438

 CHILDREN AND SLAVERY IN THE OTTOMAN EMPIRE

Anahtar Kelimeler: Osmanlı Köleligi, On Dokuzuncu Yüzyıl, Çocuk Köleler, Elit
Haneler, Istanbul.
Bu çalısma, Osmanlı Imparatorlugu’ndaki çocuk köleleri ele almaktadır. Öncelikle
modern dönem öncesi Osmanlı kölelik sistemlerinden biri olan devsirme yöntemi
ile çocuk köle edinimini kısaca tartısarak, köle seçimi ve egitilmesi hakkında diger
yüzyıllar ile arasında muhtemel bir sürekliligin var olup olmadıgını sorgulamaktadır.
Ardından, on dokuzuncu yüzyıla geçerek Kafkas ve Afrika kökenli çocukların elit
hanelerce satın alınmadan önce hangi süreçlerden geçerek kölelestirildiklerini ve nasıl
Istanbul’a getirildiklerini göstermektedir. Bunu yaparken, on dokuzuncu yüzyıl Osmanlı
Imparatorlugu’nun kölelik ve köle ticareti kosullarını göz önünde bulundurmaktadır
. Daha sonra, Istanbul’a getirilen çocuk kölelerin neden bazı elitler tarafından
tercih edildigini tartısmaktadır. Varlıgı tespit edilen bu çocukların istihdam
alanlarını anlatmaktadır. Ayrıca, köle çocuklara olan talep, onların istihdamları,
sahipleri ve bu konu üzerinde yazılan seçilmis birincil kaynaklar degerlendirilirken
toplumsal cinsiyet perspektifi kullanılmaktadır. Son olarak, bu tez çocuk köleler ve
seçkin sahipleri arasındaki iliskiyi analiz ederek, geleneksel köle-efendi iliskisinden
ziyade aralarında karsılıklı bir bagımlılık oldugunu, ancak bu karsılıklı bagımlılıgın
asimetrik oldugunu savunmaktadır.
v
ACKNOWLEDGEMENTS
I completed this thesis during the COVID-19 pandemic with the help and support
of many people and institutions. First and foremost, I would like to express my
sincere gratitude to my advisor Hakan Erdem, who patiently guided me. As his
last student before retirement, I consider myself fortunate to write my thesis under
his supervision. I also owe special thanks to my jury members Aksin Somel and
Nazan Maksudyan, whose insightful feedback and criticism allowed me to sharpen
my thinking and brought my work to a higher level. I also want to thank Tülay
Artan and Ayse Ozil for their intellectually rich courses and stimulating discussions
in the History department of Sabancı University.
I need to acknowledge that I have received a scholarship partially from Sabancı
University and Türk Tarih Kurumu (TTK), the Turkish Historical Society, for which
I am very thankful. When I was writing my thesis, I had a chance to receive
valuable contributions from various great scholars. Therefore, I would like to extend
my appreciation to Hülya Canbakal, who was always available and willing to help
me and support me in every regard whenever I needed it. I am indebted to her
guidance and feedback, which enhanced my knowledge and perspective. Likewise, I
am thankful to Aysel Yıldız, whose inspiring history classes assisted me in discovering
the irresistible appeal of history and enhanced my interest in the field. It was the
History of Civilizations class of 2015 that made me first fascinated in the subject of
slavery. Since then, I have greatly benefited from her continuous support, guidance,
and insights that have been invaluable for the betterment of the text.
I need to express my gratitude to my professors at Istanbul Medeniyet University,
where I completed my undergraduate studies. I am thankful to Elif Bayraktar-
Tellan and Ismail Hakkı Kadı, for their undying support, generosity in sharing
their knowledge, and intellectual encouragement throughout my undergraduate and
graduate years.
No words can adequately describe my gratitude to Damian Pargas, whose academic
excellence and diligence I admire. He always guided and inspired me since I conducted
my Erasmus internships at Leiden University in 2017 and 2018, when I was
acquainted with some major works of prominent scholars that widened my knowledge
and deepened my interest in slavery studies. I am also indebted to him for introvi
ducing me to Bonn Centre for Dependency and Slavery Studies (BCDSS), where I
have met great scholars of slavery studies and attended meaningful academic events.
Perhaps, it was the largest upside of the COVID-19 pandemic that I could join numerous
scholarly events organized by the BCDSS. There, I am incredibly thankful
to Veruschka Wagner for what I have learned from her. Her encouragement, unwavering
suggestions, and guidance have been invaluable for me. I also want to note
my most sincere appreciations to Fırat Yasa, Zeynep Yesim Gökçe, and Turkana
Allahverdiyeva for their helpful contributions to my work.
Friends made the process of research and writing a blast. I appreciate Elif Çonker
and Talha Katırcı’s companionship and incalculable support during every stage of
this thesis. I am blessed for Halide Kahraman’s friendship, whose continuous support
throughout the happiest and darkest of times helped me complete my thesis. They
all cheered me up and assisted me out during the painful times of the writing process.
I feel grateful for having Yagmur Yurtsever’s companionship, as I often shared my
concerns and worries with her.
More importantly, I wish to thank my family, who always financially and emotionally
supported me throughout my education and research. I would like to give a special
thanks to my mother, who devoted herself to my education and provided me with
the opportunity to pursue my dream. I have to acknowledge that this study could
not be accomplished without the lasting comfort of Ammar Omar. You have indeed
been on this ride with me from day one. I am thankful for your endless love and
belief in me.
vii
TABLE OF CONTENTS
1. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. A REVIEW OF LITERATURE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.1. Global Slavery Studies: Pioneering Scholarship . . . . . . . . . . . . . . . . . . . . . . . 7
2.2. Ottoman Enslavement Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2.1. The Demise of Slavery and the Slave Trade . . . . . . . . . . . . . . . . . . . . 11
2.2.2. A Gendered Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.3. Slavery in Legal Texts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2.4. Individual Cases of Ottoman Enslavement . . . . . . . . . . . . . . . . . . . . . 17
2.3. Global Child Slavery Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.4. Ottoman Child and Childhood Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3. CLASSICAL OTTOMAN SLAVERY SYSTEMS BEFORE MODERNIZATION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2. Pencik System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3. Devshirme System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4. CHILD SLAVERY IN NINETEENTH-CENTURY OTTOMAN
EMPIRE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2. The Concept of a Child . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.3. Methods of Child Enslavement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.3.1. Kidnapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
4.3.2. Capturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.3.3. Tribute, Exchange, and Tax Paying . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.3.4. Abandonment and Sale of Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.3.5. Enslaved at Birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
viii
4.4. Arrival of Slave Children to Istanbul . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.4.1. Nineteenth-Century Ottoman Slave Trade . . . . . . . . . . . . . . . . . . . . . 49
4.4.2. Slave Bazaars and the Purchasing of Slave Children . . . . . . . . . . . 51
4.4.3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
5. CHILD SLAVES IN ELITE HOUSEHOLDS IN NINETEENTHCENTURY
ISTANBUL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
5.2. Households of the Ottoman Empire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.2.1. The Imperial Household. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
5.2.2. Elite Households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
5.3. Child Slaves in the Elite Households of Istanbul. . . . . . . . . . . . . . . . . . . . . . . 60
5.3.1. Characteristics of Slave Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.3.2. Education of Slave Children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.3.3. Purposes of Purchasing Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.3.3.1. Marrying Slave Children Off with Fine Gentlemen. . . 65
5.3.3.2. A Tool of Investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.4. Employment of Slave Children in Elite Households: Playmate-Gifts . . 69
5.5. Relationship Between Slave Children and Elite Owners . . . . . . . . . . . . . . . 72
5.6. Gender Aspect: Slave Girls, Female Owners, and Women Writers . . . . 74
5.7. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
6. CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
BIBLIOGRAPHY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
ix
1. INTRODUCTION
“They are sold usually at about twelve or thirteen years of age, but there
are cases of sales at the early age of six or seven. This happens, however,
only a lady wishes to bring them up as her slaves, either to accustom
them to her service or to resell them at a profit when they are older.”
(Melek Hanım 1872, 158-159)
The wife of Grand Vizier Mehmed Emin Pasha of Cyprus, a nineteenth-century
Ottoman elite, notes that six- or seven-year-old child slaves were bought to be raised
by some elite women and then sold at higher prices. This reveals that some Ottoman
elite women were involved in the upbringing and sale of slaves and earned financial
profits from it. Although the existence and significance of the unfree labor of child
slaves in the Ottoman Empire was not entirely left out of the literature up to date,
it was not investigated extensively. The main reason for this is that the field of
Ottoman slavery studies is still growing, but also that Ottoman child studies is a
new, relatively less explored research area. This thesis aims to combine Ottoman
slavery and child studies to investigate the enslavement of children in the Ottoman
context. Accordingly, this study benefited from the existing global child slavery
scholarship, which recently started attracting the attention of historians.
By the 2000s, scholars had started to turn their focus to the experience of children
in bondage. The main themes amongst the existing studies on child slavery are
the trade of slave children and their social, political, and domestic roles. Nevertheless,
the growing child slavery literature has been dominated primarily by studies
on North America before emancipation, and the transatlantic slave trade. Scholars
concentrated on topics such as child soldiering (David M. Rosen 2005), the childhood
and childcare of slaves (Damian A. Pargas 2011), slave children given as gifts
(Sarah Winter 2017) or playmates (Jonathan Blagbrough and Gary Craig 2017),
children’s knowledge of and participation in the abolitionist movement (Audra A.
Diptee 2017), interactions between enslaved and free children, slave-born children
1
in North America, and African children the transatlantic trade (Wilma King 1995).
Lately, this growing literature has begun to consider various time periods and places
beyond the translatlantic slave trade; for example, research has revealed information
about conscripted children in the Ottoman Empire (Gülay Yılmaz 2009) and singing
slave girls in the Abbasid court (Kristina Richardson 2009).
Scholars quickly encountered some of the problematic aspects of studying children
in bondage: the problem of defining a child slave (Anna Mae Duane 2017) and the
need to distinguish between child slavery and child labor (Gary Craig 2017).
Even though Ottoman slavery literature is not at an embryonic level as it once was,
other forms of child enslavement in the Ottoman Empire had still not yet been
comprehensively addressed outside of the devshirme boys until recently. Since the
1980s, modern Ottoman slavery research has been published; thus, it is possible
to discuss and build upon generations of research today. The subjects and themes
that formed the fundamental axis of Ottoman slavery literature have diversified;
shortly and roughly, scholars have investigated the end of slavery and the slave
trade as well as its various forms, which mainly are domestic, kul and harem slavery,
slave ownership, slave agency, and the social and economic roles of slaves in various
places and eras. Each study has been based on different types of sources: some
utilized archival documents, while others made use of travelogues and memoirs.
These researched works intend to make the voices and stories of individual slaves
heard or to outline slavery as an institution. In almost all the conducted studies, the
smallest and most powerless subjects of the institution of Ottoman slavery, namely
children, were not fully neglected but only briefly mentioned here and there. They
pointed out that some children were victims of kidnapping and capturing, and thus
were forced into slavery. Additionally, they revealed that some elite members of
the Ottoman Empire were purchasing young boys and girls for years, similar to
the longstanding practice within the Imperial Palace. Expanding on this budding
research, I intend to dig deeper into the topic of slave children in the late Ottoman
Empire.
As an history student, it stroke my attention that there has been no comprehensive
work established on child slavery in the Ottoman context yet even though there
was a considerable demand for child slaves by some Ottoman elite. Child slaves
in the Ottoman Empire have primarily been analyzed within the context of the
devshirme (child levy) practice since child slaves were the main subjects of this
long-lasting system. There have been numerous studies conducted and debates held
on this topic. One of the main discussions is about whether collecting children to
be recruited for service in the Imperial Palace or the Janissary Corps (the Sultan’s
2
private troops) is compatible with Ottoman Sharia law since the levied children
were selected from Ottoman subjects. Research conducted to date has focused on
the ways in which these children were selected and how the process operated. The
levied children fall under the category of child slavery, as they began to act as the
“Sultan’s servants” after their conscription. This is to the point that when the term
“child slave” is mentioned in the Ottoman context, the devshirme boys rightly comes
to mind first.
Up until now, no framework has been established to specifically understand slave
children in the Ottoman context outside of those collected through the devshirme
method. In the Ottoman slavery literature, where different generations can be recognized
today, child slaves have been deprived of necessary scholarly attention and
still remain an understudied subject even though they were included in various forms
of enslavement in many fields in Ottoman society. With this in mind, this dissertation
intends to reveal the areas where child slaves were employed, why child slaves
were favored over adult slaves, and finally, which social group preferred child slaves
in the nineteenth century Istanbul. Thus, this study examines enslaved children of
Circassian and African origin working in some Istanbulite elite households.
Considering the limitations and scope of the study, this thesis does not aim to fully
discuss all the child slaves in the Empire. The present thesis confines itself solely
to some elite households located in the nineteenth-century Istanbul, where enslaved
children of Circassian and African origin were put to work. While information
on adult slaves in the nineteenth-century Ottoman Empire was still challenging to
find, sources on children were even further limited. The nineteenth century was
compulsorily chosen for this study as the available primary sources, mainly foreign
travelogues and native memoirs that this thesis mainly relies its arguments on, were
penned in this century. Countless visitors were motivated to come to Istanbul from
various states and empires for an array of purposes. In other words, the nineteenth
century could be regarded as the century of travelogues and memoirs; in this century,
not only some foreign observers logged their memories in or observations of Istanbul,
but some of the domestic elite kept memoirs as well. Numerous travellers, diplomats,
and visitors had been to Istanbul, encountered some of the elites of the city, and were
invited to their households in which they were able to observe child slavery. These
travelogues and priceless memoirs written by some elites of the Ottoman Empire
have been utilized exhaustively to reveal the perspectives on the phenomenon of
slave children in the capital city. In light of these primary sources, this thesis aims
to present an outline of the existence of enslaved children, their employment, and
their status in some of nineteenth-century Ottoman Istanbul’s elite households.
3
In doing so, this study intends to contextualize the conditions of the chosen time
period. For example, in the nineteenth century, an anti-slavery movement was instigated
under the leadership of Great Britain and pressure was put on many countries
and empires to put an end to the slave trade, one of them being the Ottoman Empire.
As a result, the Ottoman central administration took some specific measures
to curtail the slave trade in several areas:
“This pattern yielded the prohibition of the slave trade in the Persian
Gulf in 1847, the temporary prohibition of the traffic in Circassians and
Georgians in 1854-1855, the general prohibition of the African slave trade
in 1857, and the Anglo-Egyptian and Anglo-Ottoman conventions for the
suppression of the slave trade in 1877 and 1880, respectively.” (Toledano
1993, 484)
Additionally, the Ottoman Empire was in the midst of the CrimeanWar (1853-1856)
with the Russian Empire. It is known that the CrimeanWar drastically increased the
volume of the white slave trade through the Black Sea. Some people participating in
the slave trade made significant profits from the turbulent environment in the region.
In addition, significant migration from the Caucasus region to different parts of the
Empire, but especially to the capital, had begun. Among the migrants, there were
young girls and boys whose paths somehow led them to Istanbul. This carried on
in the decades following the Crimean War. Despite the ban on the African slave
trade in 1857, it was still possible to find black child slaves in Istanbul throughout
the nineteenth century.
Children who were captured and imported from farther lands and brought to Istanbul
were purchased by some Istanbulite elites. They were nurtured and educated in
these households. Their presence can be examined from many angles and perspectives.
This thesis recognizes that the institution of slavery in the Ottoman Empire
had child members as young as age six or seven, and that their labor was utilized
for domestic servitude in some elite households of Istanbul. This domestic service
was different from what typically comes to mind when forced domestic work is mentioned,
namely cleaning or cooking. The domestic servitude that some slave children
were forced into consisted of playing games if they were given as birthday presents
to free children born to the elites of the city.
On some occasions when slave children were not given as presents, they were nurtured
and educated in the household. They were not trained for no reason, but
to be married off to fine gentlemen or to be sold at higher prices. In this regard,
this thesis proposes that slave children were often a means of investment during the
4
course of the nineteenth century.
Analyzing the relationship between child slaves and their elite owners, this study
suggests that acquiring child slaves provided some Istanbulite elites a degree of
prestige, social work, and a network, while being raised by an elite from a young
age was also a factor that increased the value of a slave on some occasions.
This dissertation is organized in the following way: Chapter 1 intends to discuss
the devshirme system as one of the classical Ottoman slavery systems utilized before
modernization to provide background information on the topic of Ottoman
child slavery before hastily starting with the child slaves of the nineteenth century.
Nevertheless, it is necessary to understand why the devshirme system emerged and
was needed before discussing the details of the system itself. Considering that it
is necessary to look at the pencik system for this, the first chapter will briefly describe
the pencik system, another classical Ottoman slavery system. The chapter
will cover how it was formed, how it was developed, and why it did not work to the
point that a new system was needed, which was the devshirme method. The chapter
will then discuss the devshirme practice’s introduction, early implementation,
and compatibility with Ottoman Sharia law. In the process, secondary literature
often will be summarized, and some Ottoman chronicles will be utilized as primary
sources. This section will show that the devshirme system would not remain as it
was originally created but would be developed over time. The modifications made
for its development, especially some characteristics that were taken into account in
the selection of children, will be discussed thoroughly. Chapter 1 points out that
some characteristics that were sought in choosing the devshirme boys might have
resembled the features that the nineteenth century elites looked for in slave children.
To what extent this could be considered a continuum from the devshirme method
will be questioned.
The following chapters of the present thesis concentrate on the nineteenth century.
Chapter 2 moves on to discuss enslaved children in the nineteenth century and
attempts to explain what children went through before entering the elite households
of Istanbul. First and foremost, it discusses what the concept of a child meant for
the Ottomans and how the Ottomans recognized children before and during the
nineteenth century. In this light, it examines the process of how some children fell
into slavery and ended up in Istanbul in the nineteenth century. It aims to present
the ways children were captured and enslaved, the most likely used slave trade
routes, places where child slaves were most possibly sold, and finally, by whom they
were sold in the nineteenth century Istanbul.
Chapter 3 examines the child slaves possessed by some Istanbul elites in the nine-
5
teenth century. Before doing this, it briefly describes the households of the Empire
and the emergence of the elite households. Then, it discusses which characteristics
were considered when choosing Caucasian and African children to take for various
purposes. Here, the chapter deals with the question of the continuum from the
devshirme method mentioned earlier in the first chapter. Diving deeper, it explains
the reasons why these slave children, most of whom were girls, entered the household,
what duties awaited them after entering, and what kind of education they
went through from a gender perspective. Then the chapter moves on to investigate
what happened to the child slaves who were raised and trained. It finally analyzes
the relationship between the slave children and their elite owners and suggests that
there was an interdependent, asymmetrical relationship between them due to the
power the owners obtained of making critical decisions over their slave children.
6
2. A REVIEW OF LITERATURE
2.1 Global Slavery Studies: Pioneering Scholarship
The first wave of slavery studies in the Islamic world started in the 1980s following
the leading works of Bernard Lewis and Robert Brunschvig (Brunschvig 1960, Lewis
1971). The first wave examined archival sources and outlined essential elements that
made up the Ottoman slavery institution such as manumission and enslavement, as
well as areas of slave employment (Toledano 2011, 208).
