabitants.

Furthermore, official procedures insured that the military governors would remain loyal and obedient servants of the central government rather than set down roots and build a following among the populations they governed. The sipahi, who were assigned to the countryside as a kind of country police force, best exemplify how this was accomplished. They came into direct contact with villagers on a routine basis in the course of collecting taxes, apprehending fugitives, imposing forced labor, and carrying out court decisions. They were therefore in a good position to build a base of local support, by favoring or disfavoring the villagers for whom they were responsible. But to counter this very possibility, the sipahi were subject to all kinds of controls.

The provincial governor could renew or revoke their appointment to a tıımar, as well as allow or disallow their relatives to succeed them as asipahi. So both their assignment in the countryside and their tenure as subordinate military officers were entirely dependent on higher state officials rather than local preference or election. Normally, they were rotated from district to district during the course of their service, so that they usually found themselves among peoples with whom they had no previous social contacts. They also had limited opportunity to develop social contacts since they were obliged to report for military campaigns, during which they were replaced by deputies for extended periods.^[10] Treasury and judicial officials also subjected them to periodic inspections and were in a position to bring charges against them for malfeasance. Complaints could be lodged against them by townsmen and villagers and could result in dismissal if confirmed.^[11]

In other words, all kinds of precautions had been adopted to prevent lower state officials from doing exactly what some would do during the period of decentralization. Policies of appointment, rotation, and inspection prevented the sipahi from setting down family lines and building local followings. It was as though the Ottomans had designed the institutions of the classical period in anticipation of the post-classical period that followed it. As we shall see, this would not have been a coincidence. Those who were responsible for inventing and implementing the imperial project would have taken care to neutralize a decentralizing potential that was inherent within its logic.

Four Ottoman registers attest to the thoroughness and efficiency with which the classical system of provincial government was put into practice in the province of Trabzon. Taken once every generation during the first century of incorporation, the registers tabulate the different classes of tıımar assignments, list the names of the sipahi appointed to them, and establish the tax obligations of each household head.^[12] Although the same structure of government would have been operating in other parts of Asia Minor and the Balkans, the province of Trabzon was in certain respects unusual. Initially, as confirmed by the registers, the Ottomans appointed sipahi who had taken part in the conquest. Some of these individuals came from the western coastal region (the province of Canıık), which had been Islamized and Turkicized for many years.^[13] More than a few came from still further away, from other parts of Asia Minor, or even the Balkans.^[14] Thus, thesipahi in most of the coastal valleys of Trabzon would have been even more distant from their charges than was the case in some other provinces. The villagers among whom they resided were a newly subjected Byzantine population composed mostly of Orthodox Christians. The sipahi and the villagers therefore observed different customs, spoke different languages, and belonged to different religions.^[15]

Accordingly, the rural societies of the sixteenth-century province of Trabzon faced a bureaucratic hierarchy whose officials were part of interpersonal networks closed to ordinary townsmen and villagers.^[16] The townsmen and villagers found themselves subject to strangers who represented a foreign fiscal and legal administration. Appointments to military, administrative, and judicial positions were limited to insiders rather than open to outsiders. The entry of ordinary townsmen or villagers into the official class was unlikely. State regulations made it difficult for villagers to leave or neglect their farms. Even conversion from Christianity to Islam may have been discouraged, since every new Muslim household, no longer obliged to pay a tax (haraç) imposed on Christian households, would have had a negative effect on treasury receipts. The policies and institutions of the classical period were therefore incompatible with the peculiar character of the eastern coastal region as described in the preceding chapter.

And yet, as we shall see, the foundation had been laid for the "imperialization" of the rural societies in the province of Trabzon. The Ottomans had incorporated the inhabitants of the eastern littoral as a Christian population subordinated by imperial institutions. But the latter would eventually discover channels into the military and religious establishments that ruled them. In doing so, the inhabitants of the eastern littoral would evolve into a Muslim population participating in imperial institutions. To some degree, this transformation came about as local elites of the eastern coastal valleys took advantage of the breakdown of bureaucratic centralism during the period of decentralization. But this was not the most important process that led to the emergence of ottomanized rural societies in the province of Trabzon.

As we shall soon see more clearly, the official class of the ruling institution subscribed to a discipline of social thinking and practice that instilled qualities of individual behavior and disposition. Reinforced by bureaucratic policies and procedures, the exercise of sovereign power through interpersonal association bound the official class together even as it divided them from ordinary townsmen and villagers. But once individuals of the subject population were in a position to adopt this same "imperial tactic," they would be able to assert their prerogatives within the imperial system, even at the expense of higher state officials. The dissemination of the exercise of sovereign power through interpersonal association, its spread from insiders to outsiders, was then a cause of the period of decentralization. The breakdown and corruption of bureaucratic centralism can be seen as the corollary rather than the cause of this dissemination. To understand the emergence of ottomanized provincial societies in the eastern coastal region, it is therefore necessary to clarify the social thinking and practice of the official class.

The Palace Complex: A Device of Sovereign Association

Soon after the conquest of Constantinople, Mehmet II initiated the construction of great monumental centers on prominent hilltop locations within the city. By the end of his reign (1451–81), a palace complex and a mosque complex, by virtue of their immensity, the numbers of their personnel, and the extravagance of their budgets, had completely transformed the city from a Byzantine into an Ottoman capital.^[17] Each of the two sites had a dual purpose, both organizational and representational.^[18] They were centers for implementing a new ruling institution based on radical measures of centralized government, and at the same time, they were also theaters for displaying this ruling institution in splendor and grandeur. The palace complex and the mosque complex were therefore instruments of edification as well as of government, and as such, designed to confound and enthrall, if not intimidate and terrify.

Two hundred years or so after the time of Mehmet II, as the imperial system slipped into a period of decentralization, the counterparts of these great imperial complexes began to appear in many parts of the eastern coastal region. Aghas residing in government mansions and hodjas teaching in mosque academies were in effect rural versions of the great monumental sites of Istanbul. However, these rural replicas of imperial models mimicked their originals more perfectly at the level of interpersonal association than at the level of ceremony and architecture. Neither the aghas nor the hodjas intended to emulate the courts, domes, ceremony, and protocol of the palace and mosque complex. Rather, they drew upon the ethical underpinnings of the imperial project: the exercise of sovereign power through a discipline of interpersonal association.^[19]

Since the great monumental centers were designed as much for "showing" as for "ruling," their ceremony and architecture can be deciphered to expose these ethical underpinnings. I shall consider each of the two centers in turn, but devoting more attention to the palace than the mosque complex. The relationship of power and religion during the classical imperial period justifies this treatment. As we shall see, the foundation of the imperial project, sovereign power through interpersonal association, required that the palace hold Islam captive.

The Palace Machine

By its placement and appearance, the palace complex is contrived to manifest sovereignty and invincibility, but by a peculiar logic. Two striking architectural features, clearly visible from afar, suggest a coordination of seeing with ruling. The residential quarters of the sovereign are set on a promontory overlooking two continents, Europe to the west and Asia to the east. This vantage point is set behind and above a high fortress wall that surrounds the entire palace complex.^[20] So the sight of the sovereign is represented by an overlook from within an interior. And the association of a personal gaze with a ruling power, already implicit in its elevated perspective, is explicitly represented by a high fortress wall adorned with watch towers and gun emplacements.

The physical structure of the palace complex announces the relationship of gaze and rule not once or twice, but over and over again. Apertures, gratings, and windows are repeatedly coordinated with balconies, overlooks, and towers. Each instance of the relationship symbolizes the sovereign situated within an interior space overlooking an external space occupied by subjects. All these instances of an elevated oversight articulate the personal presence of the unobserved sovereign as the foundation of a world ruling institution.

But exactly how is gaze translated into rule? What is the quality of vision as cause and how is it related to the quality of power as effect? Does the relationship of gaze and rule perhaps have some kinship with the panopticon, as invented by Jeremy Bentham and analyzed by Michel Foucault?^[21] Or should we not anticipate that the articulation of peculiar elements—eye, interior, oversight, rule, exterior—points to a specific language of authority and obedience? The verses of an Ottoman poet, cited in Necipoğlu's definitive study of the architecture and ceremony of the palace, seem to have been composed as an answer to this very question. Cafer Çelebi said:

	What is each window, but an eye opening to the whole world, to watch ceremonies and spectacles?
	What is each tower, but from head to foot a tongue to praise and eulogize the just shah?[22]

The verses of the poet refer to the basic elements of the overall placement and appearance of the palace complex: A personal presence, symbolized by windowed interiors set in tower overlooks, bespeaks of sovereign power. But now the poet composes these elements to articulate a somewhat strange message.

According to the conceit, the palace complex works like a machine that mediates between the eye of its resident and the imperium beyond its walls. Its overlooking windows (set in towers) are telescopes that expose the whole world to his sight. Its elevated towers (pierced by windows) are megaphones that speak to the whole world of his rule. So by this combination of mechanisms, seeing is most certainly linked with ruling. The structures that expose the world to the person of the sovereign are also the structures that represent the person of the sovereign to the world as legitimate ruler. Moreover, the verses also point to the specific qualities of seeing and ruling. The windows are an eye that brings the world into view, but not as a landscape of fields, waterways, mountains, and lakes. The world appears in the form of the subjects of an imperium, engaged in ceremonies and spectacles, and so not engaged, by contrary example, in commerce and warfare. The sovereign therefore sees, or rather oversees, normative performances. The windows, as instruments of the palace machine, bring a kind of world into view, a world of propriety and sociability, pleasing in form and delightful to experience. So the sight of the sovereign, as mediated by the palace machine, has worked to compose the world it brings into view by sponsoring and supporting normative performances. The palace machine is both organizational and representational. Cafer spoke both brilliantly and succinctly as he put the spirit and logic of the palace complex into a score of words.

By the anachronistic terms "telescopes" and "megaphones," I have emphasized how the poet images the palace as an instrument of sovereignty. On the other hand, Cafer's tropes are more anthropomorphic than mechanical. He likens the palace windows to an eye and the palace towers to a tongue. Since these figures clearly do not refer to the sovereign himself, the poet is suggesting that the palace is an instrument of sovereignty that works through anthropomorphic rather than technical media, such as the scripts and documents of a centralized bureaucracy. As we shall see, the palace was indeed an instrument for extending and amplifying the personal presence of the sovereign by means of loyal and obedient officials who were something more than a centralized bureaucracy. This feature of the palace is indicated by another coordinated repetition of structures and functions, one that determined the overall layout of the palace complex (see fig. 7).^[23]

A succession of three gates, leading to three courts, organized the grounds of the palace complex. The outer gate led to the outer court, from which the middle gate led to the middle court, from which the inner gate led to the inner court. This arrangement defined a series of thresholds leading to interiors, each standing in relationship to one another as "further out" or "further in" by reference to the person of the sovereign. At the same time, each instance of a threshold and interior was also associated with apertures and overlooks representing the personal presence of the sovereign, seeing but unseen, hearing but unheard. In this respect, the successive interiors defined by the gates and courts had the status of assembly spaces that were correlated with the increasing nearness, and hence intensity, of sovereign oversight. As the visitor to the palace complex moved through them, he experienced the progressive loss of his own will and purpose as he became more and more subject to ceremony and protocol. In this way, the visitor experienced his approach to the sovereign as an approach to a locus of normative regulations that otherwise radiated outward.^[24] The series of thresholds and interiors therefore indicated the personal presence of the sovereign as a focal point of a world imperium. So an architecture of windows and towers was coordinated with an architecture of gates and courts.

An Omnipresent Personal Oversight

Following the lead of Cafer, let us begin with the Tower of Justice (köşk-i 'adl), a structure he may have had in mind when he composed his verses.^[25] The base of the Tower of Justice was situated in the residential quarters of the sovereign located beyond the north corner of the middle court. The upper section of the Tower of Justice, fitted with a latticed window, rose above the rooftops of the other buildings so that it could be seen from afar. The sovereign was therefore able to mount the Tower of Justice and view the two continents of his imperium from the seclusion of his residential quarters. By its architectural prominence, this symbol of sovereign oversight, only one example of many, dominated the entire palace complex.

A Council Hall (divan) was situated before the Tower of Justice in the forward section of the middle court (see fig. 7).^[26] Here, the highest officials of the ruling institution deliberated state affairs, reached administrative decisions, received petitions from subjects, and heard law cases.^[27] At the base of the Tower of Justice, a little room was equipped with a curtained window that opened out into the upper wall of the Council Hall.^[28] The sovereign could access the little room from the seclusion of his residential quarters and sit behind the curtained window. For all those present in the Council Hall, he was then always present, even when absent. Before the curtained window itself, hanging from the domed ceiling of the meeting room, a gilded globe represented the earth.^[29] Round about the meeting room, large grilled openings at the entrances symbolized the availability of justice to all subjects.^[30]

The window above, which overlooked two continents, was therefore punctuated by another window below, which overlooked the dispensation of justice by state officials. In the early years of his reign, and continuing for sometime thereafter but ever more rarely, Mehmet II had joined state officials in the Council Hall and participated himself in their deliberations. This practice, which was abandoned as the palace complex reached completion, points to the derivative and evolutionary character of the ruling institution. Once upon a time, it would seem, the sovereign had been more directly and routinely involved in the day-to-day matters of administration and adjudication. At a certain moment of dynastic history, one could presume, the device of the windowed overlook had begun to substitute for the personal presence of the sovereign.

