Nationalism – a highly sensitive and difficult subject at the root of shifting understandings of identity – forms an important focus of attention in the present chapter. In its essence, nationalism speaks of one dominant nationality; for example, the Turkish republic is said to rest on a Turkish identity. Yet the Ottoman Empire for much of its history brought together multiple and different ethnic and religious groups. At times their interaction was co-operative and harmonious; but under the pressures of “modern nationalism” those ethnic and religious relations deteriorated into hostilities and worse, massacres, that remain a difficult subject in memory and national accounting. This issue is particularly acute in the interactions among, for example, modern-day Turks, Armenians, Greeks, and Kurds, as well as Palestinians and Israelis.
The subject of historical intergroup relations in the Ottoman Empire looms large because of the many conflicts that currently plague the lands it once occupied. Recall, for example, the Palestinian–Israeli struggle, the Kurdish issue, the Armenian question, as well as the horrific events that have befallen Bosnia and Kossovo. All rage in lands once Ottoman. What then, is the connection between these struggles of today and the inter-communal experiences of the Ottoman past?
Let me begin with the assertion that there was nothing inevitable about these conflicts – all were historically conditioned, that is, produced by quite particular circumstances that evolved in a certain but not unavoidable manner. Other outcomes historically were possible but did not happen because of the way in which events unfolded. Nor, it is important to repeat, are these struggles ancient ones reflecting millennia-old hatreds. Rather, each can be explained with reference to the nineteenth and twentieth centuries, through the unfolding of specific events rather than inherent animosities of an alleged racial or ethnic nature. But because these contemporary struggles loom so large and because we assume that present-day hostilities have ancient and general rather than recent and specific causes, our understanding of the Ottoman inter-communal record has been profoundly obscured.
Despite all stereotypes and preconceptions to the contrary, inter-Ottoman group relations during most of Ottoman history were rather good relative to the standards of the age. For many centuries, persons who were of minority status enjoyed fuller rights and more legal protections in the Ottoman lands than, for example, minorities in the realm of the French king or of the Habsburg emperor. It is also true that Ottoman inter-communal relations worsened in the eighteenth and nineteenth centuries. In large part, this chapter argues, the deterioration derives directly from the explosive mixture of Western capital, Great Power interference in internal Ottoman affairs, and the transitional nature of an Ottoman polity struggling to establish broader political rights. Such an assessment does not aim to idealize the Ottoman record of inter-communal relations, which was hardly unblemished. Neither does it seek to explain away the major injustices and atrocities inflicted on Ottoman subjects by the state or other subjects.
The goal is to replace the stereotypes that too long have prevailed regarding relations among the religious and ethnic Ottoman communities. One’s religion – as Muslim, Christian, or Jew – was an important means of differentiation in the Ottoman world. Indeed, ethnic terms confusingly often described what actually were religious differences. In the Balkan and Anatolian lands, Ottoman Christians informally spoke of “Turks” when in fact they meant Muslims. “Turk” was a kind of shorthand for referring to Muslims of every sort, whether Kurds, Turks, or Albanians (but not Arabs). Today’s Bosnian Muslims are called Turks by the Serbian Christians even though they actually share a common Slavic ethnicity with these Christians. In the Arab world, Muslim Arabs used “Turk” when sometimes they meant Albanian or Circassian Muslim, one who had come from outside the region.
Stereotypes present distorted and inaccurate pictures of Ottoman subjects living in sharply divided, mutually impenetrable, religious communities called millets that date back to the fifteenth century. In this incorrect view, each community lived apart, in isolation from one another, adjacent but separate. And supposedly implacable hatreds prevailed: Muslims hated Christians who hated Jews who hated Christians who hated Muslims. Recent scholarship shows this view to be fundamentally wrong on almost every score. To begin with, the term millet as a designator for Ottoman non-Muslims is not ancient but dates from the reign of Sultan Mahmut II, in the early nineteenth century. Before then, millet in fact meant Muslims within the empire and Christians outside it.
Let us continue this exploration of inter-communal relations and look at two different versions of the past, taken from Ottoman Bulgaria during the 1700–1922 era. In the first version, we hear the voices of Father Paissiy (1722–1773) and S. Vrachanski (1739–1813) calling their Ottoman overlords “ferocious and savage infidels,” “Ishmaelites,” “sons of infidels,” “wild beasts,” and “loathsome barbarians.” Somewhat later, another Bulgarian Christian writer Khristo Botev (1848–1876) wrote of the Ottoman administration in a similar vein:
And the tyrant rages
and ravages our native home:
impales, hangs, flogs, curses
and fines the people thus enslaved.
In the first quotation are the words of Bulgarian émigré intelligentsia who were seeking to promote a separate Bulgarian nation state and break from Ottoman rule.1 To justify this separation, they invented a new past in which the Ottomans had abruptly ended the Bulgarian cultural renaissance of the medieval era, destroying its ties to the West and preventing Bulgaria from participating in and contributing to western civilization.
And yet, hear two other Bulgarian Christian voices speaking distinctly differently about Bulgarian Muslims, the first during the period just before formal independence in 1908 and the other a few years later:
Turks and Bulgarians lived together and were good neighbors. On holidays they exchanged pleasantries. We sent the Turks kozunak and red eggs at Easter, and they sent us baklava at Bayram. And on these occasions we visited each other.2
In Khaskovo, our neighbors were Turks. They were good neighbors. They got on well together. They even had a little gate between their gardens. Both my parents knew Turkish well. My father was away fighting [during the Balkan Wars]. My mother was alone with four children. And the neighbors said: “You’re not going anywhere. You’ll stay with us…” So Mama stayed with the Turks…What I’m trying to tell you is that we lived well with these people.3
Thus, as the various quotations demonstrate, some Bulgarian Christian writers emphasize the differences between “Bulgarians” and “Turks” while others stress the everyday, friendly relations existing between two sets of neighbors.