Three decades ago, Bernard Lewis penned his book, Race and Colour in Islam,
when there was no inclusive scholarship on slavery in the Middle East other than
studies constructed based on European sources or translated material (Lewis 1971.
Bacharach 1992, 305). Lewis’s work concentrates on the twelfth-century Islamic
world, briefly touching upon the slave trade and the abolition of the institution.
The work has since been revised and was translated to French in 1982, as Race Et
Couleur En Pays d’Islam (Paris: Payot, 1982). In 1990, Lewis extended his work
and published it with the title Race and Slavery in the Middle East, which reflects
the evolution of his thoughts over the last twenty years (Lewis 1990). This time,
the revised book presents a brief historical outline of the institution of slavery in
the Middle East from ancient to modern times. Lewis claims that “slavery has
influenced both race relations and racial perspectives in the Middle East.” (Morony
1995, 70). Additionally, he suggests that “there is no evidence for any anti-black
prejudice neither in pre-Islamic Arabian society nor in the Qur’an; however, antiblack
bias did develop historically among Muslims” (Morony 1995, 70). According
to Lewis, most of the slaves shipped into the nineteenth-century Ottoman Empire
“were Africans because the sources of white slaves were cut off, which meant that
black slaves could obtain a higher status than they could before.” Lewis put forward
that black slaves were treated well by their Muslim owners (Morony 1995, 70). The
good, proper treatment of slaves is a significant theme that other Ottoman historians
7
argued further . Zilfi 2010, 123. Faroqhi 2002, 139-151).
In the decade after Bernard Lewis wrote his pioneering work on the concepts of
slavery and race in the Islamic world in 1971, the field of slavery studies began
to flourish. In other words, the early 1980s witnessed works written by renowned
scholars and the various forms and systems within the institution of slavery had
begun to be investigated on a global level.
Firstly, a prestigious academic journal entitled Slavery and Abolition, whose editorial
advisory board was composed of leading names in the field such as Joseph Miller,
Orlando Patterson, David Brison Davis, and Suzanne Miers, published its first issue
in 1980. The first issue of Slavery and Abolition includes prominent Ottoman
historian Alan Fisher’s well-known article entitled “Chattel Slavery in the Ottoman
Empire” (Fisher 1980). The author begins by stating that inadequate attention
had been paid to slavery systems in the Islamic world, and therefore the history of
the Ottoman Empire was left least studied. Fisher cites available sources related
to the institution of slavery and its trade and explains slavery-related taxes and
state records. As for the provincial law codes concerning taxes on slave sales, Fisher
points out that there were several patterns in different cities such as Salonika, Basra,
Bolu, Diyarbekir, Mardin, Erzincan, and Damascus, as well as cities in Bosnia and
Egypt (Fisher 1980, 80).
Moses I. Finley published four lectures that were taught at the Collège de France
entitled Ancient Slavery & Modern Ideology in 1980. In his lectures, he tackles four
essential themes: the ways in which ancient slave societies came into being, how slavery
functioned in the ancient market and politically, how it was perceived socially
and morally, and finally, what contemporary scholars have made of it. Finley introduces
a dichotomy in his book, dividing societies across time into “slave societies”
and “societies with slaves.” Finley puts forward five slave societies: ancient Greece,
ancient Rome, the colonial Caribbean, Brazil, and the American South. Finley defines
a slave society as “an institutionalized system of large-scale employment of slave
labor in both the countryside and the cities” (Finley 1980, 67), not solely a society
where slavery exists. As a scholar of Ottoman and Islamic societies in early modern
and modern times, Toledano finds the distinction Finley makes between “slave societies”
and “societies with slaves” valuable and significant in differentiating societies
“from the better known, more familiar, Atlantic world societies.” However, Toledano
doubts the model’s applicability to all societies (Toledano 2018, 362).
In 1982, Orlando Patterson published his well-known work Slavery and Social Death:
A Comparative Study in which he compared the internal dynamics of slavery in
numerous societies from ancient to modern times. The author ponders questions
8
related to what slavery was and what it was like. In his groundbreaking study,
Patterson (1982, 13) defines slavery as “the permanent, violent domination of natally
alienated and generally dishonored persons.” Patterson (1982, 13) further argues that
the process of enslavement began with kidnapping people. The slaves lost their social
identities after their enslavement began; they were given different names and lost
their kin relations and social bonds (Patterson 1982, 67), which led to their social
death (1982, 13). Numerous scholars concentrating on the field of slavery studies
have benefitted from his arguments. It should be noted that this model does not
apply to all societies. For instance, Ehud Toledano argues that none of these points
apply to enslaved people in the kul/harem system of the Ottoman Empire (Toledano
2017, 137).
David Brion Davis penned his influential work entitled Slavery and Human Progress
in 1984, in which the author examined the notions people held about slavery and
emancipation. Juxtaposing “slavery” and “human progress,” Davis discusses the
perceptions people held on slavery and emancipation on a global scale from ancient
times to the twentieth century (Davis 1984). In addition to these pioneering works,
countless models were introduced in the field of slavery. Among these, it is noteworthy
to mention Dale Tomich and Michael Zeuske’s “second slavery” (Lavina &
Zeuske 2014) and Jeff Fynn-Paul’s “slaving zones” (Fynn-Paul 2009).
2.2 Ottoman Enslavement Literature
Ottoman historiography was also influenced by the blossoming slavery literature of
the early 1980s. The initial slavery research was mainly concentrated on the economic
aspect of the institution and tackled the following questions: What forms
of slavery were there in the Ottoman Empire? For example, was there agricultural
or industrial slavery in the Ottoman lands? How similar were these patterns to
global slavery studies? Was Ottoman slavery similar to transatlantic slavery? Was
there a system of slavery that made the Ottoman Empire unique? What are the
primary sources that enable historians to explore Ottoman slavery? They intended
to differentiate Ottoman slavery from transatlantic slavery by calling Ottoman slavery
“servile” or observed similarities in a broader sense and named it “chattel.” In
one way or another, this scholarly work on slavery planted the seeds that allowedg
for Ottoman slavery literature, as it outlined the fundamental framework on the
subject matter. The topics focused on in the literature mentioned are primarily the
employment areas of slaves, the diversity in the ethnicities of slaves, the enslavement
9
sources, and the manumission methods and their significance in Ottoman society.
As shown above, Ottoman slavery literature was inspired by global studies from the
1980s; nevertheless, it did not emerge in the 1980s. It could be said that the works
of Bernard Lewis initiated and accelerated modern studies. Even before the 1980s
and Bernard Lewis’s contributions, there were studies that had dealt with forms
of Ottoman slavery. Amongst them, Ömer Lütfi Barkan’s pioneering scholarship is
worth mentioning.
In 1939, Ömer Lütfi Barkan published his groundbreaking article in Iktisat Fakültesi
Mecmuası (The Journal of Faculty of Economics) in which he unearthed sharecropping
slaves (ortakçı kul) who were were employed in big farms in the Ottoman
lands. According to the registry of 1498, there were share-cropping slaves employed
in the crown villages in Istanbul and different regions in Marmara, Western Anatolia,
and Rumelia. Barkan demonstrates that the share-croppers were employed in big
farms that belonged to the Crown (hassa), the askeris, private individuals, and vakıfs
(pious endowment). Discussing the legal status of the share-croppers, Barkan claims
that their status completely differed from that of the free peasant subjects in the
Empire (1939, 31), as they were the property of their owners. Additionally, principles
in Islamic law related to slavery were applied to them, and they were exempted from
marriage or changing their workplace, which was different from the legal status of free
peasants. In this sense, Barkan suggests that share-cropping slaves resembled serfs
in Western Europe (1939, 33). Building on this, Barkan examines laws these sharecropper
slaves had to obey based on the registry of 1498 (Barkan 1943). Diving
deeper, Barkan further questions whether there was a serfdom in the Ottoman
Empire in his other work worth mentioning entitled “Türkiye’de “Servaj" Var Mı
Idi?” (Was There Servage In Turkey?) and to what extent European serfs resembled
Ottoman share-croppers (Barkan 1956). As a result, Barkan argues that there
were slaves settled on the land in the fifteenth-century Ottoman Empire and that
independent farming enterprises were established with them. He further argues that
the rights and responsibilities of land workers were very similar to those of European
serfs, and that this continued until the middle of the sixteenth century and began
to disappear after.
Before publishing “Chattel Slavery in the Ottoman Empire” in Slavery and Abolition,
Alan Fisher had first penned an article entitled “The Sale of Slaves in the Ottoman
Empire: Markets and State Taxes on Slave Sales, Some Preliminary Considerations”
which was published by Bogaziçi University Journal in 1978. While Fisher was doing
archival research on the sixteenth- and seventeenth-century Crimean Khanate
at the Ottoman archives in Istanbul, he encountered numerous archival documents
10
on Ottoman slavery and got surprised how this subject had not been studied until
that day. In Fisher’s pioneering article, in which he shares his “preliminary considerations,”
Braudelian influences can be observed as the author underlines the
ubiquitous nature of Ottoman slavery in the Mediterranean (Fisher 1978).
In contemporary times, Halil Inalcık penned an article titled “Servile Labor in the
Ottoman Empire” in 1979. In this work, Inalcık provides an overview of the employment
of slave labor in different areas such as the military, economy, crafts, and
agriculture in the fifteenth and sixteenth centuries of the Ottoman Empire. Inalcık
demonstrates that domestic slaves were acquired by different social groups such
as courtiers and high-ranking and well-to-do non-Muslims. As understood from
the title, Inalcık favors using the term “servile” rather than “slavery” to specify
slaves in the Empire, on the ground that the enslaved people—especially the elite
slaves—were different from the slaves in the Americas (Inalcık 1979).
During the same time period, Halil Sahillioglu conducted pioneering research on
slavery and its economic aspects, focusing on the city of Bursa in the Ottoman
Empire. In light of the late fifteenth-century Bursa court registers, which Sahillioglu
investigates in “On Besinci Yüzyıl Sonunda Bursa’da Is ve Sanayi Hayatı: Kölelikten
Patronluga” (From Slavery to Employer: The Business and Industrial Life in Bursa
at the End of the Fifteenth Century), he draws attention to the excessive number
of slaves in Bursa, which was a commercial and industrial center at that time,
and unveils their role in the economy (Sahillioglu 1980). Using rich court records,
Sahillioglu notes that “slave” was not a lifelong status and discusses the methods
of manumission. His work reveals that numerous slaves were emancipated through
the mükatebe contract in Bursa; however, being forced to work in the weaving
industry was stipulated in return (Sahillioglu 1980, 182-184). The research uncovers
that some slaves who previously worked in the weaving industry opened their own
weaving workshops and benefited from slave labor in doing so (Sahillioglu 1980,
184). Finally, the author goes on to say that there were women, manumitted or still
slaves, who worked in Bursa’s weaving industry in the fifteenth century (Sahillioglu
1980, 187). Shortly after, Sahillioglu revised and published his above-mentioned
article in English as “Slaves in the Social and Economic Life of Bursa in the Late
15th and Early 16th Centuries” (Sahillioglu 1985, 43–112).
2.2.1 The Demise of Slavery and the Slave Trade
Towards the end of the eighteenth century, slavery and slave trafficking began to be
heavily criticized in Europe. Denmark took initiative and made the transatlantic
11
slave trade forbidden in 1803, far from abolishing global slavery itself. In England,
a robust anti-slavery movement arose, which was supported later by the government
in London and finalized with the approval of the Abolition of the Slave Trade Act of
1807 that officially prohibited the slave trade in the British Empire’s lands. France
continued to allow slaves to be trafficked within the bounds of its empire until 1815,
and Spain and Portugal until 1820. In 1815, the Congress of Vienna declared a joint
statement about the slave trade. It stated that the slave trade should be forbidden
since it “desolated Africa, degraded Europe, and afflicted humanity” (Drescher 2009,
230).
Nevertheless, slave transport remained in various forms and dimensions throughout
the world during the nineteenth century and the twentieth century. To further
its cause, the British Empire turned its perception of colonial emancipation into a
global abolition movement. The British ministers aspired to constrain the trade “in
every part of the world.” All foreign authorities were notified that abolition was the
plan “of Her Majesty’s Government and the British Nation” (Erdem 1996, 70-77.
Drescher 2009, 273). Thus, Britain embarked on what Eve M. Troutt-Powell has
called “invasive abolitionism” (Troutt-Powell 2003, 136).
Another region where the slave trade continued into the nineteenth century was the
Ottoman Empire. The British government desired to prohibit the Ottoman slave
trade, as it did in several other areas, or restrict it where it failed to end it. This
raises the following questions: What initiatives did the British government take to
prevent the Ottoman slave trade? How did the Ottoman government resist the
pressure from the British Empire? What was the nineteenth-century Ottoman slave
trafficking like? Which slaves were brought from where? How did the slave trade
end in the Ottoman Empire?
Addressing these questions, Ehud R. Toledano completed his groundbreaking
doctoral research under the supervision of Bernard Lewis in 1979 at Princeton
University, which was published later as The Ottoman Slave Trade and Its
Suppression:1840-1890 in 1982. His research examines the ways in which the British
attempted to suppress the Ottoman slave trade between 1840 and 1890. Toledano
provides a comprehensive picture of the slave trade and its routes and network with
the help of meticulously drawn maps, statistics listing the numbers of exported
slaves distinguished by gender and ethnicity, measures taken to suppress the slave
trade, and finally, the Sublime Porte’s reaction. Toledano thoroughly benefitted
from diplomatic records, correspondences, and negotiations between the two capital
cities: London and Istanbul. He established the fact that an imperial edict was imposed
to outlaw the slave trade of Africans in 1857. Still, the traffic was suppressed
12
after 1880, when the Ottoman Empire became a signatory to The Anglo-Ottoman
Convention Of 1880. The scholar dedicates Chapter V to “Circassian Slavery and
Slave Trade,” which was utilized enormously in this thesis. He addresses the obstacles
faced after the influx of Circassians into the Ottoman lands after the Crimean
War on the grounds that some Circassians migrated with their own unfree servants
from their homeland and utilized their slaves’ labor on government-granted lands.
Unlike the African slave trade, Toledano (1982, 148) claims that “reforms concerning
Circassian slavery and slave trade were entirely the result of Ottoman initiative.”
In this subject that continues to be explored, Toledano has been joined by Hakan
Erdem. Conducting his comprehensive research on the same topic at Oxford University,
Erdem’s meticulous dissertation, later printed as Slavery in the Ottoman
Empire and Its Demise 1800-1909, contributed to this growing literature. Organized
chronologically, the well-researched work is indeed a political and diplomatic
study of the institution of slavery in the Ottoman context; however, what makes
it so significant is that it fills a gap in this literature, which was a text that thoroughly
discusses Ottoman slavery. Even though Erdem sheds light on the means
of legal and illicit enslavement and manumission in the pre-Tanzimat period, the
bulk of his research is devoted to the nineteenth century. In regards to the question
of abolition and whether the Ottomans abolished slavery officially or there was a
debate for or against slavery in the Empire, Erdem underscores that “there was no
organized movement for abolition in the Empire, no abolitionist tracts popularizing
the subject and bringing home the sufferings of slaves—real or imagined” (Erdem
1996, 19). In his work, he intends to demonstrate the ways in which slavery met its
demise without an official decree of abolition. How did that happen? It has been
debated whether the British intention was to end slavery gradually or end it right
away. In regards to this, Toledano asserts that the British government had aimed
to abolish slavery when it first began to put diplomatic pressure on the Ottomans,
but that it ultimately had to change its intention to suppress the slave trade later
on. On the other hand, utilizing the same archival material, Erdem disputes this
assertion by indicating that the British government had never intended to propose
that the Ottomans abolish slavery downright (Erdem 1996, 74).
In Tunisia, which can be referred to as one of the regions located in the periphery of
the Ottoman Empire, the nineteenth century anti-abolition process took place in a
different and unique way. Ismael Montana penned a welcome addition to the growing
Ottoman slavery studies in 2013. In his work, he addresses the unique historical
event of the abolition of slavery in Tunisia and exposes the British influence on the
abolition of slavery in the Tunisian realm. There was no anti-abolitionist movement
in the region, but the ruler of Tunisia, Ahmed Bey, had abolished it officially in
13
1846 when the Ottoman central government abolished the slave market in Istanbul.
How did this happen? How did Ahmed Bey justify the abolition of slavery in
terms of Islam, which did not prohibit slavery? Where did this idea come from?
Addressing these points, the author narrates the story of this exceptional event and
illuminates the role of British anti-slavery influence in the official abolition of slavery
in Ottoman Tunisia. Montana has shown that the end of slavery is closely linked
to the nineteenth century’s social, economic, and political events by contextualizing
the decision of Ahmed Bey (Montana 2013).
In 2015, Ceyda Karamürsel completed her doctoral dissertation under the supervision
of Eve M. Troutt-Powell at the University of Pennsylvania. The author intended
to place the end of slavery in the Ottoman Empire in the context of global slavery
studies, considering different intersecting concepts and phenomena. In her work,
Karamürsel investigates “how such categories as gender, race, ethnicity, class, or
age mattered and how their meanings and experience changed over time for the
Ottoman state, slaveholders, as well as the slaves themselves” (Karamürsel 2015, 1).
Utilizing archival documents from different categories such as “slave petitions, slaveholding
elites’ correspondences, police interrogations, legal records, and parliament
minutes,” Karamürsel unveils the intertwined story of emancipation and citizenship
in the late Ottoman Empire and the early Turkish Republic. For instance, the author
unearthed the life of a Circassian slave girl named Fatma Leman, who fled from
her mistress’s house and sought refuge in the Ministry of Justice in Istanbul shortly
after the Young Turk Revolution of 1908. In her reasoning, Karamürsel notes, “If
this revolution, under the slogans of ‘freedom, equality, and justice,’ granted freedom
to each and every Ottoman individual, then she too was free like the rest of
her compatriots.” Neither the Ministry of Justice nor the Ministry of Police knew
what to do due to the uncertainties of freedom (Karamürsel 2015, 153).
2.2.2 A Gendered Perspective
The foundational studies that constituted the first wave of the Ottoman enslavement
studies concentrated mainly on the political and economic aspects of the institution.
Consequently, they did not take into consideration gender as a category for historical
analysis until the late twentieth century, when scholars of Ottoman history
attempted to fill this gap.
To begin with, Leslie P. Peirce positions harem slavery right in the heart of imperial
politics in her book entitled The Imperial Harem: Women and Sovereignty in the
Ottoman Empire in 1993 (Peirce 1993). In her text, she first lays out a parallel be-
14
tween establishing the concubinage system and the growing empire’s interests. She
goes on to argue that the Ottomans abandoned the marriage alliances that they had
been making after the conquest of Istanbul in 1453; instead, they embraced the concubinage
system. The concubinage system in the Imperial harem was kept alive by
bringing in slave girls of different origins. After establishing this parallel, the author
investigates the beginnings of royal women’s power in the sixteenth and seventeenth
centuries when it was perceived as corruption by contemporary intellectuals. This
study is fundamental for this present thesis as it demonstrates the ways in which
slave girls from various backgrounds were brought to the Imperial Household and
raised. Considering that the elites of Istanbul mimicked the palace customs, the concubinage
system presented by Peirce resembles the elites’ purchasing of Circassian
and African child slaves and teaching them in their own households.