The sovereign had ruled by his personal presence in accordance with a kind of justice that was interpersonal in character rather than abstract and technical. With the extension of his domains, however, the sovereign came to rely on his officials as stand-ins for his personal presence so that their eyes, tongues, heads, and feet were his representatives. And then with the further extension of his domains, his officials multiplied, so that they numbered in the thousands rather than the hundreds. The windowed overlook had then been adopted as a device for extending the personal presence of the sultan. Thus, the ruling institution retained its character as an interpersonal association even as it evolved into a world system. This had come about by a resort to an architectural instrument that preserved a founding principle but enabled its imperial amplification.

This "just so story" suggests that the Council Hall had first been fitted with a curtained window, and then a Tower of Justice with a latticed window had been added to the Council Hall. The windowed overlook would have been first adopted as a device for supervising the ruling institution and then only later become a symbol of the ruling institution to be seen from afar.

A Spectacle of Interpersonal Association

In fact, the Tower of Justice also served as an internal device of the ruling institution as well as its external symbol. Its latticed window had been specifically constructed to provide a view of the core mediating segment of the palace complex, the middle court. From the seclusion of his residential quarters, the sovereign was able to mount the Tower of Justice, position himself before this latticed window, and observe an extraordinary scene of "ceremonies and spectacles" consisting of "praises and eulogies." Down below in the middle court, he could observe thousands of loyal officials who had assembled to show themselves to their sovereign, greeting and honoring him as they did so. In all probability, Cafer had these occasions in mind when he composed his verses describing the palace. He had sought to capture the wonder of a novel architectural structure by which personal oversight of normative performances had been translated into a world imperium.

In the early years of his reign, before the palace complex was completely finished, Mehmet II had not always mounted the Tower of Justice to view the assemblies down below. He had regularly joined his officials and troops in the middle court, appearing before them and hosting a meal.^[31] Necipoğlu, relying on the description of a Genoese merchant, gives the following account of these gatherings, which occurred early each day at the break of dawn:

On these occasions the courtyard was filled with eight thousand officials wearing "vests of brocade and silk of every color and type." The sultan appeared in glory under a magnificent portico (lobia) in front of the gate that led to the residential court. The ceremony lasted for a quarter of an hour, during which time servants offered food to the sultan on a gold tray and to those assembled on silver and copper trays, according to their rank. At the end of this banquet, at which not a word was spoken, the courtiers loudly acclaimed their ruler, extolling, praising, and glorifying his name. . . . Ambassadors were made to watch this ceremony of imperial glorification, after which they were conducted . . . to the sultan's seat. After bowing and kissing his hand, they were made to sit . . . until he rose to return to his private quarters. As soon as he stood he was again loudly acclaimed by the soldiers; then he sat and rose a second time to hear another acclamation, before entering his private apartment. After the soldiers left, the dignitaries of the [Council Hall] ate. Only then did [they] listen to the cases presented, which they subsequently reported to the sultan inside the private courtyard.^[32]

The derivative and evolutionary character of the palace complex has once again come into view. Once upon a time, the sovereign had been more routinely present at the assembly of his officials, not only as they dispensed justice, but also as they congregated as a sovereign association. The windowed overlooks of the palace complex therefore appear as architectural replacements of the once-present sovereign.

At a certain moment, it would seem, Mehmet II had chosen to absent himself from his officials, substituting his participation in a sovereign association with the device of a windowed overlook. The intention of carrying out such a replacement can therefore be more or less precisely dated to Mehmet II's decision to build a new palace. It did not immediately follow the conquest of the Byzantine capital.^[33] Mehmet II had first built another palace complex in Istanbul, later to be known as the "old palace" (eski saray). It was only after he had decided to launch an imperial project some years later that he undertook to build what would become formally known as the New Imperial Palace (sarây-ıı cedîd-i 'âmire).

The Dynastic Court Tradition

Given its derivative and evolutionary character, the New Imperial Palace can be best understood in terms of its place in the history of the Ottoman dynasty. I shall here tell a second "just so story."

Osman, founder of the Ottoman dynasty, had been a chieftain of a Turcoman tribe of pastoral nomads in northwestern Anatolia. But during the course of his reign (1280–1324), he had also become a leader with followers in the marchlands between the Byzantines and the Seljuks.^[34] As such, he was the principal figure of a warrior association composed of individuals who came from many places and spoke many languages. This warrior association eventually came to represent a ruling institution legitimized by the extension of the lands of Islam at the expense of the lands of Christianity. So Osman was something other than a chieftain, and his associates were something other than tribesmen. But still, during this early period, they had participated in communal banquets more or less in accordance with the steppe traditions of the Turcomans.^[35]

As successive Ottoman dynasts had fitfully risen in power and stature, these communal banquets continued as part of the dynastic court tradition. However, they gradually lost their tribal character and acquired a more Islamic character. The banquets therefore shifted from an occasion that reinforced tribal customs and habits to an occasion that celebrated the ruling institution as a normative association underwritten by the sacred law of Islam.^[36] These banquets were then not mere expressions of royal hospitality and sociability that supplemented the otherwise separate business of implementing policy and applying justice. On these occasions, the leader appeared as the personal guarantor of an interpersonal association underwritten by the sacred law of Islam. So the banquets of the dynastic court tradition became at the same time disciplinary exercises and legitimacy performances.^[37] The leader and his follower composed a ruling institution of Islam legitimized by the extension of the lands of Islam.

Accordingly, the holding of formal banquets prepared in the royal kitchens eventually determined the design of the grounds and buildings of royal palaces, just as they had also determined the layout of the royal tents during military campaigns. So long before the time of Mehmet II, the Ottoman dynast and Ottoman officials had joined in formal banquets in order to enact the normative performances of an interpersonal association. And long before Mehmet II, such occasions had featured ceremony and protocol for the purpose of impressing, if not astonishing, subjects, participants, and competitors.^[38] The palace complexes that preceded the New Imperial Palace were then likewise governmental residences that provided for a theater of state. They had also included assembly grounds, massive kitchens, and large staffs for this very purpose.

At this point, let us return to the middle court to consider what became of these formal banquets as the New Imperial Palace reached completion.

An Uncanny Discipline

During the later years of his reign, as the palace and mosque complexes were reaching completion, Mehmet II adopted new state regulations (kanunname) that decisively transformed his state system into an imperial system.^[39] Afterwards, he entered the middle court to appear before the personnel of the ruling institution only twice each year, on the occasion of the two major Islamic festivals. Otherwise, he never again joined his officials to host them with food prepared in his royal kitchens. And yet the dynastic court tradition of the formal banquet still continued, but in a new "disarticulated" form.

After the adoption of the new state regulations the sovereign was still the focal point of the ruling institution, but the sovereign himself was more than ever inaccessible, even invisible. He retreated to his residential quarters within the palace, distancing himself even from the members of the official class. Meanwhile, the representatives of his ruling institution became more than ever a centralized and exclusive official class. The highest state officials were more and more often of "slave" status, individuals who had been separated from their families to be raised and trained within the palace. Correspondingly, the dynastic court tradition changed in character to reflect the logic and sense of the new imperial project. The loyal and obedient officials now presented themselves before a sovereign who had come to resemble a divinity, seeing but unseen, hearing but unheard. Four times each week, thousands of officials came to the middle court to manifest the ruling institution as a normative interpersonal association through a display of ceremony and protocol.^[40] Necipoğlu summarizes the intent of these so-called "Victory Councils" (galebe divanıı) in the following terms:

Ceremonial [of the middle court] served to create a visual diagram of this hierarchically organized military state that was immediately graspable at a glance. This diagram accentuated the omnipotence of the sultan together with the transformation of the centralized state into a bureaucracy and a great army at the personal service of the sultan.

The perpetuation of ceremonial communicated a message of timeless order and stability, bestowing permanence and legitimacy on an arbitrary social construct. Its power lay in constant repetition, enacted in an eery silence, as if time had been temporarily suspended by an endless recurrence. It froze time in an eternal present and created the illusion of an order transcending mere human experience. [Italics mine]^[41]

A fragment of the old dynastic court tradition, a warrior leader hosting warrior followers, had been elevated to the level of an imperial project. But in the quote, which tells us a great deal, Necipoğlu has also uncharacteristically slipped past the most important point, which her study otherwise fully documents.

The imaginary potential of the displays in the middle court did not arise from the suspension of time through constant repetition. The witnesses of the ceremonies of the middle court, both insiders and outsiders, had an experience of repetition, but in the register of the uncanny, rather than the eternal.^[42] The imperial theater of the middle court had distilled an interpersonal association, rooted in the dynastic court tradition, into its disciplinary essence. The ceremonial rested on an arbitrary social construct, but only in the sense that it revealed thousands and thousands of individuals whose being had seemingly been seized and compelled by such a construction. It was this conscious display of the unconscious that had lain behind the poet's image of the place complex in terms of body parts, eye (çeşm), head (ser), foot (kadem), and tongue (dil).^[43] The display of a sovereign association had acquired the quality of an "order transcending human experience." The movements of its participants seemed governed by divine ordinances, by their reduction to figures of loyalty and obedience. Such an imperial theater would not have been effective had it been arbitrarily staged and performed by casting directors and professional actors. Its power to seize and compel its witness relied on effects that could have only been induced by radical procedures of recruitment and training.

One did not have to "know" the meaning of the imperial theater in order to be impressed. Both foreign observers and official participants were struck by its representations. Fresne-Canaye, who witnessed one of the Victory Councils in the company of the French ambassador in 1573, noted its disciplinary essence, even as he could not conclude whether the sight was pleasurable, frightening, monastic, natural, civilized, or savage. The Frenchmen had first passed through the outer gate into the outer court where they dismounted from their horses. They then passed through the threshold of the middle gate to enter into the open square of the middle court. Fresne-Canaye describes what they then saw:

At the right hand was seated the Agha of the Janissaries [a high officer of the central army], very near the gate, and next to him some of the highest grandees of the court. The Ambassador saluted them with his head and they got up from their seats and bowed to him. And at a given moment all the Janissaries and other soldiers who had been standing upright and without weapons along the wall of that court did the same, in such a way that seeing so many turbans incline together was like observing a vast field of ripe corn moving gently under the light puff of Zephyr.^[44]

The Frenchmen were then taken to the Council Hall in the forward section of the middle court where they were "courteously received" by the grand vizier and governors. Following this reception, they were then taken elsewhere in the middle court, where they were provided a generous dinner laid out under a portico. As they dined, they observed the officials and troops assembled in the middle court:

We looked with great pleasure and even greater admiration at this frightful number of Janissaries and other soldiers standing all along the walls of this court, with hands joined in front in the manner of monks, in such silence that it seemed we were not looking at men but statues. And they remained immobile in that way more than seven hours, without talking or moving. Certainly it is almost impossible tocomprehend this discipline and this obedience when one has not seen it. . . . After leaving this court, we mounted our horses where we had dismounted upon arrival [in the outer court]. . . . Standing near the wall beyond the path we saw pass all these thousands of Janissaries and other soldiers who in the court had resembled a palisade of statues, now transformed not into men but into famished wild beasts or unchained dogs.^[45]

This representative of a country distant from the Ottoman Empire was both enchanted and frightened, but the nearer neighbors of the Ottomans came away with less ambivalent impressions. When the Safavid prince visited the palace complex in 1591, for example, his attendants were "exhausted" and "weakened" by what they saw in the middle court.^[46]

According to the witnesses of this imperial theater, the crowds of officials did not behave as crowds at all. By the gravity of their movements and the silence they imposed on themselves, they took on the appearance of an exalted assembly under the governance of divine ordinances. The personal oversight of the sovereign was coupled with the enactment of a discipline of interpersonal association. The result was the manifestation of the ethical underpinnings of the ruling institution in the form of a body of loyal and obedient servants.

The Palace as Panopticon

The palace machine does indeed recall the panopticon of Bentham and Foucault. The guarantor of the discipline enacted by loyal and obedient servants was not himself present in the middle court, but positioned instead in the Tower of Justice up above. From behind its latticed window, the sovereign observed the middle court down below. And so by the device of the Tower of Justice, the sovereign retained his officials under his surveillance, even though he might not even be present. The windowed overlook was then a device of omnipresent inspection and regulation serving to inculcate self-control in each servant. But in this regard, there is something odd about the Tower of Justice, Council Hall, and middle court, at least by the measure of the nineteenth-century western European panopticon.