Concepts of the “other”, that characterize the first set of quotations, abound in history. The ancient Greeks divided the world into that of civilized Greeks and of barbarian others. Barbarians could be brave and courageous but they did not possess civilization. For Jews, there are the goyim – the non-Jew, the other – whose lack of certain characteristics keeps them outside the chosen, Jewish, community. For Muslims, the notion of the dhimmi is another way of talking about difference. In this case, Muslims regard Christians and Jews as “the People of the Book” (dhimmi), who received God’s revelation before Muhammad and therefore obtained only an incomplete message. Thus, dhimmi have religion, civilization, and God’s words. But since they received only part of that message, they are inherently different from and inferior to Muslims.
In the Ottoman world, people were acutely aware of differences, for example, those between Muslims and non-Muslims. Muslims, as such, shared their religious beliefs with the dynasty and most members of the Ottoman state apparatus. The state itself, among its many attributes, called itself an Islamic one and many sultans included the term “gazi,” warrior for the Islamic faith, among their titles. Later on, as seen, they revived the title of caliph, one with deep roots in the early Islamic past. Further, for many centuries military service primarily was carried out as a Muslim duty, although there always were some non-Muslims in the military service, such as Christian Greeks serving as sailors in the navy during the 1840s. Yet, in a real sense, the military obligation had become a Muslim one. Even when an 1856 law required Ottoman Christian military service, the purchase of exemption quickly became institutionalized as a special tax. A 1909 law ended this loophole but in response hundreds of thousands of Ottoman Christians fled the empire rather than serve. Thus, subjects understood that Muslims needed to fight but non-Muslims did not.
A variety of mechanisms maintained difference and distinction. Clothing laws, as seen earlier, distinguished among the various religious communities, delineating the religious allegiance of passersby. They reassured maintenance of the differences not simply as instruments of discipline but also as useful markers of community boundaries, immediately identifying outsiders and insiders. Apparel gave a sense of group identity to members of a specific community.
Until the nineteenth century, the legal system was predicated on religious distinctions. Each religious community maintained its own courts, judges, and legal principles for the use of coreligionists. But Ottoman realities made the Muslim courts more powerful. Since Muslims theologically were superior, so too, in principle, was their court system. Muslim courts thus held sway in cases between Muslims and non-Muslims. The latter, moreover, simply did not possess the necessary authority (velayet) and so, with a few exceptions, could not testify against Muslims. The state used the religious authorities and courts to announce decrees and taxes and, more generally, as instruments of imperial control. The ranking government official of an area, for example, the governor, received an imperial order and summoned the various religious authorities. They in turn informed their communities which negotiated within themselves over enforcement of the order or distribution of the taxes being imposed.
Because of their inherently greater authority, Muslim courts often provided rights to Christians and Jews that were unavailable in their own courts. And so non-Muslims routinely sought out Muslim courts when they were under no obligation to do so. Once they appeared before the Islamic court, its decisions took precedence over the decisions of other religions’ courts. Thus, non-Muslims often appealed to Muslim courts to gain access to the provisions of Islamic inheritance laws that absolutely guaranteed certain shares of estates to relatives – daughters, fathers, uncles, sisters. Persons who feared disinheritance or a smaller share in the will of a Christian or Jew placed themselves under Islamic law. Christian widows frequently registered in the Islamic courts because these provided a greater share to the wife of the deceased than did ecclesiastical law. Or, take the case of dhimmi girls being forced into arranged marriages by fellow Christians or Jews. Since Islamic law required the female’s consent to the marriage contract, the young woman in question could go to the Muslim court that took her side, thus preventing the unwanted arranged marriage.
With the Tanzimat reforms, the old system of differentiation and distinction and of Muslim legal superiority formally disappeared. Equality of status meant equality of obligation and military service for all. The clothing laws disappeared almost entirely and, while the religious courts remained, many of their functions vanished. New courts appeared: so-called mixed courts at first heard commercial, criminal, and then civil cases involving persons of different religious communities. Then, beginning in 1869, secular courts (nizamiye) presided over civil and criminal cases involving Muslim and non-Muslim. Whether or not these changes automatically and always improved the rights and status of individuals – Christian, Jew, or Muslim – currently is being debated by scholars. Some writers, for example, argue that women’s legal rights overall declined with the replacement of Islamic by secular law, but others disagree.
So, how equal were Ottoman subjects and how well were non-Muslims treated? Quite arbitrarily, I offer the testimony of the Jewish community of Ottoman Salonica, as recorded in the “Annual Report of the Jews of Turkey” of the Bulletin de l’Alliance Israélite Universelle in 1893. French Jews had founded the Alliance Israélite Universelle in 1860 to work for Jewish emancipation and combat discrimination all over the world. The organization placed great stress on schools and education as a liberating device, establishing its first Ottoman school in 1867 and within a few decades, some fifty more. It published a journal, the Bulletin, in Paris, to which Jewish communities from all over the world sent letters reporting on local conditions. Here then is the statement which the Jewish community of Salonica sent to the Bulletin in 1893:
There are but few countries, even among those which are considered the most enlightened and the most civilized, where Jews enjoy a more complete equality than in Turkey [the Ottoman Empire]. H. M. the sultan and the government of the Porte display towards Jews a spirit of largest toleration and liberalism.4
To place these words in context, we need to consider several points. First, the authors of the statement were quite aware that the treatment of Jews in many parts of Europe was atrociously bad and, by comparison, Ottoman Jews truly were better off. Second, the statement possibly can be read at face value since it was not prepared for circulation within the empire (but, nonetheless, the authors could surmise their views would become known to the Ottoman state). And third, Ottoman Jewish–Muslim relations were better than Muslim–Christian (or Jewish–Christian) relations. Even after all these reservations are taken into account and although this statement explicitly is about only Ottoman Jews, it likely also represents the sentiments of large numbers of Ottoman Christian subjects as well during the eighteenth and nineteenth centuries.