Building upon Peirce’s work on the Imperial harem and concubine system in the
sixteenth and seventeenth centuries, Betül Ipsirli-Argıt completed her comprehensive
doctoral research in 2009 entitled “Manumitted Female Slaves of the Ottoman
Imperial Harem (Sarayîs) in Eighteenth-Century Istanbul.” She investigates what
followed when the female slaves in the Imperial harem manumitted and left the
Imperial Palace (çırag ) in the eighteenth century. Ipsirli unveils the interactions
and patronage relationship between former palace slaves and the Imperial Palace.
She claims that this relationship, which was beneficial for both parties, had indeed
started in the palace and continued even after their manumission without seeing a
reduction in the strength of the relationship. Ipsirli’s work shows that even after
leaving the palace, former slaves looked after the interests of the royal house and
contributed to the formation of a neighborhood that was loyal to the palace through
their marriages. The research reveals that these households brought the palace
culture to the populace and contributed to city life in Istanbul through various architectural
projects. This thesis benefitted from Ipsirli’s contribution to scholarship
on Ottoman slavery as it likewise shows the ways in which the relationship between
elite owners and child slaves had started in the household, and did not necessarily
end after their manumission. As Chapter 3 demonstrates, after manumission, child
slaves often kept their relationships with their elite owners and stayed connected to
their networks.
Madeline Zilfi joined the constantly growing cluster of scholars focusing on Ottoman
enslavement in 2010, introducing a gendered perspective into the Ottoman slavery
institution. Her gendered reading of the existing sources brought new insights to
the Ottoman enslavement literature: for instance, Zilfi questions what enslavement
means for those who endured slavery and enslaved people. Mainly focusing on the
eighteenth- and mid-nineteenth-century Istanbul, Women and Slavery in the Late
15
Ottoman Empire: The Design of Difference contextualizes Ottoman enslavement
socially, politically, and culturally. The author asserts that slavery was fundamental
and inseparable from all aspects of life. Zilfi draws attention to the unpleasant side
of enslavement as it was the reality of life. For instance, the author indicates that
skin color caused different consequences in the lives of slave people (Zilfi 2010, 136).
The author also finds the distinction between the slaves artificial, arguing that a
slave is a slave and that there is no distinction between sexual slaves and all the
other types. All were used by their owners as they pleased.
Her work additionally challenges the traditional interpretation of slavery in the Ottoman
Empire and pushes back against generalizations such as the mildness of the
domestic slavery that originated when the Ottoman elite attempted to frame the
purchase of young slaves as a way of kindness and charity in response to anti-slavery
pressures. For the Ottoman elite, it was a benevolent action to purchase young
slave girls who should be given an upbringing and then be married off to nobility.
The author’s most prominent argument is that slavery gained became gendered
throughout the period between the fifteenth and nineteenth centuries, culminating
in the institution becaming feminized by the nineteenth century. There are two
main reasons for this. First, a significant number of slaves that were imported and
sold in the urban hubs of the Empire were women. Second, the Empire’s military
organization had changed from a slave-based army to a conscription-based one due
to modernization efforts. Additionally, the sources for slaves in Africa were dried up
in the nineteenth century. That is why the slave population in Istanbul consisted
mainly of female slaves of Caucasian origin. Consequently, most slaves purchased
in Istanbul were women employed as domestic servants or young Caucasian-origin
slave girls acquired in the nineteenth century. Last but not least, in parallel to these
developments, owning a Circassian or a Georgian female slave became a status symbol.
This thesis fits well with the above-mentioned main arguments of Zilfi’s significant
work. Her study aimed to show once again that child slaves in the nineteenth century
were primarily girls due to the feminization of the slavery institution, as well as
that the acquisition of girl slaves was a practice belonging to the elite circles, thus
making having a child slave a status symbol. In congruence with Zilfi’s arguments,
this thesis examines the phenomenon of child slave girls being imported from the
Caucasus region, purchased and nurtured by some elites, and then married off to
other noble people in the nineteenth century.
16
2.2.3 Slavery in Legal Texts
As mentioned above, the scholars who wrote the foundational Ottoman slavery
studies located the existing primary sources related to the phenomenon of slavery
and its various forms. The legal texts, primarily the court registries, have been
significant in shedding light on Ottoman society and its slaves, primarily the social
and economic aspects. “The vast majority of cases in the court records related to
female slaves, and to slaves in general, are acts of manumission and sale” (Ze’evi
1995, 161. Özkoray 2019, 262). Nevertheless, it must be stated that the court records
usually “fail to satisfy the historian’s craving for detail, narrative expansiveness, and
voice” (Zilfi 2005, 135) since they provide limited representation of the slaves and
owners.
Enormous effort has been put in, and numerous researched works have been done
utilizing court records, causing the literature to grow significantly. Considering the
studies relying on court records, it can perhaps be noted that the following primary
topics and aspects are most often investigated: slave ownership, slave agency, emancipation,
fugitive slaves, the role of slaves in social and economic life, the market
value of slaves, and the employment areas of slaves.
When examining the court records, it is inevitable to focus on a specific region,
leading to the capital city of Istanbul attracting the most attention (Seng 1996, 157.
Ben-Naeh 2006. Sobers-Khan 2020. Aykan 2017. Özkoray 2019. Wagner 2020.).
There might be several reasons for that, but one of them is that “there are over
10,000 registers for Istanbul and its suburbs. . . a robust survival rate” (Zilfi 2010,
5).
In addition to the city of Istanbul, there is significant scholarship on other cities that
must be mentioned here such as Aydın (Faroqhi 1991), Konya (Sak 1989), Jerusalem
(Frenkel 2020), Bursa (Sahillioglu 1985. Canbakal 2020), Cyprus (Jennings 1987),
Crete (Spyropoulos 2015), as well as cities in the Crimean Khanate (Yasa 2018 and
2019) and Egypt (Baer 1967. Walz and Cuno 2010).
2.2.4 Individual Cases of Ottoman Enslavement
In addition to legal texts such as court records, this thesis intends to make use of
existing slave narratives. Here, an important question must be posed: Was there a
slave narrative in the Ottoman Empire? Can the voiceless express their stories in
their own voices? To what extent do the sources allow historians to tell the stories
17
of slaves? In the Middle Eastern context, Erdem notes that there is no such genre
of source material as there is in the “New World” since “there was no abolitionist
public to feed these narratives” (Erdem 2010, 125). Nevertheless, such narratives
in the Americas were “published under the careful editing and supervision of white
American abolitionists,” as Troutt-Powell notes. In terms of Ottoman slavery, what
is available is a bunch of valuable documents “in the form of letters, court records,
petitions, autobiographical notes, and even graffiti” (Erdem 2010, 125).
Instead of works written by the slaves themselves, there are written analyses of
individual cases of Ottoman slaves, which still allow the voices of the voiceless to
be heard. In light of these individual cases, some titbits of information about child
enslavement and what it was like can be grasped as long as the source provides
retrospective details. The following texts inspired this thesis with their arguments,
inquiries, and conclusions, all which intended to make the voices of slaves heard.
Analyzing a police report issued in Cairo in 1854, Toledano (1993, 59-74) reveals the
story of a Circassian girl named Semsigül, who was purchased in Istanbul by a slave
trader named Deli (Mad) Mehmet in 1852. He had sexual intercourse with her on the
ship, which led to the unfortunate fate of the slave girl. Deli Mehmet decided to get
rid of the pregnancy. The owner tried very hard to cause her a miscarriage: he gave
Semsigül pills to take and then beat her to cause her to lose the child, but nothing
worked. In the meantime, Semsigül gave birth, but the tragic story of Semsigül does
not end here. Within a year, the baby would die, and Semsigül continued to be sold
to others. Her next owner was told that Semsigül had a child with Deli Mehmet.
The incident was then reported to the police to deepen the investigation and prepare
the indictment against him. It is not known how representative Semsigül’s story is;
however, forced sexual intercourse, forced abortion, and exposure to violence were
harsh realities that slave girls were subjected to.
In 2006, Eve M. Troutt-Powell published her work “Will That Subaltern Ever
Speak?” in which she was inspired by Gayatri Chakravorty Spivak’s observations
of Hindu women who committed rite of sati, the Indian practice of widow suicide.
According to Spivak, the most deprived and marginalized in society have no voice to
express themselves, represent their interests or concerns, or ask for a more equitable
portion of society’s goods (Spivak 1988). In this regard, Toledano notes that Troutt-
Powell urges historians not to leave their jobs because they have not yet collectively
managed to uncover the slave voice (Toledano 2007, 7).
Taking the attention raised on the importance of narrating the individual stories
of Ottoman slaves into consideration, Erdem told the story of a black African girl
named Feraset, who committed the crime of setting fire to her owner’s home in Izmir
18
in 1867 and consequently was charged with arson. Just like Karamürsel revealed the
story of a Circassian slave named Fatma Leman who demanded her manumission,
Erdem shows that Feraset took action to decide her fate instead of waiting for her
manumission (Erdem 2010, 125-146).
In her other work, Troutt-Powell reveals the stories of individuals who endured several
forms of enslavement, as the author believes that recently written narratives
“cr[y] out for retelling” (Troutt-Powell 2012, 206). Her work, Tell This in My Memory:
Stories of Enslavement from Egypt, Sudan, and Ottoman Empire consists of
six detail-rich case studies that reveal some of the enslavement stories of individuals
in the Middle East during Ottoman and post-Ottoman periods. The author
attracts attention to the impact of bondage on slaves’ identities, world views, and
self-perception. As the author notes, “their own languages are lost, their identities
oversimplified” (Troutt-Powell 2012, 38). In light of the memoirs written by slaves
or slave owners, the historian presents the ways in which slaves integrated into Ottoman
elite and urban households and how slavery was part of their everyday life.
The individual case that the present thesis utilized the most is the story of Halide
and Huda. They were were two contemporary feminists whose views on domestic
life and social order affirmed the reality of slavery when its analyzed comparatively.
Halide was given an African child slave as a playmate by her father when she was
a child, and Halide and that child slave were raised together; even after the slave
was manumitted, they continued residing together. Troutt-Powell (2013, 115-147)
presents the ways in which that child slave and other slaves in the household played
a role in the individual and political development of Halide Edip Adıvar.
2.3 Global Child Slavery Studies
Child slavery has only recently attracted the attention of academic research. Previously,
scholars have focused primarily on adult male slaves, and more recently on
women in slavery (Miers, Miller, and Campbell 2009, 1). It can be noted that these
global trends were respectively followed by the Ottoman historians as well. As noted
above, the gendered perspective was applied to the broad framework on the institution
of slavery was established. This raises some questions: Why were scholars late
to focus on children’s experiences of slavery? Why did what enslaved children endured
only begin to be researched now in the twenty-first century? Scholars engaged
in child slavery studies had previously believed that children were did not make up
the majority of slaves (Miers, Miller, and Campbell 2009, 1). Nevertheless, more
19
recent studies demonstrate that the proportion and role of children in slavery could
no longer be ignored.
The historians Joseph Miller, Suzanne Miers, and Gwan Campbell edited a valuable
work on the subject in 2009. The rich collection of articles in Children in Slavery
through the Ages has made a significant contribution to the current understanding
of children in slavery. The book aims to provide comparative examples of child
enslavement from the eighth to the twentieth centuries. The editors Miller, Miers,
and Campbell devote their introduction to an essential discussion on the definition
of slavery and the definition of child. In light of the existing studies, it becomes clear
that “slaves were not always chattels or deprived of basic civil rights; they did not
always pass their slave status on to their children; and some could rise to positions
of considerable wealth and influence” (Miers, Miller, and Campbell 2009, 1).
In regards to the definition of a child, it is as hard to define as the term slave.
In today’s world, the United Nations defines a child as “anyone under the age of
eighteen,” but this is not a universal standard. When contemporary times are being
considered, several aspects and conditions should be taken into consideration. For
example, the age to get married for girls could be as low as twelve, while the age
to be conscripted for military service depends on the location. The age one finishes
school could be considered an indication of adulthood, but this cannot be considered
a viable definition since education is not mandatory in some countries. Of course,
it should be noted that numerous children in the world do not have registered birth
certificates and that therefore, their ages are unknown (Miers, Miller, and Campbell
2009, 1). The editors of the book underscore that “the difficulty of defining childhood
is even greater when discussing the past” (Miers, Miller, and Campbell 2009, 1).
Again, it depends on time and space, and one should consider multiple aspects.
For instance, the determining factors for a child’s age was height and apparent
physical maturity when trafficking slaves across the Indian Ocean (Miers, Miller,
and Campbell 2009, 1).
Regarding the content of the book, the essays focus on slave children and the trafficking
of them in a wide array of regions: the Americas, West Africa, East Africa,
North Africa, the Indian Subcontinent, the Ottoman lands, China, and Europe.
Many of the essays are noteworthy. For instance, Kenneth Morgan’s essay provides
insight on the unfortunate fate of enslaved children in British Caribbean society.
Pierre Boulle reveals “non-white child slaves and servants in the eighteenth century
France” and touches upon “strategies of resistance of African children and youth
enslaved in Europe.” Similarly, Kristina Richardson’s essay focuses on her research
on slave girl singers and entertainers employed in the Abbasid courts in Baghdad’s
20
ninth and tenth centuries (Rossi 2010). Finally, Gülay Yılmaz’s essay discusses the
children recruited using the devshirme method to be a part of the Janissary Army
or taken to serve in the Imperial Palace.
Recently, another edited book has been published on this subject. Child Slavery
Before and After Emancipation: An Argument for Child-Centered Slavery Studies,
which is composed of essays written by scholars, historians, activists, and philosophers,
invites scholars to question concepts that they oftentimes take for granted
such as “trafficking” or “slavery” (Duane 2017, 2-3). The author Anna Mae Duane
urges historians that “the present has the capacity to affect our perspective on the
past” (Duane 2017, 2). She goes on to say that “it is risky to accompany invoking
the word slavery in a modern context,” therefore it is significant to question, ‘When
is a child a slave?’” (Duane 2017, 2). Regarding the question of why scholars or
writers did not include children in their analyses, Duane (2017, 5) states that children’s
vulnerabilities would complicate their work more. It is noteworthy that the
author draws attention to the fact that the perception of child slaves as properties is
problematic. Duane asserts that children are already considered domains of others,
primarily of their parents in that children do not speak on behalf of themselves, but
rather their parents speak for them. In this sense, slaves are properties, too, but in
this case, properties of their owners (Duane 2017, 15).
2.4 Ottoman Child and Childhood Studies
Ottoman historians have recently begun to address several questions that had been
left answered in Ottoman childhood studies. What is it like to be a child in the
Ottoman Empire? Is the concept of a child a modern concept? Who did the
Ottomans consider to be children? Were children like small adults? One of the
first comprehensive works in the field of the history of children in the Ottoman
context was written by Yahya Araz. His work 16. Yüzyıldan 19. Yüzyıl Baslarına:
Osmanlı Toplumunda Çocuk Olmak, (Being a Child in the Ottoman Society: From
The Sixteenth to The Early Nineteenth Century) firstly focuses on the dramatic
tales of the lives of some children reflected in the archival records, emphasizing the
effect of the time spent with parents on the happiness and peace of children and
noting the troubles and miserable lives of children whose mothers or fathers died.
Secondly, since school comes to mind when mentioning children, the author discusses
the establishment of pre-Tanzimat educational institutions, their continuation, and
the sources of their revenue. With regards to the participation of children in school,
21
it is not clear at what age they started school and how many years they continued
there. According to Araz, when children go to school, they disconnect from the
world of adults, and they start to build their own world where they spend more
time with their peers; nevertheless, there would be a clear difference between boys
and girls based on gender. For example, the lucky ones among the boys attend
the madrasas, while it is impossible for girls who have completed primary school to
receive a higher education. Last but not least, Araz points out that it is possible to
think that the decline in employment of slave labor was caused by the attempt to
replace it with the labor of adopted children, which increased in the second half of
the eighteenth century.
A welcome contribution was made to the growing Ottoman child and childhood studies
by Nazan Maksudyan. Analyzing records from the Ottoman, German, French,
and Protestant and Catholic missionary archives and memoirs, Maksudyan investigates
the situations that girls and boys had to endure to reveal the voices of the
orphans and impoverished children of the late nineteenth-century Ottoman Empire.
The author examines the encounters of destitute children in shelters, households,
vocational schools, and foreign orphanages and contextualizes the emergence and
development of these institutions. First, Maksudyan reveals the practice of child
abandonment by poverty-stricken mothers in the nineteenth century as a harsh reality
of life. Afterwards, the author argues that the political rivalry between the
Ottoman central state and ethno-religious communities came about “not [over] orphans
but their future” (Maksudyan 2014, 125). Finally, the author points out that
the Ottoman central government took initiative to fix these domestic problems; however,
the modern regulations and organizations to advance the children’s conditions
concerning modernization of the Empire failed.
Although adopting daughters and fostering them was a widespread phenomenon in
all ethnic groups of the nineteenth-century Ottoman Empire, it did not always go
well. Maksudyan provides insight on domestic servant girls’ resistance to abuse by
their quasi-owners, even though their chance of success was considerably low due to
the status inequality between them. Some of the foster daughters (besleme) damaged
or burnt down their foster families’ apartments or they ran away or attempted
or committed suicide. In one way or another, they presented their opposition to
the abuse. Last but not least, unlike Yahya Araz, Maksudyan stands against the
argument made by other scholars that fostering daughters was a practice emerged
after the demise of slavery in the late nineteenth-century Ottoman Empire.
In addition to these contributions to the growing field of Ottoman child and childhood
literature, the works of Onur Bekir (Bekir 2005) and Benjamin Fortna (Fortna
22
2010. 2016.) should be mentioned. Also, very recently, an edited work from Edinburgh
University Press by Gülay Yılmaz and Fruma Zach will be added to this
growing literature soon. Children and Childhood in the Ottoman Empire from the
15th to the 20th Century focuses on a broad geographical area including Ottoman
Romania, Bulgaria, Crimea, Greece, Bosnia, Syria, Palestine, and Istanbul from the
fifteenth to the early twentieth century. Due to the diversity of the geographical area
it focuses on, the text provides insights on Christian, Jewish, and Muslim children
that explore the peculiarities and commonalities in family structures in different
communities within the Ottoman Empire.
Last but not least, it should be noted that the exploitation of child labor in domestic
work did not, unfortunately, end in today’s world. As Özbay notes, “[Humans] were
indifferent to the exploitation of children in the domestic sphere for a long time.