There is no compartmentalization of those subjected to surveillance, one of the most distinctive features of the panopticon.^[47] On the contrary, regulation and inspection are invariably linked with assemblies and associations. The windowed overlook of the palace machine is therefore a device for internalizing a discipline that takes the form of an interpersonal ethic rather than a personal conscience. The eye of the sovereign up above is reinforced by others down below. Sovereign oversight therefore has no need of jails, workhouses, or barracks, because it is supported by the eyes and tongues of others. Sovereign oversight therefore takes the form of support and sponsorship of places of normative association, such as salons, coffeehouse, and mosques, rather than cells, exercise yards, and parade grounds.

The Tower of Justice, Council Hall, and middle court are not the panopticon of Bentham and Foucault because they refer oversight to "ethics" rather than "conscience." They are devices for instilling a discipline of interpersonal association, rather than a discipline of self-control by self-oversight. This is why apertures and overlooks indicate a person located within an interior, rather than a point of technical observation occupied by an anonymous warden. This is why the micro-forms of palace architecture are personalized, referring specifically to the location of the sovereign even in his absence. And this is why the personal presence of the sovereign can be projected through, and so has his counterpart in, the personal presence of others. And this is why the palace, insofar as it is imaged as machine, never escapes anthropomorphization. The windowed overlooks are insistently linked with the sovereign, precisely because he is not a machine, but a person who can be represented by the eyes, tongues, heads, and feet of others.

Disarticulation, Distribution, and Rearticulation

The architecture and ceremony of the palace complex would appear to have reduced the old dynastic court tradition to a disciplinary tactic of loyalty and obedience. The sovereign has left the banquet to become a figure of omnipresent surveillance. The servings of food have been removed, leaving nothing other than ceremony and protocol. The fellowship of the assembly has been transformed into a setting of individuals who have taken on the appearance of lifeless, marble statues. A windowed overlook is all that remained of the warrior leader who once hosted his warrior followers. But in fact, the totality of the old dynastic court tradition had remained in place in the palace complex. Mehmet II had not at all forsaken the old dynastic court tradition in the course of building his palace complex and launching his imperial project. He had rather disarticulated, distributed, and rearticulated it. The architecture and ceremony that supplemented the middle court illustrates this operation.

As we have seen, the layout of the palace complex consisted of a series of thresholds and interiors signaling an approach to the personal presence of the sovereign, but also by a reverse movement, the projection of the personal presence of the sovereign into world at large as an imperium. Tracing this double movement, we discover how the core mediating segment of the palace complex transformed the formal banquets of the old dynastic court tradition into a world empire. I shall start with the innermost approach and then conclude with its outermost projection.

The Inner Gate and Petition Room

When granted audiences with the sovereign, high officials and foreign ambassadors proceeded to the inner gate in order to enter the inner court. As they did so, they moved from the outer palace (birûn), a place that was subject to devices of sovereign oversight, to the inner palace (enderûn), a place that was subject to actual sovereign oversight. The difference appears in the official title of the inner gate: the Gate of Felicity (bâb üs-saâde). The approach to the sovereign is likened to an experience of happiness and delight. In this respect, an element of the scene of the warrior leader hosting his warrior followers—bountiful hospitality and engaging sociality—had been distilled into a figure of the pleasures of paradise.^[48]

But paradoxically, having become an almost god in an almost cosmic setting, the sovereign was now unfit to engage in any form of reciprocity even as he remained a symbol of hospitality and sociability. In the inner court, his servants, forbidden to speak out or to reveal themselves, communicated in signs and hid behind columns. His personal assistants, eunuchs, mutes, and dwarfs, featured physical disabilities. His personal companions—youths who were not men, men without social origins, mothers who could not be wives, wives who could not be mothers—featured status debilitations.^[49] The sovereign, as the fount and origin of an imperial normativeness, could not himself partake in horizontal social engagements, at least by the representations of official ceremony and protocols. The personal presence of a world ruler reduced every other being in his immediate environment to something less than fully human. No one "whole in being" was to be found close or near to him. The principle of sovereign oversight had in effect blasted away the imperial family. From the standpoint of ceremony and protocol, the sultan could not be a father, a son, a husband, a brother, a lover, a companion, or a friend. He could only be an ascendant or a descendant. The crossing from outer to inner palace was not a matter of entering a familial space and time, or a communal space and time. It was architecturally and ceremonially marked in order to symbolize the personal presence of the sovereign (see fig. 7).

As the visiting official or ambassador passed through the inner gate, he stood before the Petition Room (‘‘arz odasıı), where he would soon encounter the sovereign, but he was already under surveillance. The wall of the petition room was pierced by a large iron grating so that the threshold of the inner gate was exposed to the gaze of the sovereign. The visitor had therefore already come within the view of the sovereign sitting on his throne.^[50] The visitor was then brought into the Petition Room, where he was presented to the sovereign in a manner that varied with imperial fashions and policies. The occasion might involve an exchange between the visitor and the sovereign, a ritual greeting and welcoming, but this was not the usual formality. More commonly, it was an exchange without exchange, submission without recognition, supplication without acknowledgment, or sentence without trial. The official or ambassador entered a room whose walls and carpets were encrusted with gold and jewels, in accordance with a solemn ritual, conducted in "the very silence of death itself."^[51] The sovereign sat upon his throne, unflinching and immobile.^[52] The ambassador, with eunuchs gripping both his hands, was then taken down to the knees of the sovereign to kiss his robes at his feet.^[53] The official, standing before the sitting sovereign, might hear a clap of the hands, whereupon mutes might suddenly appear to carry out his execution outside the Petition Room before the grating.^[54]

And yet the personal presence of the sovereign, in whose vicinity no man could compose his own will and purpose, had to be the foundation of a sovereign association that could be projected as a world rule. This was precisely the role of the palace machine, a monster of heads, feet, eyes, and tongues constituting the person of the sovereign as a world imperium. By reversing direction, in order to move from Petition Room to the middle court, we can follow the path whereby personal presence was projected as world imperium.

The Petition Room is topped by a low roof in the imperial style, which extended beyond the walls of the structure itself. As in the instance of the coupling of an interior with an overlook, this feature of the roof symbolizes the projection of the sovereign's oversight of an interpersonal association into the world at large. The roof both covers an interior and extends beyond the limits of this interior. The symbolism is explicitly articulated by a feature of the overhanging roof that draws the Petition Room into a relationship with the middle court. On that side of the Petition Room facing the inner gate, the roof does not come to an end but penetrates the boundary wall in which the inner gate is placed. On the other side of the boundary wall, this same section of the overhanging roof reappears as a portico extending over and before the inner gate itself. Here, under a cupola set in the forward part of the projected roof, the sultan would sit on a throne to witness the ceremonies in the middle court on the occasion of the two annual Islamic festivals.^[55] Otherwise, the throne and cupola served as symbols of the sovereign on the occasions of the Victory Councils four times each week.

So the inner gate represented an approach to an interior overlook (room and grating) even as it also represented the projection of an interior overlook (court and cupola). The micro-forms of palace architecture—the room and grating and the cupola and court—had disarticulated the old dynastic court tradition, then rearticulated it to link a sovereign, who was absent rather than present, with a sovereign association. However, the palace machine has not yet completed its operations. The placing of the inner gate before the Petition Room has its counterpart and complement in the placing of the middle gate before the middle court.

The Middle Gate and Middle Court

More than any other architectural boundary within the palace complex, the middle gate represented a transition from outside to inside. As such, it revealed what was composed on the inside (an interpersonal association) and how this composition appeared on the outside (the sovereign power of a world ruler).^[56]

The visitor arrived in the outer court crowded with the attendants and horses of all the officials and ambassadors who were to enter the middle gate and take part in the ceremonies of the middle court.^[57] Approaching the middle gate, he was presented with what appeared to be a medieval stronghold (see fig. 7). The entrance was flanked by two stone towers topped by parapets and surmounted by a crenellated wall that joined the flanking towers together. The threshold itself was staffed by military officers and soldiers of the central army and festooned with banners, weapons, and armor commemorating imperial victories.

The external facade of the middle gate therefore represented the fortress of a ruler, and, in so doing, repeated the statement made by the watchtowers and gun emplacements of the high stone wall that surrounded the entire palace complex.^[58] This restatement asserts the linkage between military force and victory on the outside with the discipline of interpersonal association on the inside. Accordingly, the threshold itself articulated the relationship force and ethics. It was the site of a court where two military judges (kazasker) tried military officers accused of malfeasance.^[59]

In effect, the external facade of the middle gate is a remainder of the old dynastic court tradition of a warrior leader hosting his warrior followers. Now, however, all the symbolism of military capacities and formations has been exiled from the middle court by the logic of a disarticulation, distribution, and rearticulation. So passing through the threshold of the middle gate into the interior of the middle court, the visitor would discover the ethical "substance" behind the military "facade." The official name of this entrance, the Gate of Greeting (bâb-üs-selâm), contrasted with the official name of the Gate of Felicity behind it. The warrior leader hosting his warrior followers had been broken into pieces, consisting of military force and power (middle gate), ethical discipline (middle court), and hospitality (inner gate).

After the visiting official or ambassador passed through the middle gate, he was received under a "stately portico with ten marble columns attached to the gate's inner facade."^[60] Before the portico itself, he saw a garden "paradise of peacocks and gazelles, cypress trees and other trees."^[61] The visitor had left behind the crowds of attendants and horses in the outer court. He had also left behind the representations of the military force and power of the ruling institution. He now found himself welcomed by an assembly of officials whose dress and movement were regulated by ceremony and protocol. The inner gate represented the personal presence of the sovereign as the guarantor of a discipline of interpersonal association. The middle gate represented the discipline of interpersonal association as an instrument of sovereign power in the world at large. The relationship of inner, middle, and outer courts was then a relationship of gaze, discipline, and rule.

The Never-Ending Banquet

And yet there would still seem to be one missing piece of the old dynastic tradition in the middle court. Mehmet II had seemingly suppressed the banquet itself, when the sovereign hosted his officials to food and drink. But this is not the case. The banquet had not disappeared at all, but it had rather become routine and perpetual.

I have already noted that the Tower of Justice rose prominently above the other buildings within the outer fortress wall. Similarly, twenty paired chimneys of the Imperial Kitchens had been designed as prominent marks on the skyline.^[62] To the left of inner gate, the Imperial Kitchens comprised the transverse border of the middle court. Here thousands of "graded" meals were prepared each day in "graded" kitchens, for the sovereign, the staff of the inner palace, high officials of the ruling institution, and, finally, military officers and regiments.^[63]

With the building of the palace complex, the warrior leader hosting warrior followers had become a master leader hosting slave followers, while the banquet had become an everyday routine but no less charged as a symbol of loyalty and obedience. On campaign, military regiments were brought to shame by the loss of their cooking pots to the enemy, not the loss of their flags and banners.^[64] And similarly, military officers and troops expressed their displeasure with the sovereign by overturning the soup cauldrons where their meals were regularly prepared.^[65]

The elements of the dynastic court tradition had remained in place, as a warrior leader of warrior followers became the early modern sultan of a world empire. The symbol of personal presence (window), a normative oversight (tower), an interpersonal association (kitchens), and invincible sovereignty (fortress wall) were all the tropes of palace architecture as seen from afar. Each consisted of a disarticulated micro-form of a warrior leader hosting warrior followers. The elements were enduring because they represented something more deeply rooted in thought and practice than ceremony and protocol.^[66]

We are now ready to accompany the sovereign on those occasions when he ventured from his residence in the company of his personnel. Where did they go and for what purpose?

The Mosque Complex: A Device of Official Islam

So far as I have been able to determine, the ceremonies of the middle court did not include an occasion in which the sovereign appeared as a military commander leading a military formation.^[67] Contrary to what one might expect—given the origins of the dynastic court tradition in an association of warrior leader and warrior followers—the Victory Councils represented the ruling institution as triumphant and invincible by virtue of its ethical underpinnings, rather than military organization or technology. However, the sovereign did routinely lead a procession of officials from the palace complex in parade formation. On such occasions, he came to the middle court where his attendants had brought his horse from the Imperial Stables opposite the Imperial Kitchens (see fig. 7). He mounted and rode through the middle gate in the company of officials and attendants on foot. In the outer court, he then joined a larger formation, which included both horsemen and footmen, and exited through the outer gate. They were on their way to perform the Friday prayers (namaz kıılmak) in a great mosque of the imperial capital.^[68]

Having arrived at their destination, sovereign and officials would perform their ablutions, preparing their minds and bodies for the religious observance. This done, they would enter the great mosque and assemble themselves in ordered ranks, the sovereign occupying a separate elevated compartment. Having oriented themselves in the direction of Mecca, the constituents of the ruling institution then performed the prayers, enacting a discipline of mind and body, in accordance with required formulas of speech and gesture. The staging and performance of a normative association before the sovereign, unseen and unheard, was therefore punctuated by the staging and performance of a normative association before divinity, unseen and unheard.^[69]

Is it far-fetched to draw a parallel between sultan/palace and divinity/mosque? Is this merely a kind of free association of a ruling discipline with a religious discipline? Not according to western European observers of these ceremonies and protocol. Fresne-Canaye had noticed the parallel in his account of his visit to the Petition Room. As he waited his turn, he had observed high officials groom themselves before making their appearance before the sovereign: "And I saw the pashas . . . standing up straight and combing their beards with their hands: because they consider it meritorious to appear well groomed before God as they perform their prayers, and also in the presence of their emperor."^[70]

And he had noticed the parallel again precisely on the occasion of the sovereign's procession to Beyazit Mosque in order to perform the Friday prayers:

With the passage of the Great Lord, there was everywhere an extraordinary silence. One could say that his very look had the power to transform men, like Medusa, into marble or into mute fish; because they have the conviction that their lord is the shadow of the breath of God on earth, having learned nothing from their youth in the palace save obedience and respect for their emperor. And by that unique discipline, they are always going to increase their power, to the shame of all Christians.^[71]

The impressions of foreign visitors are of course notoriously unreliable. However, Fresne-Canaye appears to be well informed. High Ottoman officials had not explained their thoughts and feelings to him, as his citations imply. Nonetheless, the ghost of Mehmet II could have been his "literal" informant, his calligraphers having left his "personal signature" all over the palace complex.