Residential patterns – whether people of the different communities lived separately or apart – provide an important key to understanding intercommunal relations. The example of mid-nineteenth-century Salonica at first glance seems to suggest a pattern of segregation by religious community. The city map of Salonica at the time indicates separate Jewish, Muslim, and Greek Orthodox quarters and further depicts these respective quarters generally clustered together. Thus, thirty-eight of the forty-three Muslim quarters are concentrated in the northern part of the city, while eight of the twelve Greek quarters are in the central and southeast corner and all sixteen Jewish quarters in the south-central district. And yet, quarters of the three communities also are scattered about, sometimes in the middle of quarters of a different religious community. Thus, one Greek Orthodox quarter appears right in the middle of a group of Jewish quarters while another is embedded among the Muslim quarters. Also, it is unclear if quarters designated as Jewish, Greek Orthodox, or Muslim held any persons of another religion. That is, we do not know if large numbers of Christians or Muslims resided in a “Jewish” quarter in Salonica but we do know this was the case elsewhere in the empire.
Overall, residential exclusivity by community was not the rule in the 1700–1922 era. In the European provinces, Muslims in the city of Resen did not live in separate quarters of the city (although they did in Ohrid). In many regions, households of different religious communities clustered together according to wealth. This pattern held for Istanbul during the nineteenth century, where the wealthy lived near the palace. But elsewhere in the capital, different economic strata lived together in many residential neighborhoods. In nineteenth-century Ankara, an unimportant provincial town and thus very different from the imperial capital, certain quarters had been cohabited by both Muslims and non-Muslims for several centuries. Mid-eighteenth-century Aleppo provides a well-documented, strikingly clear example of residential patterns according to wealth and not religion. Here we know both the patterns by quarter and even who lived in the particular houses of the quarter. In this carefully studied case, no quarter was inhabited by only a single religious community. And, names could be deceiving: hence, the so-called Jewish quarter of Aleppo held only part of its Jewish population while many Muslims called the neighborhood home as well. The Kurdish quarter at the time in fact was empty of Kurds; none remained from the original Kurdish settlement there in the medieval Mamluk era. Indeed, in the early twentieth century, 93 percent of the residents of this so-called Kurdish quarter were Christian (Kurds almost exclusively were Muslim). Thus, while Muslims, Christians, and Jews in Aleppo often lived with their own kind, they also often inhabited mixed neighborhoods. In Aleppo, Jewish homes nestled up to a mosque while Muslim homes were adjacent to a synagogue. Instead of separation by religion, quarters in Aleppo tended to be strikingly homogeneous in social and economic status. Thus, the inhabitants of this important Arab city often preferred to live with others of similar wealth rather than the same religion. Elsewhere, as in Istanbul and Ankara, rich and poor and middling often resided together in the same neighborhoods. In sum, when Ottoman families chose their home sites, they used a host of criteria and not simply religion. Depending on time and place and whim, the economic status of neighbors, the convenience of the neighborhood as well as religion affected their selection. Overall, there was a high degree of inter-communal residential mixing.
The argument for intimate daily contact among members of the various religious and ethnic communities is further supported by the very languages spoken in the Ottoman Empire, as well as the liturgical music employed. Even the most cursory glance at the official Ottoman language demonstrates an incredibly rich intermixing rather than separation of communities. The Ottoman language largely is Turkish in syntax and grammar, but written in the Arabic script. It contains massive infusions of Arabic vocabulary (perhaps 40 percent of the total), an equal amount of Turkish and a lesser measure of Persian. Many other languages are represented as well. Among nautical terms, for example, there are perhaps 1,000 Greek and Italian loan words that entered into Ottoman usage, together with many words from Spanish, English, French, German, Portuguese, Bulgarian, Old Serbian, and Russian, among others. When new foods came into the Ottoman diet, the names given by their lenders often entered with them. Thus, tomatoes and potatoes in Ottoman were called by words derived from those spoken by the Nahuatl peoples of southern Mexico and the Taino in the Caribbean. In addition, there are numerous German, French, English, and other loan words for objects ranging from bread to carriages to the machinery of the industrial age, including steam itself. The name of the Ottoman coin kuruş, derives from the German groschen. Nor is Ottoman the only language of the empire to reflect such richness. In Cilicia, in southeast Anatolia, Armenians spoke Turkish but wrote it in the Armenian script. Similarly, Greek Christians in western and northwestern Anatolia, mainly around Kayseri, spoke Turkish but wrote it in the Greek alphabet (a language called Karamanlıca). The Greek spoken at Kayseri contained so much Turkish that knowledge of both languages was needed to understand it. Many Greeks in Istanbul spoke only Turkish in the late eighteenth and early nineteenth centuries. Similarly, consider that, in Aleppo during the mid-eighteenth century, the Christian, Jewish, and Muslim religious liturgies all were based on the same Arabic melodic system (makam). Such linguistic and musical interpenetrations demonstrate communities in constant and intimate contact rather than groups sealed off from one another.