Only after the early 1990s Anti Slavery International became actively interested in
the situation of children in domestic work” (Özbay 1999). The author’s notable
historical and sociological analysis “Turkish Female Child Labor in Domestic Work:
Past and Present” shows that using child labor in domestic tasks is not new in Turkey
or elsewhere. As for the late Ottoman Empire, Özbay’s findings have been helpful,
demonstrating that there were slaves, waged servants, and evlatlıks of Caucasian,
African, and unknown origin in Istanbul’s Muslim households between 1885 and
1907. There were both female and male non-kin members of the households aged
broadly from six to fifty (Özbay 1999, 17-22).
23
3. CLASSICAL OTTOMAN SLAVERY SYSTEMS BEFORE
MODERNIZATION
3.1 Introduction
The captives caught during military action used to belong to whomever caught them
during the early years of the Ottoman state (Pakalın 1983, 766). However, this
changed in the following decades; accordingly, the Ottomans might have started
taking a one-fifth share of the war spoils, meaning that one of the five captives
captured during raiding would go to the state treasury as a tax. This in-kind tax was
to be called “pencik,” and the captives would be named “pencik boys.” The Ottoman
state benefited from the “pencik boys” whose ages figuratively ranged from seven to
seventy in different areas, primarily and possibly by employing them to carry out
administrative and military services. Before their employment, some "pencik boys"
were sent to Anatolia to work in agriculture and to learn the Turkish language and
Islamic traditions while living with some Anatolian farmer families. Some "pencik
boys" were trained in the Acemi Corps to be soldiers in the Janissary Army (the
Sultan’s private household troops). Additionally, the Ottoman state settled on an
official amount of cash that would be considered equivalent to one-fifth of a war
captive and stated that the tax could be paid in either captives or money. This
monetary tax was also named “pencik.” Furthermore, there might have been some
shortcomings to the pencik system. Over time, the system was revised, and several
aspects of it were improved. The idea was that the rules of the pencik tax collection
could be attempted to be established in a more structured and precise way in the
course of time. In this regard, to distinguish slave importers who paid their pencik
taxes from those who did not pay, a title deed was issued to the slave importer.
Eventually, the monetary pencik tax became an ad valorem tax to be collected on
each slave imported into the Empire over the coming decades (Erdem 1996, 19).
Recruitment through the pencik system was not the only method the Ottomans
24
applied to conscript soldiers to the troops or raise loyal statesmen for its militaryadministrative
positions. After the pencik system’s early implementations, the
devshirme practice was introduced as a new method. This raises some possible
simple questions. Why did the pencik system not work? What went wrong with it?
In addition to the pencik practice, the devshirme method was also implemented,
meaning that many children were levied for military service in certain regions of
the Ottoman Empire. Nevertheless, there was a clear distinction between the two
practices: the devshirme method was applied within the Empire’s borders, while
the pencik practice involved taking in war captives from enemy lands. With the
devshirme method, the Ottoman state started focusing on children rather than
adults this time. Some children of various ages were intended to be levied based on
a specific criteria and transferred to Istanbul. There, the levied children were once
again examined and converted to Islam. Afterwards, they were required to train
through a couple of different steps. Like the "pencik boys", some of those recruited
through the devshirme method were also sent to families in Anatolia to learn the
Turkish language and Islamic culture (Uzunçarsılı 1988, 23. Pakalın 447). Some
were even selected to be trained in the palaces of high-ranking statesmen (Mustafa
Âli 1997, 168). After completing their training, they were employed in military or
administrative fields and served the state.
While the devshirme practice was in effect, the way it was implemented did not
remain the same over the centuries but rather underwent a transformation. There
were several principles and rules involved and replaced in the selection, recruitment,
and employment of the devshirme boys. The practice of sending the levied children
to Anatolia was given up in the following centuries (Uzunçarsılı 1988, Vol I, 24). Additionally,
many questions might remain unanswered. Why were children preferred
over adults? What were the expected characteristics of the devshirme children that
were educated in the palaces of statesmen? What characteristics were they expected
to have to be eligible for state service?
Exploring these questions, this chapter closely examines the upbringing of the
devshirme children and the special education that was given to some of them in
the palaces of high-ranking statesmen. Although the practice of recruiting men to
the state to serve in the army or palace with the devshirme method was abandoned
at the turn of the eighteenth century, some high-ranking statesmen may have gotten
accustomed to raising servants for state posts and offices so much that they carried
on recruiting children from different sources and raising them in their palaces or
households. It is known that some nineteenth-century elites—statesmen, prosperous
merchants, and their wives—purchased slave children and raised them with various
25
motivations and purposes. To what extent could this be considered a continuum
when the devshirme practice is taken into account?
This dissertation focuses on the slave children brought up and trained by some of the
elites in the nineteenth-century Istanbul. It discusses explicitly why some Istanbulite
elites acquired slave children. To better comprehend the employment of child slaves
in elite households in the nineteenth century, I must first address the devshirme
practice, which I devote to the first chapter of this dissertation. It was a classical
Ottoman slavery system in which children were recruited, trained, and employed in
some high-ranking statesmen’s palaces before the nineteenth century. To explain the
devshirme system adequately, I must first begin with arguing the reasons why the
devshirme was necessary in the first place. In this sense, I will briefly elucidate the
pencik system, the predecessor of the devshirme, and what went wrong with it. After
explaining the pencik system, I will proceed with the devshirme’s compatibility with
the Ottoman sharia law. Together with this, I will concentrate on the ways in which
the devshirme boys were selected and what criteria were taken into consideration in
their selection and employment process.
3.2 Pencik System
Linguistically, the word pencik means one-fifth, which is indeed a distorted version
of two Persian words—“penc” meaning “five” and “yek” meaning “one.” In the Ottoman
context, the concept initially indicates the state’s practice of taking one of
the five captives caught during a raid from the raiders. In addition, the concept also
corresponds to a monetary tax which later became an ad valorem tax, and a title
deed, depending on the time period.
Some of the early Ottoman chronicles shed light on the introduction of the pencik
system. Even though they provide different details on the issue, some of
them—specifically those written by Âsıkpasazâde Nesrî and Oruc Beg—came to
the same conclusion that the idea of taking one-fifth from the spoils of war in accordance
with God’s will came into fruition after the conquest of Edirne in 1361 during
the reign of Murad I (Âsık Pasazâde 2013, 75. Nesrî 2014, 197. Oruc Beg 2014,
24-25). The Ottoman state implemented the pencik in-kind tax to be taken from the
war captives; these captives went on to be called the “pencik boys”.1 Initially, these
1It must be stated that the term “pencik boys” should not mislead anyone to think only of actual young
boys just because the word “boy” is included in the term. Since the ages of the captives ranged from 7 to
70, so did the recruited pencik boys.
26
pencik boys were employed in Gallipoli where a new military unit, the Acemi Corps
(Acemi Ocagı), was established consisting of the pencik boys (Uzunçarsılı 1988, 9).
Understandably, giving one of the captives to the state as a tax must not have been
very appealing for the raiders. It is natural for the people to not have accepted the
tax imposed immediately and different methods may have been developed to avoid
the tax. As Nesrî notes, upon introducing the pencik tax, some of the raiders started
hiding some of their captives to avoid paying the tax (Nesrî 2014, 199).2
Following the realization that taking one-fifth of the war captives would not always
be practical since not everybody would be able to capture five captives at a time, the
Ottoman state might have decided to take the monetary equivalence of the in-kind
pencik (Pakalın 1983, 766).
When the pencik system was first established, it was not a full-fledged system, but
had its deficiencies. Based on some chronicles, it seems that it was decided to
take the same amount, which was twenty-five coins, from all captives regardless of
their characteristics (Âsık Pasazâde 2013, 75. Nesrî 2014, 197. Oruc Beg 2014,
24-25). Nevertheless, not every captive was equal or equally valued; it may have
been confusing that the same amount of tax was taken regardless if the prisoner
was beautiful, healthy, and young versus one who was elderly and frail. In addition
to that, it could be unclear who would be responsible for recording the identities
of the boys and the payments of the taxes. Further, the state might have had
more concerns about implementing this taxation system and may have worried that
not all raiders would pay their taxes and that some would attempt to bring slaves
illegally. Eventually, the Ottoman state did not stay unresponsive to these problems,
and it attempted to improve the system’s imperfections by issuing regulations.3
For instance, the Ottoman state introduced a new position, a pencik officer named
“pencik emini” or “pencikci” who managed the taxes (Uzunçarsılı 1988, 7). These
officers were responsible for keeping records of the captives’ Christian names, their
new names, and their facial descriptions (Pakalın 1983, 767).
The pencik tax did not apply to all military conquests, meaning that the state was
not eligible to take one-fifth of the spoils depending on the purpose of the military
conquest. Which military actions the pencik tax applied to was clarified as well.
The state also issued a kind of tariff that divided the war captives into different
2“Amma bu akçayı ol vakit hums deyü alurlardı. Simdi humsı esir alundugu yirde alduktan sonra yine her
esirden geçit akçası deyü âlem-beri akça alurlar ve iskelede esir içün akça aldukların bilicek iskeleye esir
getirmeyüb her biri bir tarafda gizler oldılar.” The emphasis on “âlem-beri” is on the original text. (Nesrî
2014, 199)
3There are two pencik law codes and a decree that have survived until present times. The first law code
dates back to 1493, and the second one was penned in 1510. These law codes were studied by Ismail
Uzunçarsılı and Ahmet Akgündüz (Uzunçarsılı 1988, 86-88. Akgündüz 1980, 131).
27
categories based on their gender, age, and physical appearance (Uzunçarsılı 1988,
90. Akgündüz 1990, 131-132).
The state carried on issuing regulations designed to improve the quality of the boys
they recruited. As stated above, the pencik boys were initially employed in the
Gallipoli Acemi Corps. However, it was later considered a poor choice for the pencik
boys to work in the Gallipoli Acemi Corps, so instead they were sent to Anatolia
and to serve the farmers for a small price. Thus, they would learn the Turkish
language and about Islam in Anatolia. The pencik boys brought up like this were
thought to serve more faithfully. Also, they would not be able to escape easily to
Europe compared to if they were in Gallipoli (Uzunçarsılı 1988, 11). According to
Uzunçarsılı (1988, 11), the reason why the state paid them a small amount while
working in agriculture in Anatolia was that the pencik boys should not have caused
problems to the farmers and they should have been compatible. After growing up in
Anatolia, they worked in various services. Some returned to Gallipoli and worked in
the ship service there for a daily payment of one akçe. Some were employed in the
Acemi Corps while some were assigned directly to the Janissary Corps (Uzunçarsılı
1988, 12).
The state did not only confine itself to the collected pencik boys as an in-kind tax
but was also interested in the other captives taken by the raiders. As Pakalın notes
(1983, 767), the state purchased healthy and perfect boys between the ages of ten
and seventeen for three hundred akçe. These boys were also called “pencik boys”
and were able to be employed in the palace service or the Janissary Corps.
As stated above, the state issued an in-kind tax named pencik and then introduced
its monetary form. The monetary pencik later became an ad valorem tax that was
collected on all slaves imported to the Empire (Erdem 1996, 19). Upon paying the
pencik tax, a title deed was provided to the slave owner. In such a way, the slave
owner was given legal ownership rights over his slave. For instance, in 1688, three
coins miri and forty coins were taken as a tax on the prisoners taken during the
Üngürüs (Hungarian) campaign, and they were given pencik deeds.4 In another
example, a slave importer named Rüstem Aga from Kars paid the pencik tax for
his fifteen-year-old Georgian female slave and received his pencik deed upon the
payment in 1786.5 The pencik deed had a significant function since it served to
differentiate the slaves whose taxes were paid for from those that were left unpaid.
Slaves without a proper title deed could not be sold or traded. Issuing such a
4“Sebeb-i tahrir oldur ki Üngürüs seferinde alınan üseradan üç kurus miri ve kırk akçe alınıp tezkire verildi.”
BOA, IE.AS.. 26-2315. (22 Recep 1094) (17 JULY 1683)
5“Yalnız bir Gürcü cariyenindir. Merkumun tahminen 15 yasında bir Gürcü cariyesinden pencik resmi
alınmıstır.” BOA. TS.MA.e 918-71. (30 Recep 1200) (29 MAY 1786)
28
deed enabled the Ottoman central government to detect when slaves were illegally
enslaved or imported (Erdem 1996, 20).6
Paying the pencik tax at customs may not have operated smoothly all the time;
the slave merchants or pencik officers could have encountered multiple difficulties
and problems. For instance, a slave merchant named Mustafa paid his pencik taxes
for the twelve black female slaves that he brought from Tunis in 1722 at the Chios
Island customs. Upon his payment of the pencik tax, he received the pencik deed.
Later on, he went to Edirne and was asked to pay the pencik tax again at customs
even though he had paid the tax previously. The merchant Mustafa had to pay
again, and the pencik deeds were given to him once again. Stating that he was
victimized by having to pay the deeds twice despite having valid documentation,
Mustafa wrote a petition and sought his right.7
On some occasions, slave importers might have been asked to pay the pencik tax
even though the slave traders had documentation of their pencik deeds. For instance,
a slave merchant named Ömer had paid the pencik and other customs taxes for his
female slaves that he brought to Istanbul to sell in 1847. Even though he had the
deeds proving that he had paid the necessary taxes, including the pencik tax for
each slave, the merchant Ömer was asked to pay those taxes again when he arrived
in Bursa. He was forced by Cemal’s son Ibrahim in Bursa and did not have any
choice except to pay. Since Ömer knew that Cemal’s son Ibrahim did not have the
right to force him, he complained and demanded his money be paid back.8
Finally, the ad valorem pencik tax even became included in the tax-farming system
(iltizam). In the tax-farming system, the right to collect the pencik tax was given
to those who had committed to pay a certain amount of cash annually to the state
treasury and collect the pencik tax on their own. This continued until the prohibition
on trading black slaves in 1857. After, the pencik tax was removed from the tariff
(Toledano 1994, 58). Hence, the debts of tax-farmers (mültezims) were pardoned.
The Central Treasury announced they would stop taking the pencik tax in the
customs in Istanbul. An Ottoman “pencikci” was assigned to check whether everyone
abided by this new prohibition (Toledano 1994, 58). Toledano suggests that the
pencik tax might have continued to be taken for a while, though the ban might have
6Finally, in the Ottoman archives, one might encounter multiple terminologies in searching of pencik title
deeds. Scribes used different terminology in referring to these title deeds: “pençik kagıdı,” “pençik varakası,”
“pençik senedi,” “pençik temessügü,” “esir kagıdı,” “esir senedi,” “pençik tezkiresi,” or “pençik makbuzu.”
These deeds are not new for Ottomanists, as some historians have already paid attention to these proper
titles briefly (Toledano 1994, 58. Erdem 1996, 19. Witzenrath 2016, 209. Kazıcı 1977, 132).
7BOA/Cevdet/Adliye/9-587 (25 Receb 1134) (1 MAY 1722) Cited also in Erdem, 1996, 19.
8BOA, A.MKT.DV.. 5 – 96. (06 Zilkade 1263) (21 SEPTEMBER 1842)
29
been better implemented in Istanbul and Izmir. However, the prohibition on the
pencik tax regarding the imperial decree of 1857 might not have been implemented
as anticipated. Toledano (1994, 59) noted based on local and foreign observers’
reports that the pencik tax almost turned into a bribe or a hush money.
3.3 Devshirme System
The Turkish term “devsirme,” meaning “collection,” refers to the periodical forced
levy of children amongst some subjects living in some of the Ottoman provinces
in Europe and Asia Minor (Ágoston and Masters 2009, 183). In a broader sense,
the devshirme was “the forcible removal of the children of the Christian subjects
from their ethnic, religious, and cultural environment and their transplantation into
Turkish-Islamic environment with the aim of employing them in the service of the
Palace, the army, and the state, whereby they were to serve the Sultan as slaves or
freemen and to form a part of the ruling class of the State” (Ménage 1966, 64).
Unlike their Muslim predecessors, who had developed their own system and benefitted
from imported slaves, the Ottomans supplied their slaves from within the
Empire’s borders (Erdem 1996, 1). However, it must be noted that devshirme was
not just a practice applied to the Christian subjects of the Empire; instead, it was
a system that was also applied to some Muslim subjects of the Empire located in
Rumelia—mainly Bosnians and Albanians, even after they converted to Islam. With
regard to other non-Muslim subjects of the Empire, Hakan Erdem (1996, 4) suggests
that Jews were not exempt from being a part of the devshirme even though it was
assumed that they were.
Before the devshirme method emerged, the state was using the pencik method to
gather the boys they needed to serve the state (Uzunçarsılı 1988, 13. Pakalın 1983,
Vol. I, 444). Why was a brand new system established without a rhyme or reason?
A new system could have been necessary for many reasons. Perhaps, the reason was
to fix the old system’s defects and weaknesses. If so, what was wrong with the pencik
system? This chapter discusses the reasons why the devshirme was necessitated.
The devshirme method has been studied from different perspectives previously.9
Scholars have primarily investigated the system’s origins, its introduction, and argued
whether the devshirme was compatible with Ottoman Sharia law. On one
9For some of the studies, see Palmer, 1953, 448-81; Wittek, 1955, 271-278; Ménage, 1956, 181-3; Vryonis
1956, 433-443. Ménage 1966, 64-78; Repp 1967, 137-9; Cahen 1970, 211-218; Ménage 1991, 210-213;
Demetriades 1993, 23–31; Öz and Özel, 2000, 555-557; Erdem 1996, 2; Ménage 2007.
30
hand, devshirme was an administrative-military system that the Ottoman utilized
for centuries. On the other hand, it was a form of slavery in which children were
recruited and forcefully employed in addition to adults.10
The devshirme system incorporated three main stages: selection, education, and
employment. There were some specific criteria that were utilized while selecting
the children: the boys’ social status, family background, and physical features were
considered before their selection. Upon their levy and arrival to Istanbul, some
notable people chose devshirme boys from a large selection of options in the sixteenth
century. They, too, paid attention to some specific criteria in choosing devshirme
boys for themselves, seeking the highest quality. Considering that the elites and
notables of the Ottoman Empire carried on acquiring slaves, including children, the
criteria that they paid attention to might not have changed over time, and they
might even bear a resemblance to another.
Considering the principle that different periods must be contextualized separately,
this chapter intends to question to what extent child slave selection criteria of some
nineteenth-century elites resemble devshirme boy selection criteria of some notables
of the earlier centuries. Could there be a continuum in the typical characteristics of
child slaves across centuries?
To begin with, the emergence of the devshirme system has been a subject of interest
to some historians. Victor Louis Ménage notes that the date of the emergence of
the institution of the devshirme is uncertain, nothing that there are some references
to the devshirme practice in which the earliest dates back to the 1390s (Ménage
“devshirme”, EI2, BRILL, 2007). While Speros Vryonis discussing the introduction
of the system, he writes that the metropolite of Thessalonica, Isidore Globas
recorded the devshirme as “seizure of the children by the decree of the amir” (Vryonis
1956, 433-43). An Italian source from 1397 stated that the Turks took boys
between the ages of ten and twelve for their army. In light of these references, Gabor
Ágoston suggests that the devshirme practice has already been established during
the late 1390s. (Ágoston and Masters 2009, 183). In this sense, Ménage suggests
that devshirme was established at the time when Isidore Globas referred to it, but it
might have been reintroduced during the time of Murad II, perhaps after it was suspended
in the years of confusion and conflict during the Battle of Ankara (Ménage
“devshirme”, EI2, BRILL, 2007).