Two gilded inscriptions appear over the portal of the outer gate leading to the outer court. Necipoğlu is once again our guide to their significance. The lower of these is a foundation inscription that brings to mind once again the poet's verses. The solidity of the palace complex as an architectural structure is said to be the guarantee of social order and calm. Here, however, this accomplishment is attributed to a sultan who is both a substitute for and a favorite of divinity:

By the help of God, and by His approval, the foundations of this auspicious Castle were laid, and its parts were joined together solidly for strengthening peace and tranquility, by the command of the Sultan of the two Continents, and the Emperor of the two seas, the Shadow of God in this world and the next, the Favorite of God on the Two Horizons [East and West], . . . may God make eternal his empire, and exalt his residence above the most lucid stars of the firmament, in the blessed month of Ramadan of the year 1478/883. ^[72]

The sovereign does more than acknowledge Islam. The palace is more than linked with Islam. The sovereign and his palace are a paragon of divine will and favor. When you approach the sovereign by entering the palace complex, it is as though you have crossed into the circle of divinity.

The upper inscription tells us just this. It is a Koranic quote that likens the threshold to the Gates of Paradise and the interior to the Garden of Eden:^[73]

But the God fearing shall be amidst gardens and fountains: "Enter you them, in peace and security!" We shall strip away all rancour that is in their breasts; as brothers they shall be upon couches set face to face; no fatigue there shall smite them, neither shall they ever be driven out from there. [Italics mine]^[74]

The inscription marks the portal of the outer gate, which is set in a surrounding fortress well. It stands over another inscription, which draws a relationship between the material structure of the palace complex and the eternal endurance of a world empire. Nonetheless, the material structure, that is to say, palace architecture, is but the frame or setting for an interpersonal association. You know you are crossing into the circle of sovereign power, the very shadow of divinity, when you encounter within these precincts the peace and security of a brotherly association and affection.

The palace was then explicitly analogous to the mosque. The throng of officials in the middle court assumed the quality of angels compelled by divine ordinances. The setting in which they were assembled seemed as though it was a landscape from paradise.

The sovereign himself was represented as the epiphany of a divinity. The architecture and ceremony of the Tower of Justice, Council Hall, and middle court were directly associated with, if not embedded in, the legal tradition of Sunni Islam. In the middle court itself, the ceremony and protocol of a sovereign association, its disciplined gestures and movements, were the counterpart of other disciplined gestures and movements, the ablutions and prayers. The sovereign and his palace were then affiliated with, even dependent upon, Islam. The former manifested and celebrated the ethical underpinnings of a world empire. The model for a discipline of interpersonal association was Islamic. The basis of the loyalty and obedience of the official class was Islamic. The legitimacy of such a world empire before all potential competitors within and outside the Ottoman Empire was based on its status as a paragon of Islamic authority and obedience.^[75] Or is this exactly the way to express the relationship of power and religion? Did Islam have a claim on the palace? Or did the palace have a claim on Islam?

Mehmet II had built a mosque complex at a distance from the palace complex nearer the center of the capital city. As in other mosque complexes that would be built by his successors, its principal architectural feature was a great mosque. By its size and aspect, this mosque complex rivaled the Hagia Sophia, the still-existing basilica of the Byzantine emperors (but now a mosque memorializing the Ottoman conquest). Many observers have claimed that all the great mosques of the classical period seemed imitative and derivative, and, consequently, less effective as sacred buildings than the Byzantine "original." But as Aptullah Kuran has insightfully pointed out, the Ottomans had transformed Byzantine tradition, even as they drew from it, arriving at a distinctive and original architectural statement by means of a structure that is only at first sight similar to the Hagia Sophia.^[76] What follows is my own elaboration of his conclusions.

From the outside, the great mosques of the classical period have a squat, frog-like aspect such that they appear to be among the most inelegant, if not unattractive, examples of mosque architecture in the Islamic world. But their ponderous arches and segmented domes are the price to be paid to achieve a specific representational objective. From the inside, the great mosques consist of an immense, virtually unobstructed interior space sheltered by a high dome.^[77] The transparency and homogeneity of the interior represent a universal time and space for the assembly of Muslims who came from different lands, speak different languages, and live by different customs. The high dome that floats magically above this interior space confirms the role of sovereign power as the sponsor of this assembly.

The palace complex was therefore legitimized by a mosque complex. The sovereign was fit to govern all lands and peoples because his ruling institution represented a standard of ethical thinking and practice of cosmic import, valid for all times and places. However, such a reading of the great mosque is slightly askew, more in tune with popular contemporary interpretations of the classical Ottoman mosques in the Turkish Republic than with the intentions of Mehmet II.^[78]

Perhaps Mehmet II built his mosque complex in acknowledgment of the one and true world religion. Perhaps the vast domed interior of its great mosque could be seen as a place of worship for all the faithful of his imperial domains. But even if these points are conceded, and they are far from obvious, the sovereign also built the great mosque complex to stake a claim on Islam, to subject it and so command it.

The sovereign in his palace was entangled with a specific dimension of Islam, not the totality of this religious tradition. The ruling institution had a direct interest in the sacred law of Islam, which is a piece of Islam, and, however important, not all of it. And more exactly, the ruling institution had a direct interest in the legalistic and juridical side of the sacred law of Islam, which is a piece of the law, and, however important, certainly not all of it.

The Ottomans had invented and constructed mosque complexes (imaret, külliye) long before Mehmet II conquered the Byzantine capital. They were the means by which an Islamic dynasty had supported and propagated Islam among lands and peoples with Christian rather than Muslim majorities. They included schools, academies, dormitories, hospitals, kitchens, baths, shops, and warehouses situated in the immediate vicinity of the great mosque.^[79] They served as "urban renewal projects" designed to created Muslim cores of what had been Christian towns and cities.

However, Mehmet II had set about to build a different kind of mosque complex just as he set about to build a different kind of palace complex. Now, the buildings and institutions of the mosque complex, together with their accompanying positions, salaries, meals, housing, and endowments, would become, more than ever before, the resources and instruments of a religious officialdom representing an official Islam. By this intention, the mosque complexes of the classical period served to support, hence to elevate, the learned class of Sunni Islam, but at the same time served to define the membership and formalize the activities of the learned class of Sunni Islam.

The judges of the imperial courts, drawn from the learned class, represented the first tier of officials of the religious establishment. The professors of the imperial academies, drawn from the learned class, represented the second tier of the religious establishment. The ladder of salaried appointments that led to the highest professorships was topped by another ladder of appointments that led to the highest judgeships. This meant that the representatives of the state legal system had precedence over the representatives of the broader religious tradition. And this meant the broader representatives of the religious tradition were themselves oriented to the state legal system. So the judicial hierarchy was positioned over the academic hierarchy. The courts of the religious establishment placed an emphasis on that very part of the religious tradition that could best serve as the basis for the imperial legal system. And accordingly, the academies also placed an emphasis on that side of the sacred law of Islam that was most compatible and consistent with its role in the imperial legal system.

The gigantic domed interior was therefore the symbol of a centralized and exclusive religious officialdom, not the symbol of a single humanity united by a single true revelation. Accordingly, the gigantic domed interior was a place where the sovereign and his officials regularly assembled to subject themselves to Islam during the Friday prayers. But it was also routinely a site for the display of a religious establishment by which the sovereign had subjected Islam. Judges and professors, joined by attendants and students, regularly assembled in the interior beneath the dome, displaying themselves as the representatives of an official Islam. The assemblies of religious officials that appeared beneath the dome of the great mosque were then the counterpart of the assemblies of military officials that appeared in the middle court. The performance of authority and obedience in the mosque complex served as the foundation of the performance of authority and obedience in the palace complex. But by this logic, the palace had to hold Islam captive. If an Islamic discipline of interpersonal association was the basis of sovereign power, then the launching of an imperial project required that the palace complex had to command and control the mosque complex. The ruling institution of the classical period was subjected to Islam, but it was also the case that Islam was subjected to the ruling institution.

State System and State Society

Machiavelli had both understood and misunderstood the ruling institution of the new Ottoman Empire. The ministers of the Ottoman sultan, "being all slaves and bondsmen," were loyal and obedient servants of their master, but lacking any roots among the subject population, they could not "carry the people with them." Such a figure, that of a slave official, points to the exercise of sovereign power through an interpersonal association, but mis-identifies an "ethic" as a "status." The actual Ottoman slave official (kul), a product of specific measures of recruitment and training, took his place in a centralized bureaucracy. All the personnel of this centralized bureaucracy, whether kul or not kul, were set apart from ordinary townsmen and villagers as an official class. The ethical underpinnings of this centralized bureaucracy, which constituted a tactic of sovereign power through interpersonal association, were otherwise fully transmissible, from inside to outside the imperial system. They could be the basis for training and recruiting segments of the subject population when the imperial system was in need of manpower and resources. And accordingly, they could also be appropriated and adapted by segments of the subject population determined to colonize the periphery of the imperial system. The classical imperial period therefore harbored a potential for dissemination that entirely escaped the attention of Machiavelli, as well as other western Europeans, right down to the twentieth century.

As we have seen from the outset of this chapter, the palace organized provincial government by relying on principles of bureaucratic centralism, that is to say, lower officials reporting to higher officials, ruling officials reviewed by judicial officials, and so on. However, it is now evident that these same principles of bureaucratic centralism were the rationalized and legalized expression of an imperial tactic: the exercise of sovereign power through a discipline of interpersonal association. The governors (beylerbeyi), the sub-governors (sancakbeyi), and their subordinate officers (sipahi) were the eyes, tongues, heads, and feet of the palace machine. In effect, they represented the body parts of the personal presence of the sovereign, the governors by reference to the sub-governors, the sub-governors by reference to the subordinate officers, and the subordinate officers by reference to ordinary townsmen and villagers. So each member of the official class was both ruler and ruled.

But this means that identification with and participation in the imperial system were not limited to the official class as such. Each governor, sub-governor, and subordinate officer was after all always more than a singular, isolated individual. He was a father of children, a master of a household, a relative among relatives, a friend among friends, a partner among partners, and a patron among clients. That is to say, each took his place in a world of nonofficial as well as official associations. So each was the representative of a discipline of interpersonal association as father, master, relative, friend, partner, and patron. And given that each was the subject of a discipline of interpersonal association, insofar as they were members of the official class, would it not also follow that adherence to such a discipline would also shape the ethics of families, households, patronage, kinship, partnership, and patronage? The state machine, a tiered hierarchy of bureaucratic centralism, was then conjoined with a state society, a tiered hierarchy of interpersonal associations. And moreover, the associational dimension of the regime always exceeded its official dimension. And by this fact, the associational dimension of the regime always exerted a constant pressure on its official dimension. Whenever the system of imperial regulations allowed some leeway, who would be favored and who would be disfavored? It would not be possible to answer this question without a thorough understanding of the system of interpersonal associations. And this being the case, to what extent could one be sure that the system of imperial regulations was not itself bent or warped, even corrupted and subverted, by the system of interpersonal associations? This very problem explains the precision and elaboration of the imperial system of regulations. It was built as a kind of dike against networks and connections that suffused the official class. And it was a dike that always leaked.

The problem of decentralization was then in place before, during, and after the classical period. The slave official had been conceived to contain and control such a tendency. The children of a subject population who became slave officials were outsiders to existing official circles of interpersonal association (theoretically if not in practice).^[80] They were therefore recruited and trained in the Imperial College of the palace, married to women trained in the Imperial Harem, and then posted to governorships and sub-governorships. In this manner, the palace machine seeded the further reaches of the imperial system with creatures of its making, with individuals embedded in and compromised by official circles of interpersonal association. The eyes, tongues, heads, and feet of these individuals could therefore be relied upon to reconstitute the personal presence of the sovereign sitting in his tower, looking from a window, viewing the two continents of a world empire.