Relations in the workplace, in common with residential patterns and linguistic and musical borrowings, demonstrate intimate daily contact among the various religious and ethnic communities. Here too, gross and untenable generalizations have prevailed, often under the name “ethnic division of labor.” In some of the scholarship on Ottoman history, this widely used term essentially meant that particular ethnic or religious groups in general inherently were especially well suited to carry out certain tasks. And so, they were said to dominate that activity throughout the empire. Thus, Turks (taken to mean Muslims) allegedly did certain jobs but not others, while Christians of the various denominations respectively performed other tasks. In agriculture, Turks supposedly were cereal growers while Armenians and Greeks grew fruits and vegetables. In the industrial sector, Armenians were said to be the silkweavers and Greeks the tailors; Turks for their part allegedly excelled in the applied arts, such as carpet making and woodworking. Further, according to this division of labor, Greeks and Armenians were gifted in commerce but often dishonest, particularly the former. Turks on the other hand were unimaginative and dull but honest and, to boot, made good administrators. Such crass generalizations, rightfully, are seen as inappropriate in other areas of historical writing. For example, it is considered both inaccurate and unacceptable to state that Jews are especially skilled in business or Irish-Americans in bricklaying. And yet such stereotypes still reside in Middle East history.
As in many stereotypes, there is a nugget of reality. While there was no empire-wide division of labor, certain groups in particular localities did monopolize a particular industry. In some areas, most cereal farmers were ethnic Turks and most silkweavers Armenians – but the statements do not hold true for the empire as a whole. Elsewhere, the Greek Christians were cereal growers and Muslims silkweavers. Some observer might have noticed that Armenians in a particular Istanbul neighborhood dominated shoemaking and then assumed this pattern to be true both of the whole city and of every city of the empire as well, which it was not. Indeed, in another town or city a different group dominated the same activity. In fact, in a big city such as the Istanbul capital, Armenians controlled shoemaking in one quarter while in another quarter of the city, at the very same moment, Greek shoemakers thrived. Muslims, Christians, and Jews all were active in the industrial sector of Damascus, and were well represented in the city’s famed textile industry. There, many Christians and Sunni as well as Shii Muslims wove silk and silk-cotton cloths. Sometimes one group dominated a particular craft within the general textile industry. Damascus’s dyers, for example, almost all were Christians, while those placing the warp thread on the looms, a very skilled activity, predominantly were Muslim. This is not to imply that Muslims were uniquely or more gifted than Christians, but only that they were not the dull farmers depicted by the ethnic division of labor stereotype. In the Balkan provinces, similarly diverse and non-generalizable patterns of work prevailed. In nineteenth-century Bosnia, proportionately more Muslims owned industrial firms than Catholics while Orthodox Christians were the least well represented among industrial owners. Not far away, in Montenegro, Muslims and Albanian Catholics rather than Orthodox Christian, Greek-speaking, Montenegrins dominated trade and commerce. Armenian and Greek Christians formed the majority in the silk industry of the Anatolian and Arab provinces but many Muslims and some Jews as well were employed. And, elsewhere, at Trabzon for example, both Muslims and Christians wove silk. Moreover, each of these particular patterns has a specific historical explanation. Take the vast carpetmaking sector of Anatolia, for example. Most workers had been Muslims. In the mid-nineteenth century, however, European-dominated merchant houses in Izmir began competing with Muslim firms in Uşak in west Anatolia for control of the carpetmaking business. These Izmir houses formed rural carpetmaking networks and, needing labor, they relied on their Ottoman Christian business associates who utilized existing relationships among coreligionists to provide the workers. Hence, Christian workers formed the majority of those who entered the carpet industry after 1870 while Muslims continued to occupy the older sector of the industry. Such examples show clearly that no one group dominated a particular economic activity and that the ethnic division of labor was a myth.
Occupational patterns of ethnic and religious heterogeneity also show up in labor organizations, both guilds and, at the very end of the Ottoman era, unions. These organizations’ members sometimes were drawn exclusively from one or another community. But mixed guilds were commonplace. Thus, the members of one guild might be both Christian and Muslim while another contained members of only one community. There was no general pattern. An early nineteenth-century study partially surveyed the guilds of Istanbul, revealing that about one-half of all the enumerated members belonged to mixed guilds, containing both Muslims and non-Muslims. By contrast, a listing of guilds in the city of Salonica revealed that only one-quarter of its guilds were mixed. The difference between the two cities likely derives from the fact that Salonica’s population was more homogeneous and thus had less diversity to draw on. In the empire as a whole, perhaps one-quarter to a half of all guild members belonged to labor organizations that contained members of more than one religious community.
The role of communal identities in the workplace is seen clearly when labor mobilized to present its demands, protest, or strike. In such instances, religious community affiliation sometimes seemed irrelevant and at other times important. For example, coreligionists in a guild on occasion mobilized along religious lines, even when the body as a whole was religiously heterogeneous. Take, for example, a greengrocers’ guild in Istanbul that contained both Christian and Muslim members. In 1860, some 100 members of this guild signed a petition to the government (regarding coal prices). All of the signatories on this occasion were Christians who, for whatever reasons, temporarily had banded together on the basis of their shared faith. In Aleppo, similarly, only Christian members of a mixed guild of textile merchants signed a petition in the 1840s while the tables were turned in the 1860s, when just the Muslim members petitioned. In both instances, which had no apparent religious content, the petitioners asserted that they were acting on behalf of the entire guild and not merely their coreligionists.
Unions as a form of labor organization arrived very late in the Ottoman period; some dated back to the 1880s but most evolved only after the July 1908 Young Turk Revolution. Rarely were the unions religiously homogeneous. For example, Muslim and Christian commercial employees originally organized themselves in 1908 as two separate unions but, within weeks, the two merged into a single organization. In most cases, membership of these unions was heterogeneous with many Christians and Muslims and, sometimes, Jews as well. The most important unions, and perhaps all of them, emerged in the context of foreign capital. Take, for example, the railroad unions with their Christian and Muslim members; or the Salonica-area tobacco workers’ union with Jewish, Greek, Muslim, and Bulgarian members; or the various utility company unions in Izmir, Beirut, and elsewhere, with Muslim and Christian members. The intercommunal quality of unions is vividly illustrated by a June 1909 protest meeting (against state labor policies) held in Salonica where speakers harangued the crowds in Ottoman, Bulgarian, Greek, and Ladino (archaic Spanish written in Hebrew characters).5 Salonica was noteworthy for the multi-ethnic, multi-religious character of its working class activities, some of which evolved into socialist movements.