As noted before, the soldiers were supplied to the Acemi Corps using the pencik
method at first. So what went wrong with the pencik system? Why was a new
10devshirme boys were examined in a child slavery context previously, see Yılmaz 2015, 901-930.
31
system required? According to Uzunçarsılı, due to some extreme political events in
the fifteenth century (e.g. the Battle of Ankara and the Interregnum Period), the
number of soldiers in the Empire decreased. As a result of these disturbances, some
lands fell to the Byzantine Empire and the Serbian Kingdom. Therefore, prisoners
of war could not be utilized during the reigns of Mehmet I and Murad II as they
were used in the past; a new method of recruiting soldiers was developed instead
(Uzunçarsılı 1988, 13. Pakalın 1983, 444-445).
A new system might also have been needed to avoid the difficulties and uncertainties
generated by the pencik system. Taking one-fifth of the total war captives did not
mean that only young, beautiful, and healthy men could be chosen. As mentioned
earlier, the term "pencik boys" contained a vast range of captives, from the age of
seven to seventy, as it included all prisoners of war regardless. Uzunçarsılı pointed
out that the perfect and healthiest captives between the ages of ten and seventeen
were purchased by the state with 300 akçe (Uzunçarsılı 1988, 8). They were different
from the "pencik boys" that were taken in accordance with the one-fifth tradition, as
they were able to select the captives more carefully. The state might have recruited
them to be employed in the palace service or the Janissary Army. However, this
would not be a well-established and full-fledged system and could be economically
exhausting in the eyes of the state. In addition to this age matter, the state might
not have known which family the captives come from. Moreover, the state could
not have been sure what the captives’ social status and background were; since they
were to serve the state, these aspects might have carried a lot of importance.
Having explained why the devshirme practice was needed, I now move to summarize
the debate on the devshirme’s compliance with Ottoman Sharia law. Before, the
concepts of the zimmî status and the zimmet agreement must be defined. The
monotheist non-Muslims living in the Ottoman Empire were called zimmîs. As it is
known, Christians and Jews living in the Ottoman Empire were monotheists, and
they were referred to as people of the book (ehl-i kitab). However, being considered
one of the “people of the book” did not automatically make a monotheist non-
Muslim subject of the Empire a zimmî. The zimmîs were “people of the book” who
accepted the authority and superiority of the Islamic state and made a contract with
the Islamic state. The zimmîs were granted the right to worship and were exempt
from slavery in return for their obedience to the Ottoman state and their acceptance
of their secondary status. They were obliged to pay a specific tax called the poll tax
(cizye). Paying this tax indicated that their zimmî status was valid, meaning they
could not be enslaved (Erdem 1996, 25). Nevertheless, in the case that Ottoman
subjects who had zimmî status were enslaved either by another Ottoman subject or
by a foreign person, in theory the state was obliged to do everything to end their
32
bondage (Erdem 1996, 23). In fact, zimmîs were occasionally illegally enslaved both
by other Muslim and non-Muslims groups, primarily but not exclusively corsairs
(Erdem 1996, 24).
Several scholars including Patricia Crone, Benjamin Braude, Bernard Lewis, and
Metin Kunt studied the devshirme practice from different angles and came to similar
conclusions. Patricia Crone claims that the devshirme system’s implementation was
clearly a violation of Sharia law (Crone 1980, 80. Erdem 1996, 2). Braude and Lewis
state that the devshirme practice clearly contradicted the essence of the zimmet
agreement (Braude and Lewis 1982, 1-34. Erdem 1996, 2) while Kunt asserts that
devshirme was a violation of the traditional rights which were provided in Muslim
societies (Kunt 1982, 55-67. Erdem 1996, 2). Ehud R. Toledano shares a similar
opinion with the historians mentioned previously, and states that “the devshirme
system, through which were the Janissaries recruited, was an aberration” (Toledano
1982, 6).
Erdem agrees on that the practice of devshirme violates zimmi rights, the rights
of non-Muslims living on Muslim lands, but asserts that the practice of devshirme
did not initially contradict Ottoman Sharia law since non-Muslim subjects of the
Empire did not automatically obtain zimmî status. While Christians and Jews
residing within the Empire’s borders were deemed eligible to have a status called
zimmî, it was not guaranteed to them. The Ottoman Empire agreed that those
who held the zimmî status were granted the right to worship, own property, and
be protected from slavery in return for respecting the Ottoman administration and
accepting their secondary status in the society (Erdem 1996, 2). If zimmî subjects
of the Empire did not abide by the rules of their agreement with the Empire, and
collaborated with the Empire’s enemies against Islam, their status would change
from zimmî to harbi. Thus, those who used to hold the zimmî status but became a
harbi instead were punished with death penalty or slavery (Erdem 1996, 2).11
The date of the initial introduction of devshirme is not known. The practice did not
contradict with the Ottoman Sharia law at first, but later did so. The ways in which
the system operated will now briefly be discussed. Under the Ottoman devshirme
system, many boys between the ages of eight and twenty were selected from villages
11For example, such a case occurred during the Chios Island massacre in the nineteenth century. Greeks were
one of the zimmî status holders of the Ottoman non-Muslim subjects at the beginning of the nineteenth
century. In 1821, some Greek subjects of the Empire revolted. Upon that, the Ottoman central government
approved the capture of rebel non-Muslims as prisoners of war on the grounds that the rebellion was
regarded as an action that broke the zimmî agreement between the non-Muslim subjects and the Empire,
thereby destroying the status of zimmî, which the Empire had granted them centuries ago. See, Erdem,
Yusuf Hakan. “Do not think of the Greeks as agricultural labourers: Ottoman responses to the Greek war
of independence.” In: Birtek, Faruk and Dragonas, Thalia, (eds.) Citizenship and The Nation-State in
Greece and Turkey. Social and Historical Studies on Greece and Turkey Series. (Routledge, London 2005)
pp. 67-84.
33
in Rumelia and Anatolia and sent to Istanbul (Somel 2003, 57). The rates of the
intervals and the rules of the levy were altered depending on the needs of the army
and the state. At first, a provincial governor (beylerbeyi), a military governor of
the province (sancak beyi), or Islamic judges (kadı) were in charge of selecting and
collecting the children living in the Balkan region. Not appointing an official from
the periphery might have rendered the possibilities of bribery and illegality such as
choosing children with a fake sultanic decree or taking bribes instead of levying a
child (Uzunçarsılı 1988, 15). Nevertheless, the government took measures to prevent
such issues. For instance, an officer from the Janissary Corps was appointed to
take care of the child levy. This officer had both a sultanic decree and a letter
written by the Janissary Agha in his hands. In the documents, how many children
should be levied, from what regions they should be collected and sent from, and the
characteristics of the children were clarified. There were other precautions taken
for the other processes of the system. For instance, there was a register prepared
in which the records of children were kept; the former names, villages, parents’
names, birth dates, physical features, and the name of the devshirme officer were all
recorded. One copy of that registry was transferred to Istanbul with the children
(Uzunçarsılı 1988, 16). On the way to the capital, the levied children were clothed
in a red attire (kızıl aba) and a hat (külah) in order to distinguish them and thus
prevent any escapes or kidnapping during the transfer (Uzunçarsılı 1988, 17). After
their arrival, the children were permitted to rest for two to three days. Shortly
after, they were examined for physical marks and injuries. All of them were then
converted to Islam, circumcised, and given Muslim names (Yılmaz 2009, 122).
According to Mustafa ‘Âli, they were also shown to a palace scholar who knew
physiognomy (ilm-i kıyâfe), the act of judging a person’s character from their face.
The scholar examined the children’s faces and chose those who would be the best
fit in the palace in regard to their usefulness and potential godliness. He would
also decide whom to send to the Acemi Corps to potentially be recruited to the
Janissary Army in the future (Mustafa ‘Âli 1997, 273).12 Those that were chosen
for palace service were sent to Iskender Celebi, Galatasaray, Edirne, or Ibrahim
Pasha’s palaces. There, they were taught Turkish, physical military training, and
basic Islamic knowledge such as the Qur’an, Muslim jurisprudence, theology, and
law (Yılmaz 2009, 123). On the other hand, the boys chosen for the Acemi Corps
were sent to families in Asia Minor to learn Turkish, Islamic tradition, and Turkish
customs. They would come back to Istanbul after four or five years and be employed
12“ilm-i kıyâfeden âgâh, saray-hâcesi nâmına bir ‘ârif me’ârif-güvâh devsürilüp gelen ve pîs-kes tarîı ile
gönderilüp gılmân-ı pencik olan o˙glânlaruñ sîmâlarına bakardı. Eskâlinde itimâl-i alâ u diyânet olanları
içerü aldırurdı. Sâirîni kapucılıga ve Yeñiçerilige çıkan ‘acemiyâna ilâk itdürüp nefer mertebesinde kalurdı.”
(Mustafa Âli 1997, 273)
34
based upon their skills and needs (Koçi Bey 2007, 39). It should be remembered
that the "pencik boys" were also sent to Turkish families to learn Turkish language
and custom as discussed above. One question remains: why was it necessary to
send the devshirme boys to the families for a couple of years? What could be the
reason behind that? First, it could be easier to educate and train someone who
knows the same culture and speaks the same language. Second, their integration
into society would be easier and simpler. By the seventeenth century, this practice
was abandoned eventually and the devshirme boys were not sent to the families. As
Uzunçarsılı notes, this might be because of the fact that Rumelia, the region where
the devshirme boys were levied from, was already Turkified by the seventeenth
century (Uzunçarsılı 1988, 24).
Of course, this might be the case; however, the devshirme boys were not only levied
from Rumelia, but also Anatolia. Thus, it might have been related to learning
and speaking Turkish. At this point, the perspective of a sixteenth-century elite
bureaucrat Mustafa ‘Âli on the ways in which some of the devshirme servants spoke
Turkish should be taken into consideration. ‘Âli’s opinion, which is based on an
example, might reflect the elite perception of the subject matter in the sixteenth
century. In Mevâ’ıdü’n-Nefâis, Mustafa ‘Âli makes fun of the servants whom he
calls ‘Kazaks,’ and who were settled in the Islamic lands when they were young,
and then placed in some notable households where they learned how to read and
write. The author draws a parallel between the way they spoke Turkish and the
clown of the world or the fool of humanity.13 From Mustafa ‘Âli’s opinion, it can
be interpreted that even if one learned Turkish at a very early age and carried on
learning in the household of a notable elite, the person would still have an accent
that became an object of derision in the eyes of some elite. This might be because if
Turkish was learned first through families in Anatolia, it could be hard to have an
Istanbulite accent. We still do not know what letters that they could not pronouns
well if that was the case.
Having explained how the devshirme system worked briefly, I will now take a step
back and focus on some of the criteria utilized when choosing the devshirme boys in
the villages. Although there were exceptions, the following criteria might show what
kind of children the Ottoman state may have strived to recruit and raise. Overall,
the physical appearance, social status, and individual characteristics of the boys
were considered in the selection process of the devshirme practice, which showed
who was eligible to serve the state or was worth training. First, it should be noted
that not all households were exposed to the child levy. Theoretically, if there was
13Mustafa Âli, Meva’ıdü’n-Nefais (ed.) Mehmet Seker, Ankara: TTK. “Bir dahı muzık-i alem ve mashara’i
beni-adem.....” p. 378-379.
35
only one son living in a household, he would not be taken, but would be left to serve
with his father. If there was more than one boy in the house, the most beautiful
and healthy boy would be chosen (Uzunçarsılı 1988, 17.). Second, it seems that
the children’s familial backgrounds played a role as well. The children that came
from noble families were favored and attention was paid not to take children of
shepherds or orphaned children. They were not preferred, as they were believed
to be vulgar and greedy. Instead, the officials chose children of the priests, who
were thought to have been better educated (Uzunçarsılı 1988, 17-18). In addition to
these, children who were married, circumcised, or seen in Istanbul before were not
preferred either (Uzunçarsılı 1988, 18). The list of criteria goes on; however, one
should be reminded that these rules might not have always been abided by and that
there could be exceptions.
As noted above, upon their arrival to the city, some devshirme boys were chosen
by some statesmen. Mustafa ‘Âli wrote his suggestions on the ways in which these
servants should be selected, trained in the palaces of some statesmen. He begins
with why it was needed to recruit servants in the first place and then moves on what
characteristics and personality types the servants should have had.
According to Mustafa ‘Âli, some statesmen took care of several administrative tasks,
and their households were the places where people came and applied for their needs.
That is why they needed several servants to help and work under them. As mentioned
above, some of them recruited devshirme boys upon their arrival to Istanbul.
At this stage, Mustafa ‘Âli advised the statesmen to consider some criteria in picking
the devshirme boys. First, the author suggested implementing physiognomy (ilmi
kiyâfe) to choose the best candidates better. Afterwards, the statesmen should
select the one(s) the physiognomist pointed out. When there was no specialist to
perform physiognomy, those selecting the boys were advised to check their clothing
and eliminate those whose clothes were torn, ripped, or dirty. Second, the person
making the selection might have considered serving food for these children to observe
the ways in which they have their meals. According to the author, the eligible
servants should be the ones who would eat fast and finish earlier without lingering
(Mustafa Âli 1997, 96).
Mustafa ‘Âli mentions the degenerated servants, explaining why these kinds of servants
existed and the consequences of having them. This part of his account is
crucial because from what Âli recorded, a hint of what qualities servants should
have had can also be grasped. As stated above, after the devshirme boys arrived in
Istanbul, the ones selected for the palace service were sent to the palaces of statesmen
to be raised and trained. Mustafa ‘Âli believes that if the servants were not
36
raised in one of the palaces of the statesmen, then there were lots of problems with
the way the servants started working in the palace service. ‘Âli complains that some
of the servants were lazy and avoided working, calculating their working hours. On
the contrary, decent and well-trained servants did not tell their masters that they
became tired, even if they worked hard and were exhausted. For ‘Âli, a well-trained
servant perceives this service as a compliment from his master (Mustafa Ali 1997,
169). The author also thinks that lazy and inattentive servants are useless, and
their absence is preferred to their presence because they do more harm than good
(Mustafa Ali 1997, 170). For instance, if a servant responsible for shopping became
lazy, did not go to the bazaar, and instead made excuses and said that the shops
were closed today or that material was not available in the bazaar, ‘Âli believes that
this is indecent behavior and that it deserves to be punished. Next, the servants
also needed to be intuitive: they should have blamed themselves for the cold behavior
of their masters towards themselves. They needed to apologize right away
if they made a mistake (Mustafa Ali 1997, 170-171). In addition, the author does
not welcome the act of escaping from the service of statesmen and describes it as a
despicable act (Mustafa Ali 1997, 171). For him, well-trained servants usually did
not run away from their service. Despite this, they may have sometimes attempted
to run away after falling into temptation. The servant’s apology was advised not
to be accepted when the servant was caught and brought back because this action
indicated that that servant was inclined to escape and would try again (Mustafa
Ali 1997, 171-172). In addition, servants were not to constantly complain about
their material needs such as clothing, and should have only reported their needs
to their owners when necessary. The servants should not have been hurt by petty
matters either. Another point that Mustafa ‘Âli complains about is talking behind
the owner’s back. ‘Âli criticizes the servants who did this behaviour to their masters
very harshly. As he notes, the servant should not vilify his master to others, should
not be ungrateful, and should not compare himself with other servants. Even if
their situation was very hard, they should not have complained about it to others,
the unfaming of the owner is comparable to murder or betrayal (Mustafa Ali 1997,
174-75). The author also writes about his thoughts on frugal and wasteful servants.
The servants with good manners paid attention to their clothes, saved their annual
stipend, and spent money accordingly. In this way, they were able to have many
outfits and still ensured that their owners did not suffer financially (Mustafa Ali
1997, 177).
Mustafa ‘Âli also compares his time with the previous decades in some parts of his
account. In this sense, he makes another comparison to explain why there were
degenerated servants during his time. He notes that the trained servants did not act
37
in an orderly and disciplined way as they did in the past, disregarding the palace
rules. The servants brought from far lands at very early ages did not have any contact
with the outside world previously. The servants started having contact and visits
with people outside of the palace during the reign of Suleiman I. The author believes
that this negatively impacted their courtesy, propriety, and manners; therefore, he
urged that this act must have been prevented (Mustafa ‘Âli 1997, 276).
Overall, the Ottomans implemented the devshirme practice for a long time; they recruited
children when they thought that they were useful to them in line with their
needs and the conditions of the time. The first time the devshirme method was implemented
is uncertain; the practice took place during the fifteenth century haphazardly,
and became implemented more regularly in the sixteenth century. However,
conditions altered over time and the Janissaries ceased to be the elite forces of the
sultan. This resulted in the decrease of the significance of the devshirme practice
(Ágoston and Masters 2009, 184). The practice was ended eventually; Baki Tezcan
estimates the last practice of the devshirme was around 1703, whereas Ménage
estimates it to be 1705 (Tezcan 2010, 44-45)
During the era of the devshirme method until the turn of the eighteenth century,
there had been adjustments and alterations to the practice depending on the conditions
of the time. It can be seen that while choosing the devshirme boys, some
officials might have paid attention to specific characteristics, and it can be assumed
that they may have been implemented to some extent, but it is hard to say how
well the criteria were implemented. Discussing this would be exceeding the scope
of this chapter. Nevertheless, taking a step back and analyzing some criteria are
worthy. As stated above, not knowing Turkish and not having seen Istanbul before
could have been essential for being a state servant. First of all, someone who had
seen and known Istanbul would have been awakened in a sense. The servant would
know not only the good sides of the city of Istanbul, but also the naughty sides. He
would also be aware of his capabilities and the opportunities that the city presented.
Moreover, he would have had connections and acquaintanceships with people living
outside of the palace or ruling groups. As observed in Gelibolulu Mustafa ‘Âli’s
account, this would be a behavior that the notables and the statesmen would not
wish to see. Second, if a servant selected was able to speak even a little Turkish,
he may have used foul language, publicly explained what he endured, cried, and
complained. This would have been unpleasant for the high-ranking statesmen who
owned them and could have negatively affected the servant’s education. Not speaking
the same language and not being able to communicate well would prevent all of
this and make the recruited child’s education more manageable.
38
It is known that these characteristics—not knowing Turkish and not having seen
Istanbul—were also sought in the children acquired in different ways to be employed
in administrative services in the following decades. In the seventeenth century,
Evliya Çelebi included a brief anecdote in his enormous travelogue on this subject.