The palace machine was designed to hold in check the forces of decentralization that could only accumulate with the expansion and extension of the ruling institution. But the palace machine accomplished something far more general and enduring than what Machiavelli had estimated. By the work of Mehmet II, the exercise of sovereign power through a discipline of interpersonal association had been more perfectly and elegantly expressed as world rule based on a cosmic law. There was then, in principle, no limit to the number of individuals who might participate in the state society of the imperial system, only a limit to the efficiency and effectiveness of its bureaucratic centralism.

The Period of Decentralization in the Province of Trabzon

As the central government grew weaker during the eighteenth century, provincial elites in the core Ottoman provinces assumed the status of sovereigns ruling extensive domains, passing their wealth and property, as well as their official titles and appointments, to their descendants.^[81] The provincial elites lived in spacious mansions with many servants, including both male and female slaves. Their residences, often sited on promontories or hilltops, took the form of semi-fortified mansions (konak), which were like government buildings with receiving rooms, accounting offices, and prisons. They carried out all kinds of government functions, collecting duties and taxes, apprehending fugitives and enforcing court orders, assembling irregular troops, requisitioning military supplies, and maintaining roads and bridges by corvée labor. For this purpose, they maintained a certain number of administrative staff, and they moved about with retinues of armed supporters.

Since they were sometimes able to place tens of thousands of troops in the field, the provincial elites were in a position to resist, and even defeat, the central armies of the imperial system.^[82] The Ottomans, that is, the state officials representing the palace in Istanbul, were therefore obliged to work with or against them, sometimes granting them titles and appointing them to offices, sometimes declaring them rebels and sending troops against them. The provincial elites were then both inside and outside the official class. As sovereigns of their domains, the provincial elites sometimes defied the central government, but they sometimes implemented imperial regulations and contributed troops to imperial campaigns. In most of the core Ottoman provinces, both higher and lower state officials were unable to carry out the most elementary governmental functions without their assistance. Accordingly, the Ottomans styled them in different ways at different times, for example, as "lords" (derebey) or as "usurpers" (mütegallibe), in accordance with an intent to accept or to question their legitimacy.^[83] In many places, the provincial elites had risen to prominence by amassing vast farming estates worked by large peasant populations. They had done so in many cases by manipulating the tax-farming system, which had replaced the older system of military appointments and assignments. Acquiring the right to collect and forward tax-receipts, an individual could gain control over large tracts of land as well as large numbers of villagers who inhabited them.^[84]

The local elites of the eastern coastal districts were of a different sort. Unlike their counterparts elsewhere, they did not own estates worked by peasants. Instead, they were able to assert themselves as sovereign powers—against their competitors or even against state officials—by their ability to mobilize large numbers of armed followers. But here, the foundation of their sovereignty, that is, large numbers of armed followers, has to be distinguished from the quality of their sovereignty. They were not gang leaders or warlords. Some of them appeared as such at least some of the time, and state officials and foreign consuls described them as such when they gave them trouble rather than assistance. They were also not exactly leaders of military formations, although they certainly manifested themselves in this form from time to time. More exactly, the local elites in their great mansions had a close kinship with the sultan in his palace. They were the principal figures of ruling associations based on a discipline of social thinking and practice. But unlike the sultan, their ruling associations were not closely linked with ceremony and protocol. The local elites in their great mansions were at the center of circles of interpersonal association. These circles were not formally constituted as corporate groups but rather as social networks. They were composed of agnates, affines, servants, friends, partners, and allies. They included a significant fraction of the rural population, even all of the rural population of Muslims, in theory if not in practice. In this respect, the makeup of these circles of interpersonal association directly reflected the largely Muslim makeup of rural society.

The local elites of the province of Trabzon were, much like the sultan in Istanbul, the overseers of circles of interpersonal associations. They maintained salons (oda) in their mansions where they received guests and visitors, and they sponsored coffeehouses (kahve) patronized by large numbers of partners and allies.^[85] At the same time, they were wary of any signs of ambition or rivalry among their relatives or followers. They did not allow anyone within their domains to build new houses, to add rooms to existing houses, to open coffeehouses, or even to build mosques without their permission. Each of these steps could be a first move toward building local support and influence.

Since the sovereign power of local elites, no less than that of the sultan, was linked with a discipline of interpersonal association, they aligned themselves with a kind of Islam much like official Islam. Local elites arranged for the building of mosques in the market centers subject to their control.^[86] They struck partnerships with representatives of the religious professors and students, even to the point of subsidizing them. With the assistance of these representatives, they took the liberty of intervening in the affairs of their followers, arranging marriages and settling disputes in the name of the sacred law of Islam. At the same time, they did not tolerate religious teachers and students asserting themselves as independent authorities in towns or villages.

When their collective interests were threatened by a provincial governor, the local elites of Trabzon joined in coastal coalitions that were able to raise thousands of men in arms and move them both by land and sea. But they were not generally interested in bringing down the sultan or the Empire. They were themselves the creatures of the imperial system. Having adapted and appropriated the imperial tactic of sovereign power, they had a stake in imperial legitimacy, that is, in official Islam. When they rose in revolt, they did so in order to defend or to extend their privileges and prerogatives within the imperial system. As a consequence, they could be partisans of the sultan and Empire, responding to call-outs for imperial campaigns, even as they also raised men in arms in order to force state officials to grant them titles and offices.

When French and British consuls arrived in the town of Trabzon during the first decades of the nineteenth century, they had no understanding of how the local elites of the coastal valleys had grown within the imperial system. Seeing that they were able to defy and threaten the governor of the province of Trabzon, the consuls concluded that the local elites were part of a "feudal system," an alternative political system entirely distinct from the central government.

Fontanier was among those consuls who so described the local elites; however, he eventually came to understand them in ways that directly belied his prejudices.^[87] On two occasions, he inadvertently contradicted himself. The local elites were not the principals of a feudal system exactly like that of thirteenth-century Europe. He noticed instead that they bore an eerie resemblance to the sovereign in his palace.

In the first of the two passages, Fontanier was not describing the local elites but rather belittling the Ottomans. Momentarily "off-guard," he asserts that the sultan, the provincial governor, the district sub-governor, and the agha are all similar to one another:

As we are ordinarily inclined to make judgments by analogy, we might guess that the Ottoman Empire is run by ministers with specifically defined abilities who are able to take the department assigned to them in whatever direction they might wish [as was supposedly the case in the existing monarchy in France]. This would be a serious misperception, because a ministry is nothing but the organization of each family applied to the state. Thus, the agha of a village, the bey of a district, and the pasha of a province all have their house set up exactly like that of the sultan, so that they are surrounded by officers who fulfill the functions analogous to those of the ministers. Just like the Great Lord, they have their own steward, judge, treasurer, etc., whom they choose from among their relatives or their friends. It's just the same in the case of the Imperial Divan itself. Accordingly, it is clear that the ministers are merely domestic servants without any particular standing in their own right whose power depends solely on the favor of their master. [Italics mine]^[88]

Having forgotten the reversion from republic to monarchy in France, Fontanier condemns the Ottoman Empire because of a confusion of the official with the familial. At every level of political authority, sultan, pasha, bey, and agha, the family is applied to the state. However, in making this point, he recognizes that the ministers in question are not really members of a family at all, but rather "relatives" or "friends," and by analogy with the palace, one might add, "servants" or "slaves." So Fontanier has actually pointed to a specific tactic of linking sovereign power with interpersonal association. Furthermore, he has noted that this specific tactic was characteristic of every level of political authority, not just the sultan, but also the pasha, the bey, and the agha. That is, it was a feature of both officialdom and nonofficialdom.

In the second of the two passages, buried in a later chapter of his second book on the Ottoman Empire, Fontanier again qualifies his earlier expressed opinion that the aghas of the eastern coastal region represented a feudal system. The passage occurs once again at a moment when he is "off-guard" as he considers the specific character of the local elites in the district of Sürmene.^[89] Once again, he contradicts himself as he draws an invidious comparison; for he now judges the imperial system to be even lower in stature and quality than the feudal system of western Europe:

The mode of administration [in Sürmene] was more or less the same as in Anatolia. The inhabitants put themselves under the protection of those whom they supposed to be wealthier or more powerful. This would have been a feudalism completely like that which once existed in Europe, if such a patronage were to accord positive rights, and if children were able to inherit power from their fathers. But far from that, the lord who is called a "derebey" [valley lord] is nothing more than the chief of a community. He can do whatever he likes when it comes to tormenting foreigners or pillaging neighbors, but his power over those he administrates is very limited. The democratic element is that on which the society rests; such that the society is nothing more than a collection of little republics. These are more or less advanced in civilization in accord with whether they have been more or less favored by geographic, political, or commercial circumstances. [Italics mine]^[90]

Here Fontanier recognizes that the agha is "nothing more than the chief of a community," but the phrase substitutes for an omission in his analysis. He never writes about the vertical and horizontal solidarities in which the local elites were positioned. He does not take the trouble to point out that the community in question was not ethnic or tribal in character. But he understood nonetheless that the community in question was a kind of "society" constituting a sort of "republic." He even understood that the community in question was a protean one that could rise to the level of "civilization" if it were favored by "geographical, political, and commercial circumstances." Here, he could have pointed directly to the sultan in his palace. The failure in his analysis arises from a blindness. He is unable to recognize a distinctive imperial tactic: sovereign power through interpersonal association.

By the eighteenth century, the ethical underpinnings of the ruling institution had been disseminated among the populations of the core Ottoman provinces. The local elites of the eighteenth-century province of Trabzon were far removed from the ceremony and protocol of the classical period. And yet key features of the imperial tactic displayed by the middle gate, Tower of Justice, Council Hall, middle court, inner gate, and Petition Room reappeared among them. But it was not only the local elites who had assimilated the ethical underpinnings of the imperial system. Unlike elsewhere in the core Ottoman provinces, the rise of the local elites of Trabzon had been contingent on the assimilation of the imperial system by the provincial population, hence the ottomanization of the rural societies of the eastern coastal districts. So a classical imperial system—which Machiavelli had appreciated as a remarkable example of exclusivity and centralism—had resulted in the formation of "a collection of little republics . . . more or less advanced in civilization," as Fontanier was to put it.

Notes

5. Dissemination

Soldiers and Students

Ethnic Diversity and Imperial Homogeneity

In this chapter, I examine the district of Of as a case study in the emergence of an ottomanist state society in the province of Trabzon. Documents transliterated and summarized by Hasan Umur, the local historian of Of, track three different phases of such a process: the Islamization of the district by conversion and immigration, the spread of soldiering and preaching, and the rise to prominence of local elites.^[1]

Who are the Oflus? The question has often been posed by outsiders, and their answers are bewildering. The Oflus are Lezghis (Evliya Çelebi, a late-seventeenth-century Ottoman traveler who confused them with a remote Caucasian people). The Oflus appear to be Muslims but secretly subscribe to Christianity (an early-nineteenth-century French consul).^[2] The Oflus are not like the Rizelis and Sürmenelis, but have distinct habits and customs (an early-nineteenth-century British consul).^[3] The Oflus are a melange of Byzantine peoples like the Rizelis and Sürmenelis (a late-nineteenth-century British consul).^[4] The Oflus are of Pontic Greek origin, but became fanatical Muslims (a late-nineteenth-century scholar of Greek dialects).^[5] The Oflus are Muslims who keep Bibles, crosses, and other relics and would like to become Christians (a twentieth-century Greek churchman).^[6] The Oflus are Çepni Turks who settled in the district sometime after its incorporation by the Ottomans (a twentieth-century Turkish historian).^[7] The Oflus are Laz, like all the other inhabitants of the coastal region (the villagers of the interior highlands).

Despite all the confusion of outside observers, most of the Oflus I encountered in the district during the 1960s did not hesitate to say who they were. They were Turkish Muslims. This contemporary self–identification can be dated to the beginning of the nationalist period, but it is consistent with a much longer history of state participation. By the late seventeenth century, the district of Of had become a predominantly Muslim rural society and many, if not most, male Oflus were affiliated with imperial military and religious institutions. The claim to be both Muslims and Turks in the 1960s can therefore be regarded as an updating of this ottomanist legacy. As Muslims and Turks, the Oflus represented a local state society that had come into being through a complex process of conversion, immigration, and transformation. Since the traces of this history had been preserved in local habit and custom, the identity of the Oflus, like that of many other rural peoples in the eastern coastal region, has long been available for all kinds of polemical purposes. In the next section, I shall illustrate all the moving, remaking, and mixing by contrasting the distinctive features of the population in different sectors of the district. In this way, undercurrents of ethnic and linguistic diversity will serve to highlight the homogenizing process of imperial identification and participation.

Two Valley-Systems

The district of Of is geographically divided into two systems of valleys, that of the Solaklıı River and that of Baltacıı River (see map 1). The villagers of each system of valleys interact more among themselves than they do with the villagers of the other valley–system. The major market centers of each valley are different, and the routes of seasonal movement up and down the valley are different.^[8]

According to local tradition, which is probably correct, the large majority of the population in the district of Of has been Muslim since the mid- to late seventeenth century. Regional historians, westernEuropean travelers, and government records from the mid-nineteenth century all agree that the population was as much as 98 percent Muslim at that time.^[9] Government documents from earlier centuries strongly suggest that such a situation came about no later than the close of the seventeenth century.^[10] So almost all the inhabitants of the district have been Muslim for some time.