The hiring practices of foreign corporations provide a useful tool for understanding the inter-communal tensions that became too familiar in the nineteenth-century Ottoman world. These corporations numbered in the dozens and included banks, railroads, port companies, and utilities as well as textile and food processing factories. Altogether, they employed large numbers of Ottoman subjects – more than 13,000 worked on the railroads while the Ottoman Public Debt Administration hired more than 5,000 employees. The issue here concerns the stratification of the workforce in these newly founded, often large, foreign corporations. As we have seen, there was no overall division of labor in the Ottoman workforce as a whole. But in foreign companies, over and over we find the same hiring and stratification patterns present. These corporations always hired foreigners for the very top jobs in the company, the executives who sat on the board of directors and those who most often were the department and bureau chiefs. Just below them were the Ottoman Christians who served as the middle managers and held most of the skilled jobs. Muslims rested at the bottom of these corporate hierarchies, filling the lowest-ranking, lowest-paid jobs. Moreover, in times of crisis, the corporations tended to hire disproportionate numbers of non-Muslims and foreigners, as if they distrusted Muslim employees and workers. In a roughly similar fashion, the labor unions tended to have a largely Christian leadership with a mixed, Muslim and Christian, rank and file. It should be stressed that there was nothing inherently necessary about this development. Capitalism need not generate ethnically or religiously stratified labor unions, although sometimes it has. In this particular Ottoman case, however, foreign capital interacted with the local (Ottoman) society to produce a workforce in which the coreligionists of the foreign investors were privileged. This hierarchy placed foreigners and non-Muslims in positions of superiority over Muslims and thus reversed the centuries old Ottoman pattern of Muslim political and legal predominance.
The effect of the foreign corporations’ hiring policies on the workforce of those companies stands as a metaphor for the impact of west European penetration on Ottoman society as a whole. The increasing economic, political, social, and cultural power of the West had set in motion a transformation that was overturning the existing order in the Ottoman Empire. Indeed, during the final Ottoman century, three sets of social hierarchies competed for supremacy. The first, which had existed formally for centuries until the early nineteenth-century changes, placed Muslims in positions of political and legal dominance over non-Muslims. The second, the model of the foreign corporation, began to emerge in the eighteenth century, positioning foreigners at the top, non-Muslims in the second rank, and Muslims at the bottom. The third, the Ottomanist model, called for a state administrative cadre recruited from every religious and ethnic community, ruling over a society in which all members were equal before the eyes of the law and state.
We will never know whether the new society of equality before the law or the new order of foreigner/non-Muslim superiority that the foreign corporations seemed to predict would have replaced Muslim supremacy. The old Ottoman order was fading but the new one had not yet been born. In sum, Ottoman society in the nineteenth century was undergoing an evolution; but that transformation remained unfinished because of the destruction of the empire in 1922.
Ottoman inter-communal relations, I have argued, were comparatively peaceful for most of the history of the empire. Differences among subjects always existed but only sometimes, as seen, did these lead to conflicts and violence. But, as in all societies, communal bigotry, intolerance, and violence flared intermittently for different economic, social, and political reasons. Thus, after Greek Uniates left Greek Orthodoxy and established their own church in 1701, the “hostility of the Orthodox Christians towards these perceived renegades degenerated into threats, persecution and riots in which members of one Christian sect burned down the churches of another rite.”7 In another example, Orthodox Christians in Damascus, in 1840, found the mutilated bodies of a high-ranking cleric of the Spanish monastery and his servant near some Jewish homes. And so local Christians whipped up charges of the blood libel, saying that Jews needed Christian blood for their religious rituals, forcing the arrest and torture of some wealthy Jewish merchants. Similarly, when a Greek child drowned in a river near Izmir at Easter time, local Greeks blamed the Jews and began assaulting them.8
Both the scale and the frequency of violence among Ottoman communal groups increased during the nineteenth century (see Chapter 4). Unparalleled in ferocity and scope were the attacks against the Ottoman Armenian population. These began with massacres of Armenians in 1895–1896 which were repeated in 1908, 1909, and again in 1912. In this last set of assaults, recently arrived Muslim refugees from lost provinces in the Balkans played an important role. During the Balkan Wars, vast numbers of Muslims had been driven from the European provinces, to towns such as Tekirdağ/Rodosto and Malgara on the north shore of the Marmara Sea and in Adapazarı in west Anatolia. In these places, the refugees vented their frustration and anger on hapless and innocent Ottoman Armenians. By far the worst, however, were the massacres of 1915–16. An estimated 600,000 Armenian Ottoman subjects died during and after forced deportation from their east Anatolian homes, as they moved towards the Arab provinces. These events are the centerpiece of debates around the Armenian genocide. Every year, for example, the halls of the US Congress reverberate as the Greek, Armenian, and Turkish lobbies try to win support for their respective positions for and against an official American commemoration of these World War I tragedies.
The story begins as war erupted in 1914 between Russia and the Ottomans along the east Anatolian frontier. With the Russian invaders came Russian Armenian soldiers as well as some Ottoman Armenians who had defected to the enemy. In 1915, Ottoman Young Turk ruling circles issued orders for the deportation of the entire Armenian population of east Anatolia out of the battle zone, southward to the Syrian deserts. These orders exist and can be examined and read; they are authentic materials and not forgeries or part of a hoax and are full of directives commanding the protection and care of the deportees and their properties. Order after order speaks of the need to guard the deportees and their property and assure their safety. Those deported often walked since there were few trains. As they walked, they suffered and some died of malnutrition or an accompanying disease. Others died at the hands of bandits or other Ottoman civilians who preyed on the weak. But, the solicitous state documents notwithstanding, there is abundant evidence that low and high Ottoman officers, soldiers and bureaucrats – the very persons who had the sworn responsibility to defend and protect the lives of all Ottoman subjects regardless of religion or ethnicity – murdered vast numbers of Armenian civilians, men, women, and children alike. Moreover, the patterns of the killings were chillingly similar in the various areas, powerfully suggesting the presence of a coordinated program.