While writing on the Tophane region of Istanbul, Evliya Çelebi mentions a group of
people who would take their one- or two-year-old children in Istanbul to Caucasia
by ships, not to be urban dwellers, but to be servants of the Empire. They were
given to milk mothers and brought back to Istanbul when they were between the
ages of ten and fifteen years old. Later, they were sold for the sultan’s service by
whom Evliya calls “Tophane Abazaları” (Abkhazians of Tophane). Interestingly,
those who engaged in this work were two of the grand viziers of the seventeenth
century: Melek Ahmed Pasha and Siyavus Pasha, who were servants of the Sultan
of Abkhazian origin.14
In the following chapter, it will be seen that some high-ranking state officials and
elites of Istanbul would carry on to recruit children from different channels for different
purposes. Some elites purchased child slaves of Caucasian and African origin
who were shipped to Istanbul in the nineteenth century. These slaves had some
characteristics that might have resembled the criteria that were sought for selecting
some of the devshirme boys. Just like them, some imported slave children who did
not speak Turkish were complete strangers to the city of Istanbul. In this vein, the
way some slave children were raised and educated in some elite households might
bear a resemblance to how the devshirme boys were trained in some states men’s
palaces when it was practiced. There could be continuum in the selection criteria
of slave children over centuries. This connection will be discussed further in the
following chapters along with the differences in slaves, and their employment.
3.4 Conclusion
This chapter discussed one of the classical Ottoman slavery systems, the devshirme
system, in which the state recruited children to raise and train to be employed in its
administrative and military positions. It briefly addressed the pencik system first
14“Hâlâ beher sene bu Tophâne Abazaları evlâdları bir yasında ve ikisine bâlig iken sehir oglanı olmasın kul
olup satılup devlet-mend olsun içün her sene evlâdların süd analara verüp niçe besik kundak, oglan usak,
gemiler ile Abaza diyârına gönderüp on, on bes yasında Islâmbol’a getirüp pâdisâh musâhiblerine pîskes
yâhûd fürûht edüp çerâg ederler. Iste bizim merhûm Melek Ahmed Pasa ve Siyâvus Pasa böyle Tophâne
Abazası idi. Bunun emsâli niçe bin evlâdları Tophâne’den Abaza’ya gidüp gelmededir. Ve a‘yân-ı kibârları
azdı.” Evliyâ Çelebi Seyahatnâmesi I. Kitap Topkapı Sarayı Kütüphanesi Bagdat 304 Numaralı Yazmanın
Transkripsiyonu – Dizini ed. by Robert DANKOFF - Seyit Ali KAHRAMAN - Yücel DAGLI. (Istanbul:
Yapı Kredi Yayınları, 1996), p.456
39
to understand better why the devshirme system was necessitated in the first place.
Afterwards, it moved on to describing the system and explained its introduction,
implementation, and compatibility with the Ottoman Sharia law. It then proceeded
with the devshirme boys’ selection, education, and recruitment stages. The selection
process revealed that the Ottoman state aimed at recruiting better servants for its
administrative-military posts: more loyal, easily educated, and high quality servants.
After their levy, some children were selected for the palace services, while others
were conscripted in the corps. The ones selected for the palace services were taken
in by some statesmen and trained at their palace or households. This chapter
concentrated on devshirme boys and examined what features those statesmen likely
look for in selecting the boys. It further discussed the ways in which they should
behave and be punished. Finally, this chapter suggested that there could be a
continuum in high-ranking statesmen recruiting children and training them since
some of the characteristics that were deemed important in picking the children
during the devshirme method resemble the features that carried on in the following
centuries as well.
40
4. CHILD SLAVERY IN NINETEENTH-CENTURY OTTOMAN
EMPIRE
4.1 Introduction
Several slave children were brought from different regions into the borders of the
Empire in the nineteenth century. The slaves imported to the Empire were particularly
of white Caucasian and of black African origin. Nevertheless, they were not
brought alone, but with adults combined, perhaps with their parents or relatives as
well. Several methods were implemented to enslave a child: kidnapping, capturing,
abandonment, exchange, or even birth. Upon their enslavement, the ones who were
able to survive against the harsh conditions of the slave trade could end up in the
city of Istanbul. This chapter primarily aims to examine this whole process before
the purchases of slave children by some nineteenth-century Istanbulite elites. With
this aim in mind, I will discuss the methods of child enslavement, following a brief
description of the concept of the “child” in the Ottoman Empire’s context. Then, I
will proceed to discuss the Ottoman slave trade to reveal the ways some slave children
were brought to the city of Istanbul. In what follows, I will discuss the slave
market and other possible places in which slave children might have been sold in
nineteenth-century Istanbul to understand the whole process before discussing slave
children’s employment in some elite households.
4.2 The Concept of a Child
In the modern world, childhood is a stage with several periods broken down into
years, months, or even days. The Ottoman perception of a child was rightly and
understandably different from the modern understanding. However, this does not
suggest that childhood was a fixed, universal concept that refers to a period from
41
birth to adulthood in pre-modern times. On the contrary, there were various differing
views for what the period of childhood meant in the pre-modern times.
As far as Ottoman history is concerned, the connotations of child and childhood
are still obscure for the researchers of Ottoman history. There is considerably a
rich vocabulary to describe the child or childhood in the Ottoman context, and
different concepts were used in different periods of kids (Araz 2017, 32-33). More
specifically, in Islamic law, childhood is divided into two main periods: from birth to
age seven, the “gayr-ı mümeyyiz” (undifferentiated) period, and from age seven to
adulthood, the age of “temyiz” (differentiated) (Araz 2017, 89). Children were often
described using these two terms that derive from Arabic: “saghir/e” or “sâbi/ye”.
These Arabic words mean small, minor, and young to refer to infants, children, and
those who have yet not reached puberty. In the transition to adulthood, there were
additional descriptive words such as “murâhık/a,” “mumeyyiz/e,” and “emred”. The
“e” at the end of the word indicates the gender distinction between boys and girls
(Araz 2017, 88).1
In the view of the Ottomans, the transition from childhood to adolescence required
meeting specific criteria in sexual, physical, and mental maturity. The boundaries
of these criteria were not clear, both in theory and practice. Children’s sexual,
physical, and mental development each carried a different weight in the transition
to adulthood (Araz 2017, 24). Because a ten-year-old girl must have reached the
physical maturity to have sexual intercourse, her de facto marriage was deemed to
be physically possible. However, the same girl could not decide for herself on legal
matters because she was not considered to be adequately competent to do so (Araz
2017, 25). Contrary to the notion that “children were little adults,” Araz argues
that physicians, jurists, and scholars who wrote about children all frame their writing
around a different period of childhood, each with its own characteristics and social
equivalents (Araz 2017, 33).
As said earlier, girls had to be nine years old at the earliest to reach the point of
physical maturity in which they could have sexual intercourse.2 During this period,
they learned about “femininity” (âdet-i nisâ) from their mothers. The children that
were considered self-sufficient were seven years old. They could dress alone, so there
1Ottoman children were depicted in the western sources, see Marianna Yerasimos, “16.–19. Yüzyılda Batı
Kaynaklı Gravürlerde Osmanlı Çocuk Figürleri.,” in Toplumsal Tarihte Çocuk: Sempozyum, 23-24 Nisan
1993, ed. Bekir Onur (Istanbul: Tarih Vakfı Yurt yayınları, 1994).
2For child marriages in Istanbul and Anatolia, see Yahya Araz, “17. ve 18. Yüzyılda Istanbul ve Anadolu’da
Çocuk Evlilikleri ve Eriskinlik Olgusu Üzerine Bir Degerlendirme,” Kadın/Woman: Journal for Women’s
Studies 13 (2012): 98–121.; for some examples on how children died or harmed accidently, see Yahya Araz,
“‘Ölmek Için Çok Erken!’ 17. ve 18. Yüzyılda Anadolu’da Kazaların Sebep Oldugu Çocuk Ölümleri ve
Yaralanmları Üzerine Bir Degerlendirme,” Tarih Dergisi 2012/2, no. 56 (2013): 25–54.
42
was no need to pay them alimony (Araz 2017, 54). The age of legal competence, in
which the right to take inheritance was bestowed upon them, was when the children
reached puberty; this period could vary from person to person. Until the children
reach puberty, they had a legal representative (vâsî ) entitled to administer children’s
property (Araz 2017, 65).
At that time, there were training books that explained how to educate children and
how parents should treat them. These books also contained advice tailored to the
ages of the children. To illustrate, it was written that boys who were starting to
become sexually active were not to be left alone with their elders because they may
become targets of homosexual desires (Araz 2017, 95-96).
From the age of four or five, children began to move beyond the boundaries of the
house: the waterfront, streets, and fields were among their most preferred destinations
(Araz 2017, 82). Of course, the children who went outside would be busy with
other things as well; they might have been forced into begging or stealing or had
been employed as maids.
As in most of the pre-modern world, the Ottoman children could be vulnerable to
disease, malnutrition, and epidemics. Once Lady Montagu asked women of Istanbul
how they could take care of ten to twelve children, and in response, she heard that
“half of these children will die from the plague anyway” (Araz 2017, 77).
In the modernization process, the Ottomans continued to define the boundaries of
childhood in line with the physical and sexual development of the kids. According
to the civil code Mecelle, prepared by Ahmet Cevdet Pasa in the second half of the
nineteenth century, it was accepted that girls could reach puberty at the age of nine
and boys at the age of twelve (Araz 2017, 178). Fifteen years of age is accepted as
the upper limit for both girls and boys to begin puberty. These borders were drawn
in accordance with Islamic law. In other words, the way the Ottomans perceived
children before the Tanzimat period continued its existence until the end of the
Empire (Araz 2017, 178).
Having discussed what the concept of a child means in the Ottoman context, this
chapter will proceed with the methods of child enslavement to show the ways in
which a child could be enslaved in the nineteenth century.
43
4.3 Methods of Child Enslavement
In the nineteenth century, the lands that provided slaves to the Ottoman markets
were outside the Empire’s borders. The primary areas where black slaves were
sourced from were the Upper Nile, West Sudan, Galla, and the Sidama principalities
of Ethiopia, as well as Central African lands such as Vaday, Bagirmi, and Bornu
(Toledano 1982, 15). The bulk of white slaves were of Caucasian origin, primarily
from Georgia and Circassia. The methods used to acquire slaves varied in the
Ottoman realm depending on time and place: capturing, kidnapping, exchange,
abandonment, sale of children, and birth were the methods of child enslavement.
4.3.1 Kidnapping
Kidnapping was one of the ways of child enslavement; it was implemented within the
borders of the Ottoman Empire, disregarding the official attitude of the Ottoman
government. There are significant patterns about how and where kidnappings happened:
usually, some ethnic or outcast groups were kidnapped in the frontier zone.
These can be cited as an example of the abduction of Yaz¯ıd¯ıs, which will be explained
in more detail below. Another pattern is that violent atmospheres and settings are
suitable environments for kidnapping: periods of war or revolt can be given as examples.
The target group for kidnappings was not a large village community that was
always on alert but rather single and unprotected individuals (Toledano 1982, 16).
In the nineteenth century, a particular group of people came forward in the context
of kidnapping: rival Caucasian tribes kidnapped slave girls and boys from one another’s
tribes and sold them to Ottoman slave merchants. Sometimes, professional
slave merchants used marriage as a pretext to acquire Circassian and Georgian girls
to later sell them as slaves. Laz people had also generally been accused of kidnapping
Georgian women and girls. Their reputation as kidnappers persisted during the
Tanzimat Era as well (Erdem 1996, 46). Circassians were also blamed for kidnapping
Tatar, Rum, and Bulgarian girls in Bulgaria (Toledano 1982, 17) and non-Muslim
Ottoman subjects in the Empire after their forced migration at the beginning of
1860 (Erdem 1996, 46). Kidnapping was not only peculiar to the Caucasus region,
but also was widespread in some regions of the African continent, primarily Ethiopia
and regions around Lake Chad where there was a lack of slaves (Toledano 1982, 17).
Another ethnic group that was responsible for abducting people was Kurds from
various tribes that inhabited Southeastern Anatolia and Northern Iraq. Kurds en-
44
slaved Yaz¯ıd¯ıs, Nestorians, and Ya’q¯ubî women and children on some occasions.
For instance, during the Kurdish uprising led by Bedirhan Bey in 1847, Nestorians,
Ya’q¯ubî, and Yaz¯ıd¯ı women and children were taken prisoner and sold as slaves in
the Diyarbakır-Mosul region. Some were released at the request of the British and
the regulations implemented by the Sublime Porte (B¯ab-ı ¯ Ali) (Toledano 1982, 16).
In 1844, the government released some Tayyarî Christians whom the Kurdish leader
had enslaved during the last phases of the Bedirhan Rebellion (1836-1847) in response
to the protests of the British. In another uprising in the region in 1855,
Yezdan Sir enslaved Yaz¯ıd¯ıs and Nestorians in the Gerzan and Siirt regions of the
Empire and sent them to the Mosul region to be sold. These were some examples of
illegal enslavement of both children and adults living in the borders of the Empire.
It must be noted that during the nineteenth century, some foreign embassies, primarily
British and French, closely followed the events in the region; they intervened
and demanded illegally enslaved subjects’ manumission (Erdem 1996, 46).
The kidnapping method could be used to enslave children most of the time in most
places, especially in rural and frontier areas. Undoubtedly, children were the most
vulnerable and unprotected, which made them easy to carry and kidnap, especially
in a violent and insecure atmosphere created by any upheaval and rebellion. To
illustrate, a child was sold as a slave to Iranian Hacı Abdulvehhab by a slave dealer
named Osman in 1855. However, Osman died, and it was understood that the
child was initially free. Thereupon, it was decided to take the money paid by Hacı
Abdulvehhab from the probate inventory of the slave-dealer Osman.3
Enslaving children through kidnapping could be very profitable at first sight, as
children can be considered easy to hide and trade. Additionally, it might have been
easier for them to learn a new language and to adapt themselves to new conditions.
There was probably a difference between the slave boys and girls due to the gender
roles at the time. Their employment areas would differ; for example, slave girls
could be raised as concubines and utilized as domestic servants, while slave boys
could be trained for outdoor tasks. It should be noted that slave girls could be
abused sexually, and similarly, slave boys could possibly be used to satisfy the sexual
desire or carry out fantasies. On the other hand, dealing with slave children could
be difficult since they often did not survive the harsh shipping conditions and could
fall sick more quickly than adult slaves.
The Ottoman government did not always remain silent against the illegal methods
3“Hacı Abdülvehhab’ın Esirci Osman nam sahıstan alıp, Tebriz’den getirmis oldugu 1 nefer gulamın hür
oldugu iddia olunması mebni... mezkur Osman’ın vefatı cihetle terekesinden istifa kılınmak üzere... fevt
olmus olmasıyla bu hale göre mezkur 21 bin gurusun mezkur Hacı Osman’ın merkum Abdulvehhab’a
verilmek üzere bu tarafa gönderilmesi babında..” BOA, A.}MKT.MVL. 74- 59 (H-07-12-1271)
45
of enslavement or illicit use of slaves. In the Tanzimat Era, there were new laws and
rules issued that showed a difference in the punishments. For instance, kidnapping
was a subject matter addressed in the penalty law code of 1858. According to this
penalty code, child abductors would be sentenced to one year in prison. The content
of this code was expanded in the following years (Toledano 1982, 17). However, this
state protection did not cover Yaz¯ıd¯ıs who were kidnapped and enslaved (Erdem
1996, 46).
4.3.2 Capturing
Taking people captive during the war and enslaving them was one of the most
widely used methods of acquiring and enslaving slaves in the “classic” period of the
Ottoman Empire, which corresponded to the expansionist years of the Empire from
1300-1600. Capture was a widespread practice that took place inside and outside of
the borders of the Ottoman Empire. However, the Ottomans stopped taking people
captive during the war and enslaving them after the Ottoman-Russian War of 1828-
29 (Erdem 1996, 44). The Ottoman central government even disapproved of the
enslavement of the war prisoners captured during the Crimean War of 1853-1856.
Nonetheless, wars created opportunities to capture and enslave people of all ages.
Some individuals or small groups wanted to take advantage of the wartime turmoil
and the disorder and confusion of wartime. This method of enslavement through
capturing people during the war should actually be categorized under kidnapping,
not captivity, since the Ottoman central government perceived enslavement by individuals
as illicit (Erdem 1996, 45). In addition, some slaves were caught outside of
the Empire as a result of other small-scale conflicts or battles. For instance, intertribal
conflicts were taking place in some regions of Africa, and several free people,
possibly including children, were captured at the end of these wars to be sold to
slave dealers (Toledano 1982, 17). These slaves were transported to Ottoman slave
markets through the slave trade.
4.3.3 Tribute, Exchange, and Tax Paying
Children could be acquired through indirect methods: they could be exchanged
for valuable or needed products or given as tax or tribute. In general, tribute
played a particular role in the slave acquisition of the Ottoman Empire even though
enslavement through tribute was not implemented since the Empire was not a state
46
that paid tribute. However, it was customary to accept slaves from states dependent
on the Empire or friendly with it. To illustrate, the Crimean Khanate often sent
female slaves as gifts before Russia annexed the Crimea in 1783. In another example,
Tunisia had paid tribute by sending slaves to Istanbul until 1842 (Erdem 1996, 48).
In many African societies, especially during times of famine, children and adults were
exchanged for grains; those who could not pay their debts or committed a serious
crime like murder were sold as slaves. In addition, many African principalities and
kingdoms had sold their subjects into slavery: the rulers of Ethiopia, Galla, and
Sidama regularly sold their subjects, covering some of the expenses of imported
goods consumed by the elite in their principalities (Toledano 1982, 18). Amongst
them were sometimes children; they were sometimes bought in place of tax or sold
in times of famine.
4.3.4 Abandonment and Sale of Children
Another method of child slave acquisition was through the sale or abandonment of
children. Parents, family members, or slave owners sold children into slavery, sometimes
voluntarily or by force. There may be different motivations and reasons for
children to be abandoned or sold; poverty, famine, or war could be the first reasons
that come to mind. The abandonment and sale of children as a method of enslavement
occurred in different parts of the Empire during different time periods. For
instance, some Ottoman subjects had to abandon or sell their children on some occasions.
For instance, some Christian and Muslim children inhabiting Mosul were sold
by their families due to extreme poverty (Erdem 1996, 197). During the migration of
Circassians by the 1860s, with the approval of the slave owner and the slave parents
of the child, some Circassian child slaves were sold, and somehow, they arrived in
Istanbul, considering the great opportunities that were believed to be awaiting them.
Most of the time, some slave girls were willing to endure this painful separation, as
they imagined they were probably going to the elite or Imperial harems in Istanbul
(Toledano 1982, 17).