On the other hand, the villagers in different sectors of the district have a mix of ethnic backgrounds. The population in the eastern valley-system is (and, in local memory, has long been) almost exclusively Turkish speaking. The population in the western valley-system recently included a large number of Greek-speakers and was not so long ago largely Greek-speaking.^[11] This contrast correlates with the transit systems of the two valleys (see map 1). The eastern valley was open to settlement by peoples of the interior highlands. What was once the principal coastal market for the district, Eskipazar ("old market") was the terminus for a trade route that moved up the eastern valley, across the mountains, to reach the Anatolian town of İİspir, and thence Erzurum.^[12] In contrast, the western valley, lacking a natural trade route, was less conveniently connected with the Anatolian town of Bayburt. Its residents engaged in seasonal movements up and down the valley, but the valley was less accessible from the coast and from the interior.^[13]

My interlocutors in Of drew conclusions about the social attributes of each valley that are consistent with their contrasting topographies. The villagers of eastern valley are said to be of "diverse" origins by virtue of their physical appearance and social behavior.^[14] Many had accents that brought to mind the speech patterns of the people of Erzurum, across the mountains, suggesting that most immigrants had come from that area.^[15] In contrast, the villagers of the western valley are said to feature older, non-Turkic customs and practices. Although many spoke Greek as their first language, they also shared dances, stories, and songs from village to village, regardless of their mother tongue. This conservatism suggested that the rural societies in the western valley had been less unsettled by arrivals and departures.^[16]

Still, one cannot conclude that the villagers of the eastern valley were essentially of Turkic origin while the villagers of the western valley were essentially of Greek origin. More exactly, Turkish became the local language of assimilation in the eastern valley, while Greek became the local language in the western valley.^[17] In general, both Turkish and Greek as spoken in Of obey the rule of ethnic fragmentation and imperial appropriation. They are simultaneously marked by archaic traits and yet responsive to contact with the outside world. Brendemoen has found old Turkish usages in the district of Of that have vanished in most other parts of Asia Minor.^[18] Similarly, the Greek spoken in the district of Of is an old Pontic dialect unique to it and yet strongly influenced by Turkic, Arabic, and Persian.^[19]

Documenting Immigration and Conversion in Of

The traces of past immigration and conversion in different sectors of the district are consistent with published descriptions of the four Ottoman registers compiled during the first and second century of Ottoman rule.^[20] Hasan Umur carried out a study of three of these registers and summarized those portions that specifically referred to the district. Reorganizing his household counts by geographic sectors, one can detect the very beginnings of trends in migration and conversion that eventually led to the contemporary attributes of the population.^[21] The results of this reorganization are presented in table 1.[

The total Christian population in the district was increasing rather than decreasing during most, if not all, of the sixteenth century.^[22] The Muslim population, still quite small in 1515—more than fifty years after the fall of Trabzon—becomes significant only in 1583, that is, well into the second century of Ottoman rule. Thus Ottoman incorporation of the district, which would have included the allotment of tıımars and the assignment of sipahis, was not initially correlated with large numbers of arrivals of Muslims, departures of Christians, or conversions. At the time of the register of 1554, almost a century after the fall of Trabzon, the Christian population had risen slightly even in the lower valleys. On the other hand, there are now indications that the Christian population was beginning to come under some kind of pressure. Since the compilation of the previous register there appear to have been a significant number of Muslim arrivals; moreover, a proportion of the Christian population has begun to resettle in the upper valleys, where new villages are being founded. At the time of the register of 1583, there is a clear indication of the onset of the Islamization of the district population by a combination of immigration and conversion. Since the compilation of the previous register, the Muslim population has significantly risen and the Christian population has significantly declined in the lower eastern valley. This was the sector where Turkish would eventually become the language of assimilation. One might estimate that the Christian population in this part of the district was already headed toward collapse by the close of the sixteenth century.^[23] And since the Turkish spoken here now bears traces of an Erzurum dialect, arrivals of Turkic-speaking Muslims, coming directly or indirectly from the interior highlands of Erzurum, were already playing a role in the process of Islamization.^[24]

The three sixteenth-century registers give us no exact information about the role of immigration as compared to conversion in bringing about the Islamization of the district. In all likelihood, both were equally important. Some of the new Muslims must have been Turkic-speakers by the evidence of contemporary linguistic studies. Some of the new Muslims must have been converts by the evidence of personal names recorded in the registers.^[25] In other words, during the sixteenth century there was a gradual intrusion of Muslims and a gradual conversion of Christians, together resulting in an average increase of about twenty Muslim families a year from the second to the third register.

On the other hand, the registers point to the eventual establishment of Turkish and Greek as the languages of assimilation in the lower eastern and upper western sectors, respectively. In the former sector, the Muslim population was rising and the Christian population was faltering, while in the latter sector the Christian population was rising and the Muslim population remained small in number, although it too was rising. It is only the upper eastern sector that was inconsistent with the contemporary attributes of the population, since the formation of new Christian settlements did not lead to the persistence of Greek-speakers in this area. This inconsi stency reinforces the point that the population of neither valley can be considered to have been exclusively of Turkic or Greek background.

The situation in the district of Of during the sixteenth century contrasts with what was taking place in the regional capital. Toward the latter half of the sixteenth century, the Ottomans had begun to move Muslims into and Christians out of the town of Trabzon.^[26] Perhaps in response to these measures, large numbers of conversions had begun to take place among the Christians in the town. In any event, the Christians would have been inclined to convert for other reasons as well. The church hierarchy and institutions had declined, and the economic benefits of Muslim status were considerable. By the time of the registers of 1583, more than half of the population in the town of Trabzon had become Muslim, largely as a result of conversion rather than immigration, as evidenced by the fact that most of the citizenry still spoke Greek.^[27] Judging from the example of Of, the advantages of both immigration and conversion were less immediate in the outlying districts than in the regional capital itself. Unlike the regional capital, the outlying districts were not the site of major emporia serving sea and overland routes. So the Ottomans were less concerned about the composition of their populations, Muslims were less motivated to immigrate there in the absence of opportunities, and Christians were under less pressure to convert.^[28] Later, during the seventeenth century, the situation would change as the inhabitants of the eastern districts began to participate in imperial institutions as soldiers and students. The population of the eastern districts would become almost entirely Muslim, more so than the regional capital.

Christian conversion and Muslim immigration gradually gained ground in the district during the course of the seventeenth century.^[29] An official document dated 1615/1024 indicates a substantial Christian presence in the district of Of.^[30] Others, however, hint at steadily decreasing numbers of Christian households. One document dated 1631/1040 refers to 441 Christian households in the district of Of, while one dated 1673/1083 refers to only 90.^[31] Probably there was no one moment of mass conversion among the Christians, but rather one or more occasions when a large number of people, and perhaps several villages, passed over to Islam en masse. But what did it mean when an individual or a group chose to be registered as Muslim rather than Christian? Does the fact of official registration as Muslim rather than Christian indicate anything about the relationship of an individual or group to the imperial system?

As we saw in the last chapter, the ruling institution of the classical period sponsored and supported the sacred law of Islam. However, this did not mean that ordinary Ottoman subjects, whether Muslim, Christian, or Jewish, were forced or encouraged to conform to official Islam. The sacred law of Islam accords legal standing to local customs and habits (örf veâdet) of Muslims, just as it accords legal standing to the religious leadership and communities of Christians and Jews (ehl-i kitab). So, in terms of legal status and rights, the inhabitants of the eastern coastal districts did not need to become official Muslims, or, for that matter, to become Muslims at all. Nonetheless, those individuals whose family and commercial affairs conformed to official Islam would have enjoyed certain advantages. So at best the imperial regime would have only exerted mild financial and political pressures on Christians and Muslims to move toward official Islam.

In response to such pressures, it is likely that villagers and townsmen in the province of Trabzon presented themselves as official Muslims when it came to their dealings with the state. Otherwise these same villagers and townsmen would have continued to follow religious beliefs and practices that did not conform at all to official Islam. The best example of this situation is those groups of villagers in Trabzon, such as the Kurumlis, who registered themselves as Muslim with the central government but otherwise followed Orthodox beliefs and practices.^[32] And judging from the results of my fieldwork, I would expect that groups of heterodox Muslims also followed a similar strategy, presenting themselves as official Muslims before state officials, but then following Shi'i beliefs and practices when among themselves in their villages.

Thus the steady increase in the Muslim population during the seventeenth century is not in itself an indication of the emergence of a state-oriented society in the district of Of. It is possible that the local population featured all kinds of religious orientations even while the Oflus were officially registered as Muslims. But it is highly unlikely that this was in fact the case. The steady increase in the Muslim population was headed toward the disappearance of a Christian population, and this disappearance is associated with traditions of flight and apostasy. This suggests that the Oflus, both Muslim and Christian, were engaged in a transformation of their family and social lives, something far more drastic than adapting themselves to the state system by presenting themselves as official Muslims. The signs of such a transformation are clearly evident by the late seventeenth century. The Christians were converting and the Muslims were reforming, as both groups merged together to form a new kind of state society. The energy and conviction that went into this remaking of family and society was such that it soon became impossible for any of the Oflus to exempt themselves.

Documenting Imperial Participation in Trabzon

Toward the close of the seventeenth century, the Oflus were most surely identifying with and participating in imperial institutions. One of the documents transcribed by Hasan Umur indicates the extent of this change. In the year 1695/1106, the central government called on the people of the southern Black Sea coast to send 7,700 troops to report for an imperial campaign in Hungary. The interesting feature of this call-out is the distribution of troop assignments among the coastal districts. The further eastern districts, whose populations included large numbers of non-Turkic peoples, were expected to send the largest number of troops. In the document, each of fourteen districts, from Arhavi in the east to Şile in the west (beyond Bafra), were assigned a specific quota of recruits as follows (see map 2):

2,000 persons from the Kaza of Trabzon, 500 from Sürmene, 300 from Giresun, 1,000 from Of, 700 from Rize, 300 from Mapavri, 500 from Atine, 500 from Arhavi with Vetse, 250 from Keşap, 100 from Viçe, 500 from Ünye with Balya, 300 from Şile, 300 from Görele with Tirebolu, 100 from Pazarsuyu, 300 from Sinop, and, besides the total of 7,700 individuals, the resident aghas and sergeants (ağa ve çavuş) and house property owners and men of wealth, whoever they may be, are to unfurl the banners, appoint commanders (üzerine başbuğ nasb), and gather provisions for their appearance in the field at Edirne.^[33]

The large majority of the troops were called from the districts that would have had the highest proportion of non-Turkic peoples descended from the Orthodox, Byzantine population, 5,500 of the total of 7,700. So that part of the coastal region, which had so recently had a large Christian population, was to send a much larger contingent than the western coast, which had been predominantly Muslim for centuries.

The district of Of was assigned the largest quota of troops (1,000), with the one exception of the district of the provincial capital (2,000). The quota of troops for Of surpasses the total number of Muslim households attributed to the district in the last Ottoman register compiled in 1583. At the same time, the quota of troops also represents a substantial fraction, about 25 percent, of the total number of households, both Christian and Muslim, recorded for the district. So, even taking into account that the population may have increased, the call-out for the imperial campaign indicates a substantial change in the character of the population. By this date, in all probability, the large majority of Oflus had become Muslim by virtue of the combined processes of immigration and conversion. And more than this, the central government clearly expected that these "new" Muslims would respond to the call-out in especially large numbers.

Two other court documents suggest that centers of religious study had also come into existence more or less at the same time as large numbers of soldiers began to be drawn from the district.^[34] One of these, dated 1699/1110, records the following circumstances: On the complaint of one İİbrahim of the learned class (ulema), one İİbrahim Beşiroğlu of Paçan [Maraၟlıı] village in the district of Of, did sell property for a sum of money in the year 1689/1100, after which İİbrahim the purchaser went to Istanbul for educational purposes (tahsili ilm için), whereupon İİbrahim the seller did attempt to overturn the sale of the property, driving the wife and household of the purchaser from the premises and reselling it to another.^[35] The document indicates that the residents of Paçan were already engaged in religious teaching and learning toward the close of the seventeenth century. The second document, which refers to an incident that occurred in 1737/1150, indicates that a center of religious teaching and learning, with a large library, was in existence in the upper western valley by that date. According to the document, to be discussed later in more detail, an individual had come to the court with a complaint against various persons who are described as brigands. They are accused of exacting retribution on the inhabitants and killing as many as forty teachers and students while stealing seven hundred books or registers.