How can we reconcile the orders commanding care and diligence with the murderous and apparently coordinated slaughter by state military and civil officials? Consider this assessment of the events, one that seems to be gaining acceptance among scholars on both sides of the controversy.9 There was a circle, acting like a state within the state, within the ruling Committee of Union and Progress group. Coming to power in early 1913, members of this circle secretly sought to use deportation as a guise for exterminating the Armenians. As World War I developed, they increasingly feared the potential ability of Armenian revolutionary organizations to overthrow the Ottoman state and/or the consequences of mass Armenian defections in east Anatolia to the Russians. Under the leadership of Talat Pasha, a major Union and Progress figure, the group employed the Special Organization (Teşkilat-ı Mahsusa) to carry out the massacres, outside the formal government apparatus and lines of communication. This parallel Special Organization organized and coordinated the killings, often using government officials and troops who were its members. Those who were not members or objected to the orders were overruled or replaced. The Special Organization sent directives to the many locations where the killings occurred, using its own networks rather than state channels of communication. Since the records of both the Special Organization and the Committee of Union and Progress were either lost or destroyed, this argument cannot be established without doubt. On the evidence presented, it seems plausible that high-ranking officials of the Ottoman state, utilizing the Special Organization, directed a concerted, centrally orchestrated program that murdered massive numbers of Ottoman Armenians.
Asking the question whether this crisis was the first twentieth-century genocide runs the risks of being submerged in semantic arguments and thus avoiding the real issues. After all, most Armenians died because of their presumed identity, not because of their own actions or beliefs. On the one hand, these atrocities in 1915 were not Nazi-style events that sought to concentrate and eliminate every single member of a group as such. There were numbers of Armenians outside of the battle zones who were not targeted for deportation or murder. In some of these nonconflict areas, Armenians were treated brutally but it does not appear that either the government or the Special Organization sought to deport or murder the majority of Ottoman Armenians living in western Anatolia and the Southern Balkans. In places such as Istanbul and Izmir, large Armenian communities in 1915–16 remained intact, going about their lives in the midst of war. On other hand, this situation does not gainsay the slaughter of hundreds of thousands of their compatriots in the war-torn eastern provinces. Debates now seem to be centering on the context to this slaughter in the east. The new perspective asks a different question: would such atrocities have occurred in the absence of World War I?
The fate of the Ottoman Armenians likely is linked intimately to the role that nationalism played in the destruction of the Ottoman Empire. Was the empire destroyed from within by separatist or nationalist forces or from without, by the imperial powers? This is a highly controversial question. In my own view, external rather than internal factors played the key role. The overwhelming majority of Ottoman subjects were not seeking separation or withdrawal. Rather, they would have remained within an Ottoman state framework had that political entity continued to exist into the 1920s and 1930s.
To be sure, important changes in personal and group identities were occurring. During the nineteenth century, ethnic identities were becoming more important while the designators “Muslim” and “Christian” became more complicated. Earlier, in the eighteenth century, the Greek Orthodox clergy had eradicated many formerly separate clerical institutions in the Balkans and brought them under its own hegemony. Hence, in 1766, it suppressed the Serbian patriarchate of Peç and, in 1767 the Bulgarian archbishopric of Ohrid. Similarly, the patriarchate of Antioch gradually became the dominion of Greek prelates. Thus, by the end of the eighteenth century, Greek Orthodoxy reigned. That is, at the end of the eighteenth century, the term Greek Orthodox Christian covered many Christian groups of quite different ethnicities.
During the nineteenth century, ethnic distinctions became more important among Ottoman Christians in the Balkans, a process accelerated by the emergence of the separate church organizations. Indeed, the separatist movements of the nineteenth century often fought against Greek ecclesiastical and cultural imperialism as much as Ottoman rule. In 1833, after the formation of the Greek state, an autocephalous Greek Church emerged there, while in the same decade a separate Serbian Church similarly followed upon formation of a Serb state. Later on, a Bulgarian exarchate emerged in 1870 and then an autocephalous Rumanian Church in 1885. Thus, each separate Church sought to create or reinforce a sense of separate ethnic, for example, Serb or Rumanian, identity: the “Orthodox Church” went from embracing almost all Orthodox subjects to, largely, only the ethnically Greek ones. At the same time, nationalists in the various communities worked to purify the various languages of “alien” elements. Hence, for example, Greek nationalists worked to eradicate the Turkish spoken by many Ottoman Greeks. In sum, there can be little doubt that new notions of separateness were at work in the Ottoman Balkan world.
And yet, in common with developments elsewhere in the globe, nationalist movements in the Ottoman Empire were minority movements, orchestrated and promoted by a few. In (probably) every case of successor state formation in the Ottoman Empire, state preceded nation and not the other way around. The formation of independent states derived not from groundswell movements but rather from the actions of certain groups in the societies who sought economic and/or political privilege that they believed they could not obtain under Ottoman domination. That is, a relative handful of individuals established a government apparatus, drew boundaries on a map and prepared the national flag and anthem. With these in place, the creation of a national community actually began to be based on a shared feeling of being Bulgarian, Serbian, Greek, etc. In the Balkan lands, Russia, Austria-Hungary, Britain, and/or France supported these aspirations since they believed (usually correctly) that the new states were likely to fall under their own respective influence. Throbbing in the breast of every Christian in the Balkans was not the idea of breaking away from the Ottomans. The foundation of independent Balkan states in the nineteenth century is no proof of mass discontent with Ottoman rule on the part of the Balkan Christian subject populations. Their creation, however, is testimony to the determination and organizational skills of the separatists and the assistance of the Great Powers. On this basis, they created the new states and within them began constructing the new nationalities, often using the foil of the “savage infidel.”