In regards to Circassians, it can be noted that the sale of children was common
amongst Circassian people (Toledano 1982, 17). Even though they migrated to
the Ottoman Empire, understandably, some carried on their local customs in the
Empire’s borders. To illustrate, an official report was addressed to the Armenian
Patriarchate in 1856. According to the document, it seems that a Circassian named
Ishak had been using an Armenian boy as a Circassian child slave. An Armenian
bishop saved a child’s life by paying 5 thousand kurus. The report requires Ishak to
47
pay the bishop’s money back.4
At this stage, it is necessary to briefly discuss the structure of Circassian society to
better understand why some Circassians brought their own customs to the Empire
and continued enslaving children within its borders. There were two main social
classes in Circassian society: slaves and freemen. Additionally, there was the concept
of “hereditary slavery” in their society. Freemen were the slave owners, and they
sold the children of their slaves. Theoretically, only children from the slave class
could be bought or sold. However, there were occasions when this was not followed;
for example, several Circassians from the free class had to sell their children during
their migration due to the severe conditions created by the process. Not only were
poverty, migration, and “hereditary slavery” reasons for child abandonment and sale
(Erdem 1996, 50), but also “searching for better-living conditions” could be added
to the list. Finally, the demand for slaves coming from the Ottomans should not be
disregarded as well.
Some parents or family members did not always sell children; instead, some children
were sometimes willing to be sold into slavery themselves. At this point, regarding
the nineteenth-century pressure on the slave trade, Erdem (1996, 49) argues that
this willingness of children hardens the intervention in the white slave trade. Also,
if the slave children sold were free initially, they could demand their freedom in the
future, just as seven free children who were sold to the Egyptian governor in 1853
later demanded their freedom (Toledano 1982, 19).
4.3.5 Enslaved at Birth
Another source of slavery is being born of a slave. In Islamic law, the slave is
essentially subject to the status of the mother. Thus, an offspring born to a free
father and a slave mother is a slave, and the property of this slave belongs to the
mother’s master. The exception is that the offspring is born to the master and his
concubine; this offspring becomes free. Likewise, a child born to a slave father and
a free mother becomes free (Aydın, TDV, “Köle”). This means that an offspring
whose parents were slaves could inherit their slavery status in Islamic law. Children
born of a free man and a slave woman were also considered slaves. In other words,
it was technically possible to produce slaves in Islam, but Circassians stand out in
this regard because they had a caste of slaves born into “hereditary slavery” (Erdem
1996, 52).
4BOA, HR.MKT.161- 41 (H-30-01-1273)
48
Nonetheless, keeping the children of slaves as slaves was not reserved only for Circassians;
in some provinces of the Empire where slave purchasing power was limited,
some slave owners benefitted from the Ottoman Sharia law that allows “hereditary
slavery” (Erdem 1996, 53). For instance, in nineteenth-century Trebizond, the
British consul Palgrave observed that children of slave parents or children born to
a free father and a slave mother in Anatolia were considered slaves and sold on
demand. This practice was not applied only in the Anatolian part of the Empire.
In another example, Herbert, the Consul-General of Bagdad, reported that it was a
frequently repeated practice in Arab provinces and that the children of Arab fathers
and African slave mothers were sold (Erdem 1996, 53).
When all methods of enslavement are considered, it can be noted that enslavement
by birth was not the most effective form of enslavement in the Ottoman Empire, as
slaves were manumitted regularly (Erdem 1996, 52). This was in striking contrast
to transatlantic slavery, in which it was easier and more common to obtain a slave
population through the birth of new slaves (Lewis 2017, 13). This difference between
Ottoman slavery and transatlantic slavery also attracted the attention of foreign
observers in the nineteenth century. For example, Elliot records in a report he sent
to the British Foreign Ministry that slaves were granted their freedom after serving
up to seven years; plus, the number of children born into slavery was relatively low
(Erdem 1996, 54).
4.4 Arrival of Slave Children to Istanbul
4.4.1 Nineteenth-Century Ottoman Slave Trade
Upon enslavement through one of the methods discussed above—kidnapping, birth,
sale of children, and tribute-tax payment—women, men, adults, children, and people
that were old, young, black, white, and overall from many identities and origins were
brought to different parts of the Empire through the slave trade. This section will
discuss the ways in which enslaved children were traded to Istanbul during the
nineteenth century. First and foremost, it should be noted that not all slaves who
were enslaved and put on each ship could reach the desired address, and “only the
strongest among the slaves survived the hardships of the road to reach the relative
comfort of city life in the Ottoman Empire” (Toledano 1998, 4). Slaves brought to
the Ottoman Empire were brought from four main trade routes: the North African
route, the Red Sea route, the Persian Gulf-Iraq route, and the Circassian-Georgian
49
route. The North African slave trade sources were some of the regions surrounding
Lake Chad, the main ones being Vaday, Bornu, and Bagirmi. The black slaves were
first brought to Benghazi, Tripoli, and Alexandria by caravan from the coastal cities
of North Africa. Later, the slaves would stop at coastal towns on Mediterranean
and Aegean islands such as Malta, Crete, Cyprus, and Rhodes and from there were
taken to cities, primarily Istanbul, Izmir, and Thessaloniki (Toledano 1982, 21). The
black slaves that were brought along the Red Sea slave trade were mainly Nubians
and Abyssinians and were mainly sourced from the regions of Kordofan and Darfur
in Sudan and the Sidama, Galla, and Gurage principalities in Ethiopia. From here,
they were first taken to the port cities on the African side of the Red Sea by land,
and then they were distributed to various regions. Finally, they were mainly sent to
the Arabian Peninsula by crossing the Red Sea and cities such as Damascus, Mecca,
and Medina (Toledano 1982, 21). The African regions that were the sources of the
Persian Gulf-Iraq slave trade were mainly Zanzibar and Ethiopia. Many slaves came
to Zanzibar from the regions surrounding Nyasa Lake in the South. Both sea and
land were used for this trade route. A ship sailing from the coast of Zanzibar would
enter the Red Sea and the Gulf of Basra. Overland trade was also carried out from
the Basra region towards the inner parts of Iraq and the eastern Anatolian regions
(Toledano 1982, 27).
As stated earlier, some white slaves were imported from the Circassian and Georgian
regions of the Caucasus. Georgian and Circassian slaves were transported by sailing
ships. They were brought to small ports on the Eastern Black Sea coast and from
Trabzon, Samsun, and Sinop to Istanbul, Izmir, Thessaloniki, and Egypt (Toledano
1982, 28).
In regards to the ethnicities of these imported slaves, it should be noted that the Ottomans
classified all Africans as “zenci” (black) or “Arab”, but that they considered
there to be a difference between Africans and Abyssinians. Abyssinian slaves were
neither white nor black slaves, and were darker than whites but lighter than blacks,
so they were characterized as somewhere in between.5 With regard to the white
slaves and their ethnicities, it was understood from their more detailed classification
that the Ottoman central government had more information about the white
slaves, as they knew the tribal names to which the white ethnic groups belonged
and the differences between them. On the other hand, some Western observers were
more precise about African slaves but could not demonstrate the same elaborateness
for the Caucasus region due to the language barrier. So how did the Ottomans
accomplish this? With the help of Circassian interpreters, the Ottoman central gov-
5See Jennings, “Black Slaves and Free Black in Ottoman Cyprus, 1590-1640,” p. 288 for the difference
among the terms “zenci,” or “zengi,” “siyah,” “arab,” or “habesi”
50
ernment overcame this problem and understood the region and the slaves coming
from the region (Erdem 1996, 58).
Having portrayed the ways in which African and Caucasus origin slaves were shipped
to the capital city of Istanbul in the nineteenth century, this chapter now will proceed
with the places where child slaves could possibly meet with some elite purchasers
upon their arrival to the city.
4.4.2 Slave Bazaars and the Purchasing of Slave Children
So far, this chapter has presented the ways in which child slaves could possibly be
enslaved. Above, it discussed the trade routes through which the slave children were
probably brought to Istanbul. Upon their arrival, they were sold in different venues,
including slave markets. In what follows, I will discuss the slave market in Istanbul,
which was one of the places where child slaves were sold to their owners. In this
regard, I will also explain other alternative places where slave children were possibly
sold and discuss who were probably selling them.
Until a monolithic building was built during the reign of Ahmed I, Süleiman Pasha
Mansion, the Old and New Bedestens functioned as slave markets. These places are
near the Grand Bazaar, resulting in the sale of slaves in a central place. A market
for the sale of slaves was located in the exact location in Tavukpazarı, northwest
of the Nuruosmaniye Gate to the Covered Bazar. To be built in the place where
all other valuable commercial commodities were being sold, slaves were also sold in
the center of Istanbul, making them into a commodity easy to access. The slaves
brought to Istanbul were unloaded at the Bahçekapı Square Pier and shipped to
the slave market by ships (Toledano 1982, 52). Slave commerce was conducted in
the Tophane district of Istanbul, which used to be a region where the port was
located and numerous slave merchants resided. Tophane was the central region for
“the larger and more lasting slave-selling enterprises that still stood in the 1860s”
(Zilfi 2010, 216.). Istanbul had “the largest and busiest slave market in the Empire”
until it was shut down in 1846. At this point, one could wrongly interpret that
slaves were not sold and that the slave dealings were discontinued upon the slave
market’s abolition in 1846; however, it should be noted that after its abolition, the
slave business “reverted to the back alley” and implemented private sales at slavers’
or buyers’ households (Toledano 1982, 53.). In this sense, a nineteenth century
contemporary historian named Julia Garnett records that:
“Since the abolition of the public slave market, this traffic is carried on
51
to a great extent by ladies of high rank, some of whom are themselves
emancipated slaves. . .On the arrival of a fresh batch of girls, a broker
is dispatched to the houses of the lady dealers who, if they have any
vacancies, either drive to the establishment of the professional dealer or
have the slaves brought to them for inspection. Children of from six to
ten are most sought after by these amateurs who pay for them some 80 in
seventeen years of age ten times that amount.” (Garnett 1891, 403-404)
The slave dealings were carried out even in coffee houses near the shut-down slave
market (Karamürsel 2015, 134). Since the slave business moved to other indoor
spaces such as homes, slavers and slave businesses were not restricted solely to
Tophane (Karamürsel 2015, 135).
A nineteenth century Ottoman elite, Melek Hanım, recorded that to acquire a slave,
one would go to the Tophane neighborhood to visit the slave merchants there. The
prices of the slaves varied depending on their abilities, appearance, and physical
features. The slaves in the slave market were usually between the ages of twelve
and thirteen; however, there were slaves between the ages of six and seven as well
(Melek Hanım 1872, 157-158).
Who was trafficking in the slaves? Was it a profession only operated by professionals?
Needless to say, the slave dealings were not exercised only by professionals; this
profession was also fulfilled chiefly by non-professionals, which is the essential difference
between the slave business in the Antebellum South and the Ottoman realm
(Karamürsel 2015, 136). In addition to professional traders, several other ordinary
people purchased and sold slaves in the nineteenth century. As Karamürsel reveals,
upon moving to Istanbul, a Georgian man used his networks to acquire Georgian
children to sell in Istanbul or Cairo (Karamürsel 2015, 137).
4.4.3 Conclusion
Before moving on to the employment of some slave children in some of the elite
households of nineteenth-century Istanbul, this chapter first intended to outline the
condition of slave children in the nineteenth-century Ottoman Empire. Following
a brief discussion on the concept of a child in the Ottoman context, the chapter
went over various enslavement methods to better understand how children were
enslaved. In this sense, it explained that some children were kidnapped, captured,
or exchanged in return for a tax or tribute and consequently fell into slavery. Some
children were sold into slavery by their parents or relatives in certain regions, while
others were enslaved at birth. The imported slave children in the nineteenth century
52
were mostly of African and Caucasian origin. Upon their enslavement, they were
sent to the city of Istanbul through varying slave trade routes that utilized both sea
and land. Finally, this chapter explained that some slave children were placed in
the slave markets in Istanbul to be sold until their abolition in 1846 and stated that
after that, the slave trade was carried on through different illicit channels.
53
5. CHILD SLAVES IN ELITE HOUSEHOLDS IN
NINETEENTH-CENTURY ISTANBUL
“One of my companions, he had only one wife but filled his mansion with several
white and Arab concubines and a bunch of little Circassian girls...”
Ahmet Midhat Efendi 1
5.1 Introduction
Regardless of whichever method had been applied to enslave children, some child
slaves of Caucasian and African origin were sent to Istanbul in the nineteenth century.
Before their arrival to the capital city, they had to go through a long, possibly
treacherous journey, passing through the harbors or the polls of the Empire, sometimes
against their will, sometimes by their intention. They were not always alone
in that journey; at times, their mothers or siblings might have accompanied them.
They were likely examined attentively, chosen, and purchased by different social
groups of Istanbul.
In the nineteenth century, some of the well-to-do Ottoman subjects—some of whom
held high-ranking state posts or were successful in their businesses—bought these
imported slave children for various reasons. As they purchased them, they paid
attention to specific criteria regarding the origin, ethnicity, gender, and physical
appearance of the child slaves. Additionally, not knowing Turkish and not knowing
the city were possibly and understandably one of the criteria in the selection process
of the slave children. When they were taken into a home, the slave children were
1“Ahbabımdan yalnız bir zevcesi olan, fakat konagını müteaddit beyaz ve arap cariyelerle ve bir takım
küçücük Çerkez kızcagızlarıyla doldurmus.” Ahmed Mithad Efendi, Esaret. (Ankara: Akçag, 2014), p.14.
Ahmed Mithad Efendi (1844-1912) was an Ottoman writer and journalist in the Tanzimat Era. In his
long story named “Esaret”, two Circassian servants in a household fell in love and were married with the
approval of their master. In their first night, the two servants learned that they were paternal siblings who
were kidnapped from a Circassian village and enslaved when they were children.
54
nurtured depending on how they were intended to be used in the future. They
were taught Turkish, music, embroidery, good manners, along with the etiquette
of the households of their well-to-do owners. In the households, the slave children
were employed as playmates upon being given as presents to the free children of the
household owners.
After examining how children were enslaved and how Caucasian and African children
were imported to Istanbul in Chapter 4, I now analyze the reason why some Istanbulite
elite favored children as opposed to adults. I also discuss what characteristics
they paid attention to in buying slave children. There were child slaves in some elite
households of Istanbul in the nineteenth century, but how was their unfree labor
utilized in the household setting? What role did the intersection of their identities,
such as sex, origin, and age, play in their employment? Finally, I inquire about the
relationship between enslaved children and their well-to-do Istanbulite owners. I
argue whether this relationship could be interdependent asymmetrically. With this
aim in mind, this chapter first briefly discusses the Imperial Household to point out
how and why some high-ranking state officials started separating from the Imperial
Household and establishing their own households. In what follows, this chapter will
discuss the characteristics of the slave children, the motivations and plans of the
elite slave buyers, and the ways in which these slave children were educated and
employed in some elite households. In doing so, it will mainly rely on well-known
contemporary European travelogues and memoirs of Ottoman subjects.
5.2 Households of the Ottoman Empire
5.2.1 The Imperial Household
In the pre-modern period, the dynastic rulers were located right in the hearts of
the imperial structures of the empires (Duindam 2011, 1). This was a characteristic
pattern in which the Ottoman Empire was not an exception. From a top-down
perspective, the Sultan was the head of the Imperial Household, and ruling class
members were recognized as the servants of this household (Findley 1980, 7). Overall,
the Sultan’s household was the basic organizational unit of Ottoman society
(Göçek 1996, 22). These servants (kuls) of the sultan held specific prestigious titles
and posts such as that of pasha or vizier, and belonged to the ruling class (askerîye)
of the Imperial Household (Bâb-ı Hümayûn). The literal definition of the Ottoman
term askerî means soldiers, and the Imperial Household is indicated by the Imperial
55
Gate rather than a regular gate (bâb).
The ruling class was profoundly privileged, prestigious, respected, and excused
from paying taxes in return for their service (Göçek 1996, 21-22). This ruling
class (askerîye) was grouped into three subclasses: those engaged in militaryadministrative
service (seyfiyye), religious service (ilmiye), and scribal service
(kalemiye). Relying on one another, all groups made up the Sultan’s extensive
Imperial Household. The members of this household fulfilled four primary services
for the Empire: administration, defense, tax collection, and justice. At this point, it
must be noted that in this patrimonial structure of the Empire, those of the ulema
ranks like kadı (judge), were exempted from being the “kul” of the Sultan (Erdem
1996, 7).
Besides the ruling class, reâyâ, meaning the flock, was the group of the Empire
that paid taxes. “The ruled had no access to the Sultan’s authority or any of the
privileges associated with it,” (Göçek 1996, 22) yet they were still expected to pay
their fair share of tax to uphold the Empire.
There were specific sources and methods that were utilized to supply members to
the Ottoman Imperial Household (Bâb-ı Hümayûn): first the pencik and devshirme
methods, and later the purchasing of slaves from slave markets at leisure (Göçek
1996, 23-24) by the Imperial Household members or servants. The slaves who were
sent as presents to the Sultan should also not be left out of the discussion. As
seventeenth century Armenian historian Eremya Çelebi Kömürcüyan records below,
it seems some people within the Ottoman upper-class also sent slave girls to the
Sultan as gifts.2
“The captives are lined up in the Slave Market like herds of animals, and
those who come there choose the ones they like and pay the price. Fresh
and stately girls, who are still virgins, are detained if the chief customs
officers like them. The cruel Tatars destroy them after capturing them
and keep only the very nice ones to sell at high prices. After they are
bought at the customs, they are taken to the Turkish bath, washed,
perfumed, and dressed in beautiful clothes. The girls’ price was seven
hundred kurus, but we also saw that they were taken under five hundred.
In Turkish history, it is written that there are captive girls who have
been sold up to a thousand flori. These girls, Georgian, Circassian, and
Russian, are presented to the Sultan as concubines by the aghas, pashas,
khans, and sultans.”3 (Kömürcüyan 1988, 56-57)
2In this present thesis, I utilize the following edition: Eremya Çelebi Kömürcüyan. Istanbul Tarihi – XVII.
Asırda Istanbul. tran. by Hrand D. Anreasyan, ed. by Kevork Pamukcuyan. (Istanbul: Eren, 1988).
3“Esirler, Esirpazan hanı’nda hayvan sürüleri gibi sıraya dizilirler ve oraya gelenler begendiklerini seçip
56
It should be noted that not all servants of the Sultan were slaves. There were freeborn
Muslim officials as well. “The term kul (slave) came to be applied also to those
officials who were not of slave origin, but were freeborn Muslims” (Erdem 1996, 6).
Regardless of the way the servants were acquired, all members of the Imperial Palace
went through a series of special education and training events. As Halil Inalcık
notes, they were trained at the Palace after the recruitment. An educational series
specialized to the Palace was created, consisting of physical training and literary and
religious education. Then, these recruited boys were assigned to particular posts or
services based on their abilities (Inalcık 1973, 79). “The training at the Palace
was highly selective and those pages who were deemed of limited capability were
graduated in the early stages and usually employed as privates or petty officers in the
household cavalry regiments. The rest of the boys, by far the larger group, would
be ’sold’ to Anatolian Turkish farmers to enable them to learn the language and
customs of the Turks” (Erdem 1996, 8). The aim was to recreate the social identity
of these people, introducing them to their new roles in society (Göçek 1996, 24). As
they moved to the provinces as military and administrative officials, these members
formed households of their own after living in the Sultan’s household (Göçek 1996,
22). Hence, they became fully integrated into the Ottoman Imperial Household.
“The members of the ruling class were servants of the Sultan. The sultans may
even have tended to execute askeris in order to confiscate their estates as soon as
possible” (Erdem 1996, 7). However, how should this servant status be perceived?