By the middle of the eighteenth century, if not earlier, the district of Of had become famous for its many specialists in religious teaching and learning, to a degree that was almost unparalleled in any other rural area of Asia Minor. In his study of commercial conditions along the Black Sea littoral, M. de Peysonnel, referring to conditions in 1750, writes that the district of Of was famous for its "infinite number of Men of the Law known for their erudition."^[36] Since the reputation presumably followed the fact by some years, the remark indicates that large numbers of religious teachers and students were to be found in the district no later than the early eighteenth century.^[37] By later memory and tradition, centers of religious study were distributed all over the district, both in the lower and upper valleys; however, they were especially concentrated in the upper western valley, where Greek was more commonly spoken than Turkish. Sermons were delivered in Turkish, but also commonly in Greek (in each instance with Koranic citations in Arabic). Koranic texts as well were discussed in Turkish, but also commonly in Greek.^[38]

Pontic Greek historians of the last century believed that a mass apostasy eventually took place in the district of Of, provoking the flight of those who refused to convert to Islam.^[39] They traced the cause for this mass conversion and flight to conditions of insecurity that followed the rise of aghas and agha-families (derebey). There is only a grain of truth in these otherwise erroneous traditions. The process of Islamization by conversion, which had probably begun during the first half of the sixteenth century, was gradual rather than abrupt.^[40] Eventually, by incremental processes of both Muslim settlement and Christian conversion, the district of Of became almost completely Muslim by the late seventeenth century. But while no mass conversion and flight ever occurred, the close of the seventeenth century was a turning point in the eastern districts of the old province of Trabzon. Some considerable number of Christians, both Greek-speakers and Armenian-speakers, abandoned the eastern districts at this time in order to seek refuge in the province of Canıık, significantly, that part of the coastal region that had not been part of the Greek Empire of Trebizond.^[41]

These departures occur too early to be correlated with the rise of aghas and agha-families. More probably, the remaining Christians were pressured to convert or leave by a heightening of imperial identification and participation. The rural societies of the coastal districts were becoming not just Muslim but, more specifically, imperially Muslim toward the close of the seventeenth century. Large numbers of Oflus were making their way into the ranks of regular troops and irregular militia.^[42] A smaller number were locally achieving appointments as janissary officers and soldiers.^[43] A few were even able to advance themselves in the imperial military establishment, serving outside the district in various towns or cities.^[44] At the very same time, in no less remarkable numbers, other Oflus were becoming religious teachers and students. And of these, some were also using what they had learned locally to propel themselves into the imperial religious establishment, financing their further education, when they could, in more prestigious institutions of the major Ottoman cities.^[45]

There was probably some kind of correlation between ethnic background and imperial service. Individuals of Turkic and Kurdish background were probably more typically successful in military careers, while those of Greek background were probably more typically successful in religious careers, in view of the fact that the Greek-speaking villages of the district are so strongly correlated with religious teaching and learning. But given the large numbers of both soldiers and students in Of, and adding in the factor of intermarriage, such a correlation could not have been very significant. Local participation in imperial institutions transcended ethnic identity. With the new avenues of appointment and position that opened up during the later seventeenth century, there were new possibilities for both immigrants and converts. The two could join together in hopes of making a place for themselves in the imperial system, something that had not been previously possible.

Documenting the Origins of Aghas and Konaks in Of

Since imperial participation was such an important factor in the district of Of, social changes in the district should be all the more closely correlated with changes in state policies and structures. In other words, the history of social relations in Of should follow those changes in the central government that everywhere touched the populations of the core Ottoman provinces. Documents summarized or transliterated by Hasan Umur indicate that this was indeed the case.^[46] Overall, the two hundred or so court cases and imperial decrees that he summarized are well coordinated with the major trends in the relationship of state and society. At the same time, the documents also point to the distinctive features of local elites in the province of Trabzon. When combined with local tradition, the documents suggest that imperial participation proceeded in a fashion that was not exactly duplicated in other Ottoman provinces, with results that also differed.

My interlocutors in the district of Of generally believed that the individuals who became aghas and established agha-families during the eighteenth century were often newcomers to the district of Of. I do not know of any definitive proof of this, but it is the overall opinion of both district outsiders and insiders.^[47] Local tradition also holds that the different aghas and agha-families arose at different times during the period of decentralization, some of them very early and some of them very late. Most of them then endured right down to the present day, some declining and others flourishing.^[48] The documents transcribed or summarized by Umur tell us something more than this. Given that some names that appear in the documents are unrecognizable, together with the fact that aghas and agha-families were always emerging, it seems likely that some individuals rose to prominence but then failed to establish family lines. Furthermore, the documents also suggest that the individuals who did rise to prominence, whether or not they did set down a family line, had some kind of connection with the government, but were of no special official eminence. In this respect, they are usually mentioned in terms of personal names supplemented by an "official" patronymic in the "oğlu" or "zade" form.

Of the documents Umur reviews, the earliest mention of a patronymic that is eventually associated with aghas and agha-families appears in three separate documents that address the attack, theft, and pillage of a settlement in 1679/1090. A certain Hacıı Ahmet Ayazoğlu is one of several individuals alleged to have participated in these aggressions.^[49] Some of the other accused individuals also have patronymics, but I am unable to link them with contemporary family names. The names include some with various epithets of a military (kanlıı, bey) or religious (hacıı, molla) character, but no titles or ranks. Other epithets added to their personal names indicate that the individuals in the group were of different homelands (Azaklıı) and different ethnicities (Çerkes). So the accused consist of a motley group of individuals of different familial, regional, and ethnic origins who have adopted low-level religious and military epithets. If the three documents are set in the larger context of state and society, they point toward a tentative conclusion.

During the final years of the seventeenth century, irregular troops and religious students without employment were commonly responsible for the kind of incidents described in the documents. For example, the group of individuals accused and convicted of the incidents in Of in 1679 would appear to have been so composed. So it would seem that some of the first aghas and agha-families in the district of Of arose from among individuals who were associated with the very lowest level of official military and religious activities. Alternatively, Hacıı Ahmet Ayazoğlu may have been a black sheep from a more prominent family line. This possibility is unlikely when compared with the same pattern that appears in other documents. About thirty years after the above incidents (1708), Kanlıı Hasan Ayazoğlu, acting together with brothers, a cousin, and associates, is accused and convicted of usurping the authority of the district janissary agha by appointing his son-in-law to that position. Then two years later this same individual, still other brothers, a cousin, and associates are named as well-known brigands who have been attacking marketplaces, committing homicides, and assaulting unmarried women for more than a decade.^[50]

A second set of documents describes even more serious incidents occurring in the year 1737/1150.^[51] A group of individuals is accused of hundreds of crimes, including assault, homicide, theft, looting, burning, rape, and kidnapping in the course of separate incidents. Among the names of the leaders one finds individuals who bear the patronymics Fettahoğlu, Çap-oğlu, Selimoğlu, Hacııhasanoğlu, and Keleşoğlu, all of which are patronym-ics later associated with aghas and agha-families of the Five and Twenty-five parties.^[52] In one of the documents, these individuals are described as brigands who act in concert with other brigands from the Sixty-fourth Regiment of Janissaries (also named) and from the Fifth Regiment of private soldiers (not named).^[53] As in the previous documents, the individuals named have adopted various epithets of a military (alemdar, deli, kanlıı) or religious (hacıı, molla) character; however, there is no clear indication of diverse familial, regional, or ethnic origin. So the individuals in question are clearly of the lowest military and religious stations. This is especially significant since the targets of their aggressions are residents of Of bearing titles and ranks of the religious and military branches of government. Moreover, the crimes of which they stand accused are of such a serious character that it would appear they were attempting to annihilate the family lines and household organizations of prominent individuals.

In one incident, the brigands attack a retired janissary officer, Piri Çavuş zade Mustafa, kill him with a bullet not far from his house, then attack the house, seize the house and his wife, and then turn the house and the wife over to other outlaws. They later attack the house of the father of Mustafa, also an individual of some distinction, loot it of cash and valuables, and then burn it down. They track down the father, who has fled to Rize, and they finally kill him as he leaves the mosque after completing his Friday prayers. Then they return to his residence in Rize, where they kill his wives and servants and burn the house. On yet another occasion, the same individuals are said to have caused considerable destruction in and around Paçan village while engaged in a battle with private soldiers. Attacking the village, they burn buildings and steal property, including seven hundred books or registers. After being taken to court and ordered to pay restitution, they refuse to pay and are declared to be in a state of judicial noncompliance. In the document transcribed by Umur, they are accused (but not convicted) of returning to the village, pillaging it once again, carrying off women and children to their "towers," and committing numerous homicides. They are accused, but not convicted, of having killed as many as forty among the religious teachers and students and as many as five hundred altogether.^[54]

The last set of documents provides many indications regarding the social origins of the new local elites. First, the documents are a further confirmation that the individuals whose descendants later become aghas and agha-families are drawn from the milieu of irregular troops or local militias (bölük).^[55] Second, they indicate these individuals began their rise to prominence by assaulting and threatening individuals with titles and ranks. Third, they show that individuals with different patronymics acted in concert with one another to challenge and intimidate prominent local residents. Fourth, they indicate that professors and academies were in existence in the district during the first half of the eighteenth century. Fifth, they suggest that the aghas and agha-families only arose after the professors and academies had already been established rather than in tandem with them. And sixth, the documents also point to tantalizing links between the regimental affiliations of irregular troops and the later emergence of the Five andTwenty-five parties.

The individuals with the aforementioned patronymics are described as brigands residing in Of who combine with brigands from the Sixty-fourth Regiment of Janissaries to act together with other known individuals of the Fifth Regiment of Soldiers.^[56] By a report of Peysonnel, the province of Trabzon was troubled for many years by the rivalry of the Twenty-fifth and Sixty-fourth regiments of janissaries.^[57] By combining the Umur documents and the Peysonnel report, one discovers an indication of the origins of the Five and Twenty-five parties in the district of Of: Members of the Fifth and Sixty-fourth regiments join in an attack in Of, while other members of the Sixty-fourth and Twenty-fifth regiments are bitter rivals in the province of Trabzon.^[58] The patronymics attributed to the individuals who collaborated in the attacks in 1737/1150 eventually appear among the Five and Twenty-five parties. The Selimoğlu have become the leading agha-family in the Five Party just as their name is associated with the Fifth Regiment of Soldiers in 1737.^[59] However, the Fettahoğlu and Çapoğlu are later associated with the Twenty-five rather than the Five Party. This change is consistent with a tradition that tells how the Fettahoğlu and Çapoğlu fell out with the Selimoğlu in the early nineteenth century and joined the Muradoğlu, who assumed the leadership of the Twenty-five Party.^[60] So then, by a web of interrelated factual and hypothetical linkages, the regimental affiliations of the individuals involved in the attacks on prominent local residents point directly to the early origins of leading individuals, large residences, family lines, and district social formations.

Altogether eight patronymics that are eventually associated with aghas and agha-families during the period of decentralization are mentioned in the documents. These are Ayazoğlu, Bektaşoğlu, Çapoğlu, Fettahoğlu, Hacııhasanoğlu, Keleşoğlu, Nuhoğlu, and Selimoğlu. All of these names are consistently associated with regiments or janissaries, but in a way that confirms that they were individuals of the outer and lower fringes of the military establishment. The individuals with known patronyms are sometimes described as members of regiments, but only those composed of irregular soldiers or local militias (bölük). They are sometimes accused and convicted of attacking local janissary officers, but they never appear in the capacity of true janissary officers themselves. This pattern is supported by other incidents in which individuals accused of wrongdoing seem to be attempting to worm their way into the position of official janissaries. For example, individuals might carry out aggressions while falsely claiming to be acting as official janissaries. Or contrariwise, they are the victims of aggressions carried out by others who falsely claim to be acting as official janissaries.

So far the documents I have cited strongly suggest that the aghas who founded agha-families arose for the most part from lower-level regiments and militias. Some of them may have held janissary titles and ranks, since it became common for all kinds of individuals to do so. More typically, they belonged to regiments and militias that imitated and emulated the janissaries by their tattoos, insignia, and banners. Other documents offer further support for this same conclusion in a negative way by excluding the possibility that the aghas and agha-families were of other social origins.

One document, for example, describes how high state officials without appointments assembled groups of armed men from Of and then went about raiding and looting the villages of the district. During the year 1711/1123, the head steward (kethüda) of the former governor of Trabzon, in the com pany of close associates, is accused of raiding and looting villages in the district of Of. The head steward, leading a band of brigands recruited from various villages in the western valley-system, conducted a massive assault on fifteen villages (unnamed and unlocated). They burned and looted two hundred houses and warehouses. They destroyed seven mosques and raped eighteen girls. They shaved the heads of eight women and paraded them about, tied some of them to horses and dragged them, tracked down others and raped them. Fleeing in terror, many of the villagers subsequently died of exposure and starvation.^[61]

The document illustrates how higher state officials, temporarily without positions or appointments, mustered a large band of brigands by recruiting soldiers in nine separate villages in Of.^[62] It gives the names, patronyms, and villages of more than a score of Oflus who took part in the raiding and looting. On the other hand, among the accused Oflus there is not a single mention of a patronym that later emerges as an agha-family. If this and three other documents transcribed by Umur are representative, these kinds of incidents do not appear to have played a role in the origins of aghas and agha-families in the district of Of.^[63] The individuals who are the ascendants of aghas and agha-families are not higher state officials without appointments, nor do they appear to have been followers of such higher state officials.