We also need to understand the unimportance, until after World War I, of Arab, Turkish, and Kurdish nationalisms on the territories that remained under Ottoman sovereignty. Here, too, the basic point deserves reiteration: most Ottoman Muslims of whatever ethnicity remained fundamentally content with Ottoman rule and did not actively seek separation.
Several issues are important here. First, the nineteenth-century state-supported ideologies of Ottomanism and pan-Islamism were failing to protect the empire: territories continued to fall away. Nonetheless, Ottoman state elites, including the Young Turks who came to power after 1908, by and large remained loyal to Ottomanism and did not opt for Turkish nationalism, although it is often alleged that they did. It is true that some leaders, after 1908, personally pursued a new cultural identity as Turks and came to believe in Turkish superiority to others. And yet, they and their political party continued to argue for and promote the imperial policies of Ottomanism and pan-Islamism. And it is also true that, despite the personal secularist tendencies of many Young Turks, the Islamist component of Ottoman identity became more important after 1908 because of the accelerating dismemberment of the (largely Christian) European provinces of the empire. Within months of the 1908 revolution that had promised an end of territorial dissolution, many lands nominally still Ottoman became formally separate or independent: Bulgaria, Crete, and Bosnia-Herzegovina. Such fragmentation meant that, in 1914, the majority of remaining subjects were Muslims, mainly ethnic Turks, as well as Arabs and Kurds, although considerable Christian Armenian and Greek populations remained. It nonetheless is clear that a secularist, Ottomanist world view prevailed among the Young Turks, who remained determined to mold a new identity among subjects. One measure of their effort to create this common Ottoman identity is the Election Law, passed after the 1908 Revolution, that sought to eliminate representation by religious community and replace communal politics with party politics. Overall, the post-1908 policies of the Ottoman regimes reflected strong centralization policies, a pressing for close control and an imposition of uniform, standardized, imperial standards rather than Turkish nationalism.
How, then, can we explain the accusations of Armenian and Arab nationalists of our own day, that the Young Turk Ottoman regimes were harshly Turkish nationalist? They point, for example, to the famous Young Turk leader Cemal Pasha who executed a group of local notables in Damascus during World War I. And, most significantly, they recall the Armenian massacres of 1915–1916. Rather than viewing these as the actions of fierce Turkish nationalists aimed at Turkish racial dominance over others, it may be more accurate to see them as policies enacted by centralizing state officials ruthlessly determined to stamp out threats to its stability. In the first case, the hangings reflected the Istanbul government’s relentless determination to impose and maintain control over Damascene notables who were trying to replace central authority with a decentralized regime that they themselves would lead. Regarding accusations that the regime was pro-Turkish, consider that the post 1908 Young Turk regimes aggressively recruited more Arabs into the state apparatus than at any other time, except the reign of Sultan Abdülhamit II, who was exceptional in this regard. In the second case, the Armenian massacres, the state may have killed not from racialist or nationalist reasons but rather because it feared the Armenians as actual or would-be rebels seeking to break with Ottoman control and ally with enemies of the government. The state warred against its own subjects; but it was not anationalist civil war among competing groups of equals or near equals.
Neither Turkish nor Arab nor Armenian nor Kurdish nationalism pushed a dying Ottoman state over the nationalist cliff after 1914. Indeed, there were scarce few of these sentiments during the final decade of the Ottoman Empire. Some Armenians did call for a separate nation state but the overwhelming majority continued to opt for the Ottoman system. Very few Kurds spoke of autonomy. Similarly, most Arabs acted as if they expected to remain within the Ottoman polity, although it is true that a few leaders sought a separate cultural identity and promoted regionalism with greater autonomy within the Ottoman imperial system. In sum, the vast majority of Ottoman subjects in 1914 – of whatever religion and ethnicity – were not seeking to break away but instead retained their identities as Ottoman subjects.
A key to understanding the accusations of Turkish xenophobia and nationalism lies in the Middle Eastern events following World War I. The Great Powers forcibly dismantled the empire. Britain and France divided the Arab provinces between them, setting up “mandatory regimes” under their own supervision within the League of Nations framework and ruling these regions in various guises until the mid 1950s. They had intended to hand over a large chunk of Anatolia to their protégés in Athens and to leave a rump Ottoman state. Instead, Ottoman resistance forces gathered and, unable to restore the empire, settled on founding a smaller state in its Anatolian fragment, one that later became the Turkish nation state. In both the Arab and the Anatolian areas, nationalist movements after the Ottoman demise worked to create nations in the states that had emerged from the imperial debris: notably Turkey, Syria, Lebanon, Iraq, Jordan, and the special case of Palestine. Leaders respectively were working to create and propagate Turkish and Arab nationalist identities. Each found it useful to invent, find, or magnify – for quite different reasons – the Turkish nationalist elements that were present in the late Ottoman period. For the Turkish group of state and nation builders, who viewed these elements positively, finding Turkish nationalism in the Ottoman era served to legitimate the new Turkish state and gave it historical roots. For Arab state and nation builders, Turkish villainies both reinforced and helped justify their own separate state identity. And, perhaps, such villainy made more palatable the Great Power occupation that had ensued without their consent after World War I. Ironically, this anti-Turkish interpretation also helped Britain and France to justify their destruction of the empire itself. Thus, insisting on the presence of significant Turkish nationalism before 1918 promoted many post-World War I agendas, including that of Britain, France, the Turkish Republic, and the Arab politicians and intellectuals struggling to gain independence from the Great Powers.