Another set of documents diminishes the possibility of a very different kind of social origin for aghas and agha-families. There are other individuals who organize the raiding and looting of villages but have no detectable connection at all with imperial institutions. They are not drawn from either high or low-level positions in the military or religious establishment. They are instead "tribesmen." The documents describe the following circumstances. During the year 1709/1121, the residents of seven villages in the upper western valley-system are accused of all sorts of crimes: They combined with the Ayaslo and Kolotlo tribes (kabîle), who came from outside the district and settled among the villagers. Acting in concert, tribesmen and villagers disturbed the peace of the main market in the district of Of. They attacked this market, making it necessary to move it to another place. They attacked houses, kidnapped women and girls, held them prisoner, sold them to one another, and married them as they wished. They attacked the house of the judge of Of and threatened him with death. They attacked individuals with knives. They did not pay dues to the fief-holders (tıımar and zeamat) of their villages. They held the judge of Of in contempt, saying they had given him documents when he attempted to make a case against them. In punishment for all these violations, the order was given for the pasha of Trabzon to undertake the mass deportation of the two tribes and seven villages. They were to be rounded up and transported to Anakra Castle, where they would be settled at the Ottoman frontier with Georgia.^[64]

The tribesmen organizing villagers are not named as individuals who bear patronymics. The names of their tribes do not appear among the patronymics of aghas and agha-families. The tribesmen are not associated with official titles and ranks, either truly or falsely. The tribesmen directly attack state officials and subvert the normal operations of the central government. Most telling of all, state officials are now capable of making a vigorous and ruthless response. Soldiers are sent to the seven villages, where they surround the villagers and descend on them like a flood.^[65] Both the tribesmen and the villagers charged in these incidents are rounded up and deported to the frontier. The tribesmen, being from outside the state system altogether, were perhaps less able to retard or block the overzealous but effective response of state officials. In contrast, the cases lodged against the janissaries and soldiers who attacked prominent individuals in Of dragged on for three decades in the courts.

Of all the documents that Umur has assembled and discussed, about forty refer to incidents that can be plausibly linked with the emergence of local elites who challenged existing military, administrative, and judicial hierarchies. All of the documents date to the final decades of the seventeenth century or the first decades of the eighteenth century.^[66] Usurpers or brigands are accused of having assaulted, robbed, intimidated, or assassinated individuals with military or religious titles. Military officials appointed by the central government are unable to exercise the authority vested in them by their office.^[67] Men in arms who are "from among the usurpers and evildoers" (mütegallibeden ve eşirrâdan) and pretend to be janissaries descend on the market and battle one another with sword and shield, pistol and rifle. Committing many wrongs, they seize goods and attack the court.^[68] They extort taxes from villagers, intimidate state officials, raid and loot village populations, seize the lands of non-Muslim subjects, and illegally arrest and punish people.^[69] The documents that describe challenges to the authority of military officials occur in tandem with other documents that describe the subversion of judicial procedures. One individual with the title of molla, who acts in concert with another from Paçan village, is accused of improperly assuming the duties of court officials (naip and kâtib).^[70] Other individuals who style themselves as notables (âyan), but are said to be "from among the usurpers" (mütegallibeden), are accused of putting pressure on the district judge (kadıı), or his representative (naip), by receiving them in their houses and requiring they hear cases there.^[71] Men in arms appear in the court carrying their weapons and disrupt the proceedings.^[72] Men are accused of descending on the court while it is in session and freeing their confederates who are standing trial.^[73]

Documenting the Aghas and Family lines in Trabzon

From the middle of the eighteenth century, Umur's collection of documents begins to change in character. The signs of conflict between new social elements and religious and military officials at the local level are no longer apparent. There are no reports of incidents involving attacks on state officials and brigands falsely claiming to be janissaries, and there are fewer cases that mention brigands and outlaws with the patronymics of agha-families. There are no further cases of mass assaults on villages, deportations of village populations, illegal land seizures, or the failure to pay taxes. All kinds of social disorders may have been occurring in the districts, but if so, they were infrequently brought to the attention of higher military and judicial officials. The local elites were perhaps usurping offices, interfering with the courts, improperly imposing taxes, refusing to forward revenues to the capital, and illegally confiscating property. But if they were, they were able to prevent the lodging of complaints against them.

Toward the close of the eighteenth century, a new order of aghas, mansions, family lines, and regimental parties was in place. This is indicated by three documents that call on local elites to muster troops. The earliest of the three, dated 1774/1188, was written by a military commander representing the palace. He calls on "the judges, religious officials, the righteous, officers, commanders, craftsmen, and community leaders" (kadıılarıına, ûlema, sûleha, zabitan, serdarlar, iş erleri, söz sahiblerine) of nine districts of the province to assemble and dispatch troops for participation in imperial military campaigns against the armies of Moscow.^[74] However, in doing so, he reveals that the central government is unable to compel them to furnish troops, but entirely dependent on their willingness to do so. He chastises his correspondents for having failed to respond to previous orders to assemble troops for imperial campaigns. They have said they were ready to come and have unfurled their banners, but then, offering various excuses, have failed to mobilize. They have engaged in quarrels among themselves, a kind of behavior that cannot be called anything other than disobedience to the sultan. Begging them to mend their ways, the military commander exhorts his correspondents to assemble for liberating the Crimea from Moscow. Their previous faults will be forgiven. Their military service is a religious duty.

The other two documents, dated 1788/1202 and 1789/1204, provide more direct evidence that aghas from agha-families governed their districts more or less independently of the central government at this time. The earlier is written by a palace official.^[75] It addresses the notables (âyan), officials (mütesellim), judges (kadıı), and officers (yeni çeri) of the districts in the province of Trabzon, from Görele in the west to Hopa in the east. It calls on them to assemble troops and report for a campaign on the northeastern shores of the Black Sea (Anapa, Soğucak). The document lists 26 names of individuals or families in 16 districts, referring to them as "aghas" (ağavat). They are required to contribute either 100, 150, or 200 troops. About two-thirds of these names specify a patronymic, each in the "son of" (oğlu) format. Some of these add a personal name to the patronymic, as in "Tuzcuoğlu Hüseyin at Rize" and "Kuğuoğlu Süleyman at Görele."^[76] Others merely list the patronymic alone without specifying an individual, as in "Bahadıır oğlu at Büyük Liman" or "Hacıı Hasan oğlu at Pulathane." In a number of instances, the patronymics listed in the document correspond with the patronymics of aghas who held official appointments and positions fifty years later, when European consuls began to report on the coastal region at the close of the period of decentralization. In other instances, the patronymics correspond with local traditions of aghas and agha-families during the nineteenth century.^[77] The number of troops required from three families at Of are 150, 200, and 200, respectively. The total for the district of Of, 550, is larger than for any other district in the province. The total number for the three eastern districts of Rize, Of, and Sürmene is almost 40 percent of the total for the entire eastern coastal region.

The later document, dated 1789/1204, is a call-out for troops for a military expedition to the northeastern shores of the Black Sea (Anapa).^[78] The troops are to be assembled in various ports of the Black Sea coast, including Samsun, Trabzon, Sürmene, Rize and others. The government will pay merchant ships to transport these troops, and these ships will be accompanied by galleons of the Imperial Navy. The document is divided into five sections, which refer in turn to the vicinity of Trabzon, the district of Of, the district of Rize, the aghas of the Laz, and the western districts. In the first four sections, there are lists of individuals and families, each of which is called upon to contribute a certain number of troops. The names are usually the same as those in the earlier document, but there are now more than forty names for the four mentioned areas. The total call-out for each of these areas is roughly 5,200 for the vicinity of Trabzon, 1,800 for Of, 2,300 for Rize, and 1,800 for the aghas of the Laz.^[79] The individuals are usually labeled in accordance with their regimental affiliations, including the Sixty-fourth, Twenty-fifth, and Fifth. The document therefore confirms that the parties in Of were associated with janissary regiments, just as it also confirms the depth and breadth of local military affiliations.^[80]

Taken together, the three documents indicate that the social changes taking place in the district of Of during the period of decentralization were consistent with social changes taking place in all the core provinces of the Ottoman provinces. By the late seventeenth century, large numbers of Oflus were serving as irregular soldiers or studying as religious students. Occasionally, some of these elements joined up with imperial elites who, while out of office or back from campaigns, extorted money from villagers and looted their farms. By the fourth decade of the eighteenth century, individuals with low-level military appointments were asserting themselves, sometimes by banding together to usurp the authority of local officials or to intimidate local villagers. At the same time, individuals with low-level religious qualifications had organized religious academies and were recruiting and training considerable numbers of religious students. But these teachers and students were eventually confronted with the emergent aghas and agha-families. Just as the palace had captured the mosque in Istanbul during the classical period, so too would aghas and agha-families capture the religious teachers and students during the period of decentralization, neutralizing them as alternative local sources of political legitimacy or social organization.

Aghas and mansions together with hodjas and academies were to be found in virtually every segment of the eastern coastal region. Coalitions of local elites constituting regional social oligarchies comprised the entirety of the province of Trabzon. However, some coastal districts of the province of Trabzon featured a broader and deeper participation in imperial institutions than others. In general, those coastal districts that had large Christian majorities and little Turkic settlement during the sixteenth century, such as Rize, Of, and Sürmene, are associated with broad and deep participation in imperial institutions. By contrast, coastal districts that had a sizable Muslim population and considerable Turkic settlement by the sixteenth century, such as Tirebolu, Giresun, and Ordu, were associated with narrower and shallower participation. It was then those populations that had a background in market and state participation that became part of the imperial system. To some degree, a legacy of participation in Byzantine institutions had helped their otherwise mixed populations to recognize and to exploit the opportunity for becoming part of the Ottoman state system and state society.

Two exceptions to the preceding pattern prove the rule. First, the hierarchy and institutions of the Orthodox Church had survived Ottoman incorporation precisely in that part of the coastal region where the Byzantine state system had been centered and developed, that is to say, the central districts around the town of Trabzon. According to Bryer and Lowry, the survival of these pieces of the older state system had been the basis for the survival of a large Greek Orthodox population in this same area. So an alternative to the imperial system had retarded conversion and immigration, and hence imperial participation, in this section of the province of Trabzon.^[81] Second, one of the coastal districts in the west bears some resemblance to the district of Of in the east. The population in the district of Tonya included a large number of Muslim Greek-speakers, as well as broad and deep participation in imperial institutions (military rather than religious). But Tonya also stands as an exception among the western coastal districts. By its more isolated location, it did not receive as many Turkic settlers as other western coastal districts. So it is an example of a coastal district in the west that fits the pattern of the coastal districts in the east.

Conclusion

The Ottoman Empire is usually conceived in terms of a strict division between rulers and ruled. The rulers consisted of an official (askeri) class composed of military, administrative, judicial, and religious specialists. The ruled consisted of a diverse population of tribes, peasants, and townsmen. The rulers accorded legal status and rights to the ruled in various ways, one of the important ways being recognition of their religious leadership and communities.

The district of Of, as an example of the rural societies of the old province of Trabzon, is not consistent with such an analysis. The Oflus, those of Muslim as well as those of Christian background, set aside their attachments to standing beliefs and practices as they oriented themselves toward an imperial project that claimed universal import. In doing so, they sought to organize everyday life in accordance with a system of ethical thought and practice as represented by authoritative texts and learned experts. What had always been said and done was discounted as degenerate and corrupt. What should be said and done was accessible only through literacy, teaching, learning, argument, and consensus. This reorientation of family life and social relations can be described as an effort to become good and proper Muslims, rather than remain ignorant and impoverished villagers. But it included a political dimension that exceeded piety and spirituality. The Oflus had set about to remake themselves in order to remake the world. They claimed a place in the state society of the imperial system, challenging the distinction between officials (askeri) and nonofficials (reaya). To do so, they acquired all kinds of military weapons, and they organized themselves in military formations. The idea of an empire of the future was perhaps more alive in the mansions and academies of Trabzon than in the great monumental centers of Istanbul.

The period of decentralization was the direct result of crises of state stability and competition. But the latter had been indirectly provoked by technological inventions and institutional disciplines occurring in the Euro-American sphere. Print, school, steam, factory, rifles, and barracks had led to new, more efficient forms of state power which lowered, rather than raised, the need for manpower and resources. In response, the Ottomans were pressed to respond to new rounds of imperial competition by expanding, rather than contracting, their ability to mobilize manpower and resources. The dissemination of the imperial project in the eighteenth century therefore anticipated the emergence of nationalism in the Euro-American sphere. Local consciousness in the province of Trabzon had become linked with state identification and participation at an earlier phase of the global modernizing process. This means that the period of decentralization was a preparation, and hence set the conditions, for the reception of national democracy.

Notes