Entries marked with a * designate recommended readings for new students of the subject.
Adanir, Fikret. “The Macedonian question: the socio-economic reality and problems of its historiographic interpretations,” International Journal of Turkish Studies, Winter 1985–6, 43–64.
Ahmida, Ali Abdullatif. The making of modern Libya: State formation, colonization and resistance, 1830–1993 (Albany, 1994).
Akarlı, Engin. The long peace: Ottoman Lebanon, 1861–1920 (Berkeley, 1993).
Anastassiadou, Meropi. Salonique, 1830–1912. Une ville ottomane à l’ăge des reformes (Leiden, 1997).
*Andric, Ivo. The bridge on the Drina (Chicago, translated edition of 1945 Serbo-Croat original, 1977).
Braude, Benjamin and Bernard Lewis, eds. Christians and Jews in the Ottoman Empire, 2 vols. (London, 1982).
Cleveland, William. The making of an Arab nationalist: Ottomanism and Arabism in the life and thought of Sati al-Husri (Cleveland, 1971).
*Cole, Juan. Colonialism and revolution in the Middle East: Social and cultural origins of Egypt’s Urabi movement (Princeton, 1993).
Davison, Roderic. “Nationalism as an Ottoman problem and the Ottoman response,” in William W. Haddad and William Ochsenwald, eds., Nationalism in a non-national state: The dissolution of the Ottoman Empire (Columbus, 1977), 25–56.
Edib, Halide. Memoirs (London, 1926).
Hasluck, F. W. Christianity and Islam under the Sultans, 2 vols. (London, 1925).
Hovannisian, Richard G., ed. The Armenian people from ancient to modern times, II: Foreign dominion to statehood: The fifteenth century to the twentieth century (New York, 1997).
Kahane, Henry, Renée Kahane, and Andreas Tietze. The lingua franca in the Levant: Turkish nautical terms of Italian and Greek origin (Urbana, 1958).
Karpat, Kemal. The politicization of Islam: reconstructing identity, state, and community in the late Ottoman state (New York, 2001).
*Kayalı, Hasan. Arabs and Young Turks: Ottomanism, Arabism and nationalism in the Ottoman Empire, 1908–1918 (Berkeley, 1997).
Kevorkian, Raymond H. and Paul B. Paboudjian, eds. Les Arméniens dans l’empire ottoman à la veille du genocide (Paris, 1992).
Levy, Avigdor, ed. Jews Turks, Ottomans: a shared history, fifteenth through the twentieth century (Syracuse, 2001).
*Lockman, Zachary. Workers and working classes in the Middle East (Albany, 1994).
*Marcus, Abraham. The Middle East on the eve of modernity: Aleppo in the eighteenth century (New York, 1989).
*Quataert, Donald, ed. Workers, peasants and economic change in the Ottoman Empire 1730–1914 (Istanbul, 1993).
Rodrigue, Aron. French Jews, Turkish Jews: the Alliance Israélite Universelle and the politics of Jewish schooling in Turkey, 1860–1925 (Bloomington, 1990).
*Suny, Ronald Grigor, Engin Deniz Akarlı, Selim Deringil, and Vahakn N. Dadrian, “Exchange” Armenian Forum: A journal of contemporary affairs, Summer 1998, 17–136.
Tibi, Bassam. Arab nationalism: A critical inquiry (New York, translation of 1971 German original, 1981).
*Tunçay, Mete and Erik Zürcher, eds. Socialism and nationalism in the Ottoman Empire, 1876–1923 (London, 1994).
*Vatter, Sherry. “Militant textile weavers in Damascus: waged artisans and the Ottoman labor movement, 1850–1914,” in Donald Quataert and Erik J. Zürcher, eds., Workers and the working class in the Ottoman Empire and the Turkish Republic, 1839–1950 (London, 1995), 35–57.
Zürcher, Erik. The Unionist factor: the role of the Committee of Union and Progress in the Turkish nationalist movement of 1905–1926 (Leiden, 1984).
*Turkey: A modern history, 3rd edn (London, 2004).
1The quotations provided from the oral interviews conducted in Bulgaria by Barbara Reeves-Ellington.
2Interview with Simeon Radev, 1879–1967, describing his childhood before 1900, provided by Barbara Reeves-Ellington.
3Interview with Iveta Gospodarova, personal narrative, Sofia, January 19, 1995, provided by Barbara Reeves-Ellington.
4Paul Dumont, “Jewish communities in Turkey during the last decades of the nineteenth century in the light of the archives of the Alliance Israélite Universelle,” in Benjamin Braude and Bernard Lewis, eds., Christians and Jews in the Ottoman Empire (London, 1982), I, 221.
5Yavuz Selim Karakışla, “The emergence of the Ottoman industrial working class, 1839–1923,” in Donald Quataert and Erik Zürcher, eds., Workers and the working class in the Ottoman Empire and the Turkish Republic, 1839–1950 (London, 1995), 19–34.
6There is an enormous scholarship on this subject, much of it polemical but some of it increasingly constructive. See the Hovannisian and Suny works cited in bibliography for this chapter as well as studies by Michael Arlen, Michael M. Gunter, Heath Lowry, Robert Melson, and Justin McCarthy for different points of view.
7Youssef Courbade and Philippe Fargues, Christians and Jews under Islam (London and New York, 1997), 69.
8For example, Lucy M. J. Garnett, The women of Turkey and their folk lore (London, 1890), 6–7.
9Derived from Erik J. Zürcher, Turkey: A modern history, 3rd edn (London, 2004), 114–17. This line of argument also is present in some of the writings noted in Suny et al., above.