Were they slaves? If not, how can their status be understood? These questions were
debated exhaustively by Ottomanists. Metin Kunt believes the term kuls refers
to the servants of the dynastic family, while well-known historian Suraiya Faroqhi
favors referring to the servants of the Sultan as servitors (Kunt 1983. Faroqhi 1994,
564. Toledano, 2016, 137). Ehud Toledano, Madeline Zilfi, and Leslie Peirce share
the same view, and they claim that they should be regarded as enslaved people
(Toledano 2016, 137. Peirce 2003, 315. Zilfi 2010, 15, 101–102). The aspect we
should emphasize is not the fact that they could have specific abilities or high ranks,
but the fact that they were highly dependent on the Sultan’s will and that Ottoman
Sharia law did not protect them. Zilfi accepts that these servants of the Sultan were
not humiliated by the act of being sold in an auction, but they were still subject
to being “the property of the Sultan” (Zilfi 2018, 124). Hakan Erdem suggests that
bedelini verir alırlar. Henüz bakire kalmıs taze ve endamlı kızlar, basgümrükçüler tarafından begenilirlerse
alıkonulurlar. Zira insafsız Tatarlar, onları esir ettikten sonra berbad ederler ve yalnız çok güzel olanları,
yüksek fiyatlarla satmak üzere muhafaza ederler. Bunlar, gümrükte satın alındıktan sonra hamama
götürülüp, yıkatılır, kendilerine kokular sürülür ve güzel esvaplar giydirilir. Kızların fiyatı yedi yüz kurustur,
fakat bes yüz altına alınanı da gördük. Türk tarihlerinde ise, bin flori’ye kadar satılmıs olan esir kızlar
oldugu yazılıdır. Gürcü, Çerkes ve bir çogu da Rus olan bu kızlar, agalar, pasalar, hanlar ve sultanlar
tarafından odalılk olarak padisaha hediye edilir.” (Kömürcüyan 1988, 56-57) Translation belongs to me.
57
the complexity of the juridical status of ruling-class members (askeris) increased in
accordance with the rise of the slaves’ ranks in the hierarchy (Erdem 1996, 7-8).
The complexity of their kul status was due to the political culture of slavery among
Ottoman administrative officials (askeris) (Erdem 1996, 10).
5.2.2 Elite Households
The classic method of acquiring servants for the Iimperial Sservices did not remain
as the predominant solution forever. As discussed in the first chapter, the pencik
practice transformed over time, and the child levy method (devshirme) did not
continue until the end of the Empire. According to Victor Louis Ménage, the last
surviving devshirme record dates back to 1705 (Ménage “devshirme” IE2, II, 212).
This had inevitable consequences which led to irreversible changes in the Empire.
These changes and transformations started appearing by the second half of the
seventeenth century and persisted until the fall of the Empire.
When the ruling class (askerî ) realized that they could not rely on imperial methods
and sources, they began to assemble their own men from their network to administrative
positions (Göçek 1996, 33). Rifaat Abou El-Hajj observed that plenty of
men in administrative positions in the Empire had been raised in the households
of Ottoman viziers and pashas appointed to high offices in the second half of the
seventeenth century, or related to them (Abou-El-Haj 1974, 439).
According to Jane Hathaway, Ottoman elite households consisted of one’s relatives
and individuals protected under that particular household’s patronage (Hathaway
2002, 17). All of the men living in the households were not all from the same source.
For example, some household members were ex-slaves bought at the slave market,
while others were relatives of existing household members. Some others were
displaced peasants or artisans who attempted to attach themselves to these households
presumably to receive patronage (Göçek 1996, 23). Nevertheless, it should
be noted that slavery played a crucial role as it was a fundamental “means of expanding
one’s household and networks in Ottoman society” (Erdem 1996, 112). For
instance, Charles Wilkins examines elites who utilized slaves to build their households
in seventeenth-century Ottoman Aleppo. His study shows the ways in which
“slaves were integrated into selected military-administrative, merchant, and ulema
(interpreters of religious knowledge) circles and families” (Wilkins 2013, 345-391).
Regardless of the method of recruitment, being attached to a household provided
shelter, profession, and social status for the person within it. Moreover, a household
membership gave “a sense of belonging and identity” in social and political senses
58
(Toledano 2007, 29).
By the seventeenth century, some offices of the ruling class were physically separated
from the Imperial Palace. The members started moving their residences out of the
Palace. As Ismail Hakkı Uzunçarsılı notes, it was Grand Vizier Dervis Mehmed
Pasha who first brought his office to his residence in 1654 (Uzunçarsılı 2009, 249),
blending the home and workplace This practice was subsequently followed by many.
For instance, an Ottoman defterdar, or head of a financial department, started his
business in his residence in 1676. By the eighteenth century, the Imperial Council
(divân-ı hümayûn) was running the Ottoman administration in grand viziers’ households.
This shift meant that governmental administration was being held outside
of the Palace and inside of the living spaces of many of the administrators (Göçek
1996, 51).
The ruling class members’ houses were also state offices; therefore, men from their
network gathered in the elite households and were trained there. The candidates
were instructed on how to engage in bureaucratic correspondences and develop administrative
or military skills. These newly recruited members of the households
became the new candidates for opening administrative offices and posts (Göçek
1996, 32). These posts and positions could be anywhere in the provinces. Once
they were assigned to a certain post, they became deputies of whoever’s household
they belonged to. The provincial notables consisted of these people, who went on
to establish their own households and form rivalries with one another, competing
for status (Göçek 1996, 52). According to Hakan Erdem, the best example of this
rivalry might be observed between Mustafa Resid Pasha and Grand Vizier Husrev
Pasha. The Grand Vizier, a former slave of Abkhazian origin, famed for educating
and training slave boys among the Ottomans, would assign former slaves to high
administrative posts and exercise power through them. Among several servants attached
to his household, Ibrahim Ethem Pasha, a slave of Chios origin who belonged
to Husrev, became the Grand Vizier of Abdülhamid II. The Tanzimat pashas did
not have this same system of having slaves trained by private individuals to hold
official service positions (Erdem 1996, 64-65). Not only was there consistent rivalry
amongst the households (Hathaway 2002, 19), but also they created alliances with
other households (Hathaway 2002, 17). One of the ways of creating alliances was
through marriage. Inter-household marriage was an essential element of forming
alliances between households to preserve common interests and goals, and to inherit
income-generating economic assets (Toledano 2007, 27).
The decline of the child levy (devshirme) tradition is not enough to entirely explain
the changes and transformations of the rise of the elite and the elite household. In
59
the seventeenth century, a crucial development enabled the upper-class—primarily
pashas and viziers—to accumulate financial power and prestige. According to Metin
Kunt, one development was the introduction of the tax-farming system (iltiz¯am).
The t¯ım¯ar system became outdated, and was gradually replaced with the tax-farming
system (Kunt 1983, 95). Meaning “care or attention,” the word t¯ım¯ar referred to
“non-hereditary prebends to sustain a cavalry army and a military-administrative
hierarchy in the core provinces of the Ottoman Empire” (Inalcık 2000, 501). According
to Mehmet Genç, iltiz¯am meant “a private person collecting any state tax
revenue in return for a certain annual fee” (Genç 2000, 154). Over time, the Imperial
Household started relying on this tax-farming income. The more the Imperial
Household leaned on the revenues of the tax-farming system, the more intense the rivalry
amongst members of the Ottoman ruling class over the provincial tax revenues
(Göçek 1996, 22). Mehmet Genç notes that a remarkable phenomenon emerged in
1695: a lifetime tax-farming system (malikâne) that assisted the Ottoman ruling
class (askerî ) to boost their power through provincial tax sources. (Genç 2003,
516-518)
Not only did these methods of collecting tax revenues help the ruling class members
accumulate their wealth and power, but also their connections with the Imperial
Household aided them. For example, if the sultan’s servant was married into the
sultan’s household, he benefited from his wife’s links to gain close friends and supporters
from the Imperial Palace (Göçek 1996, 26).
5.3 Child Slaves in the Elite Households of Istanbul
To discuss the ways in which child slaves were employed in some of the nineteenthcentury
Istanbulite elite households and why they were chosen over adult slaves, this
chapter has discussed the emergence of the imperial and the elite households at first
to better understand the owners of the enslaved children and the setting in which
they were raised and educated. This part will discuss the characteristics of the slave
children and the potential reasons for purchasing them. Following an examination
on the employment of slave children in elite households, this chapter will proceed
with the relationship between elite owners and the child slaves. This chapter will
conclude by discussing the gender aspect of the child slavery phenomenon.
As my survey of the pre-existing literature on this topic shows, child slaves in the Ottoman
context are not being investigated for the first time. The previous scholarship
was established by utilizing different archival materials and sources to reveal some of
60
the enslavement stories, which sometimes reveals children’s or young adult’s experiences
in bondage in different regions of the Empire. Toledano told the deplorable
story of a Circassian slave girl Semsigül; Troutt-Powell unveiled the stories of two
African slave girls named Yekta and Yasemin; Erdem wrote the story of a young
African slave named Feraset; and recently, Karamürsel has told the story of another
young Circassian slave girl named Fatma Leman.
This chapter attempts to analyze some of the contemporary domestic memoirs and
foreign travelogues on this topic to draw attention to enslaved children’s presence
and their enslavement experience in some Ottoman elite households in Istanbul.
Almost all sources utilized in this research were previously studied in the context of
Ottoman slavery. In this discussion, I bring them up again, attempting to analyze
within the child slavery context.
In the nineteenth century, Ottoman Istanbul had been visited by numerous female
and male travelers, authors, ambassadors, intellectuals, and visitors from diverse
cultures and backgrounds for different purposes. Most wrote about their views,
remarks, and memories of their encounters and ventures. They all are invaluable
to be used to have insight into numerous subject matters, including slavery. In this
thesis, I do not dismiss some of the records of the pre-nineteenth century such as
Lady Montagu’s letters.
Thankfully, some Ottoman individuals, both male and female, recorded their autobiographies
and memoirs. Rich in context, their accounts can be utilized for various
subjects, including exploring the presence of slave children in Ottoman society as
well as in some of the elite households. The degree of how much they witnessed
varies; a few of the writers grew up in a household filled with adults or children, or
some of them even owned child slaves as their playmates during their childhood. A
couple of writers married former child slaves, while some only could observe their
existence in other venues. With regard to domestic observations, I benefit from the
memoirs of Hagop Muntzuri and Asçıdede Halil Ibrahim, Melek Hanım and Leylâ
Hanım’s accounts, and finally the records of Halide Edip Adıvar and Emine Fuat
Tugay.
5.3.1 Characteristics of Slave Children
As will be discussed and analyzed in more detail below, most of the child slaves
mentioned in nineteenth-century memoirs and travelogues were slave girls. This
does not suggest that there were no slave boys. Rather, this might have been
61
related to the association of slavery with women. As Madeline Zilfi rightly argued,
the role of women in domestic servitude gave the domestic slavery a distinguishing
characteristic in the nineteenth century (Zilfi 2004, 1).
Another remarkable aspect is the fact that these slave children were brought from
far away places such as the Caucasus and Africa. They, therefore, did not know the
city, speak the Turkish language, or have any relatives or acquaintances to build a
network. The age of these slave girls started at six or seven at the youngest and was
between ten and twelve at the oldest.
Leylâ Hanım, a nineteenth-century female elite, noted that there were three categories
of female slaves based on the type of service they conducted: domestic
servants, concubines, and child slaves. To be utilized for domestic service, they had
to be relatively tall, well-built, and of moderate beauty. The ugly or badly proportioned
girls among them were called molada, and their price was moderately and
comparatively lower. To be in the concubinage category, they had to be beautiful,
graceful, well-proportioned, and between fifteen and twenty years old. Some child
slaves consisted of those who were bought for the purpose of profit and chosen when
they were minors between the ages of eight and twelve years old. They had been
selected attentively; those who were pretty, well-proportioned, and held promise to
improve even more with age were chosen for this category. They were kept apart
from the common domestic servants, thinking that they could damage their attractive
bodies or hands by engaging in that work (Leylâ Hanım 1994, 60-62).
The sources mention mostly slave girls rather than slave boys. However, there
were slave boys as well. Leylâ Hanım also records some information about male
slaves, especially the African eunuchs who constituted one of the most elite groups
of Ottoman slaves in the mid-to-late nineteenth century (Troutt-Powell 2012, 131).
She explains that most of the male children were kidnapped and enslaved from inner
Africa. “Male children, usually Abysinnian, were captured in the depths of Africa
by men without heart or pity, who then emasculated them between the ages of
eight and twelve years, sold them secretly to Arab slave traders who, in turn, sold
them secretly to others with considerable profit. They were then taken secretly to
Istanbul” (Leylâ Hanım 1994, 92. Troutt-Powell 2012, 92). Once they arrived in
Istanbul, they were tutored in basic hygiene principles, as well as taught how to
read, write, and pray (Troutt-Powell 2012, 131).
A few points should be briefly mentioned from her memoirs in regards to her perspective
on African slaves in general, which indeed reflects the prejudices of some
of the elite circles within nineteenth-century Ottoman society. First, Leylâ Hanım
distinguishes black slaves from Sudan and Abyssinian origin: Sudanese were com-
62
pletely black, and “those from Abyssinia were more of a chocolate color.” Second,
she takes note of the way Circassian slaves perceived African slaves and vice versa.
“The Circassians were frightened when they saw the Negresses; the Negresses sadly
viewed the first white people they had ever seen with great suspicion and considered
them enemies” (Leylâ Hanım 1994, 70). Third, she underlines the language barrier:
“The question of language was for them yet another cause of difficulty because they
could never find anyone who could understand their dialect. Besides, they had no
hope of becoming the wives or concubines of some great figure. They were thus
fatally condemned to the most difficult of household work” (Leylâ Hanım 1994, 71).
To sum up, there was a clear distinction between slaves; their employment differed
based on their gender, origin, and age. The child slavery was one form of slavery. The
child slaves who were imported to the nineteenth century Istanbul and purchased
by some elites of the city were mainly of Caucasus and African origin. They were
enslaved through many enslavement methods, but the most featured ways amongst
them could be the kidnapping. When they arrived in Istanbul, they knew neither
the city nor the language. They were carefully selected by some well-to-do buyers.
Last but not least, there was a racial biases and discrimination amongst the slaves.
5.3.2 Education of Slave Children
As stated above, child slaves acquired through the methods of child slave enslavement
explained in Chapter 4 of this thesis were brought to Istanbul somehow and
did not know Turkish and the city of Istanbul. More importantly, since they were
purchased by some Istanbulite elites and started to live in mansions, they might
have been deprived of the household rules, elite etiquette, good manners, and much
more in the eyes of their owners. Thus, some questions arise: What was taught to
children in some elite households? What was the purpose and motivation in teaching
them?
Unlike slaves levied through the devshirme method, the slave children purchased by
some Istanbulite elite in the nineteenth century were not sent to some families living
in Anatolia to learn about native language and local culture anymore. Differently,
the slave children in the elite households were tutored by their owners or someone
in the household. They were taught Turkish, reading, writing, music, playing instruments
and singing, embroidery, sewing, swimming, and good manners (Leylâ
Hanım 1994, 60-62.). As noted before, they were not allowed to do anything which
might negatively affect their hands and figures (Tugay 1963, 309-310.). It should
also be noted that not every child slave might have gone through the same standard
63
education since this is closely linked to which elite households they were employed
in. It could be also probable to speculate that there might have been a gender-based
difference in the way they were educated.
For example, Leylâ Hanım, who was the daughter of Hekim Ismail Pasha, owned
African and Circassian young slave girls. However, there were two Sudanese slave
girls in particular whose names were Yekta and Yasemin that Leylâ wrote extensively
about. Their story begins with them both being for sale together in the market and
Leylâ’s father purchasing them both at the same time. Yekta was taught how to
read and write, as well as how to do housework, sewing, and embroidery, and was
even instructed in swimming. Based on Leylâ’s account, “Yekta enjoyed swimming
through the Bosphorus every night” (Leylâ Hanım, 1994, 85-89. Troutt-Powell 2012,
130).
In another example, Halide Edip, who was one of the most prominent female Turkish
literature writers and a distinguished social activist for women’s rights, indeed grew
up in households where slaves helped to raise her (Troutt-Powell 2012, 5). Her slave
girl named Reshe was originally from Yemen before she was kidnapped and brought
to Istanbul. She was taught Turkish by Halide, but until she became able to speak,
Reshe was able to calm down by “singing songs from her homeland until she was no
longer able to recall her native tongue” (Zilfi 2010, 151). Reshe learned Turkish and
proclaimed to others that she was traumatized in the way she was kidnapped when
she was playing nearby a group of slave merchants, explaining that she was forced to
walk for days in the Abyssinian forests. When Reshe became able to speak Turkish,
she continued to explain herself retrospectively, explaining the reason why on the
first night in her new home, she was so scared. As Halide notes, Reshe had been
told that “white” people in Istanbul would eat her (Adıvar 1923,168). Speaking the
same language might also be unexpected for the elite owners to observe the ways
in which how they were being perceived by some “black” people living far from the
capital city.
It is unlikely that this phenomenon did not bring to mind the practice of devshirme,
to which the first chapter is devoted. This similarity does not guarantee that the two
different phenomena were closely tied to one another or a continuum from one to the
other. This might only allow historians to consider that the practice of acquiring
and raising child slaves was indeed a long-lasting practice that was well-established
in society but could have been embodied in different forms over centuries. In this
case, it might also suggest that the thesis that the elites may have imitated the
Imperial Household should be reconsidered.
64
5.3.3 Purposes of Purchasing Children
In the nineteenth century, some of the elites of the city of Istanbul bought child
slaves and raised them as described above. It can be said that they did this out
of nowhere and that they had various motivations and goals. As Leylâ Hanım
mentioned, as can be understood from the training given, children were not trained
to be employed in the domestic sphere doing housework such as cooking or cleaning.
The child slave could not have been taught music and reading and writing for these
domestic purposes. On the contrary, they may have been prevented from doing these
activities because they could have been physically harmed. Then why were these
children bought and trained? What were the aims of the elites of Istanbul? This
question, which is the starting point of this thesis, has more than one answer. Now
in this part, I will try to answer in the light of examples from the primary sources.
5.3.3.1 Marrying Slave Children Off with Fine Gentlemen
First, we observe that some slave girls in elite households were married after being
brought up when it was time for marriage. Some slave girls were married to the sons
of the owners of the mansions in which they were raised. Charles White, a British
traveler who was in Istanbul in the first half of the nineteenth century, records that
high-rank ladies bought and raised Circassian girls between the ages of eight to ten
to be good and well-educated wives for their sons (White 1847, Vol II, 321). Some
slave girls were married to not the household members but also other individuals
living in the Empire. For instance, Mustafa Efendi, charge d’affairs of Egypt, was
planning to marry one of his Georgian slave girls who “had already been in his
house two years; and he hoped someday to give her a marriage portion and to see
her comfortably established, as she was a good girl, and he was much attached to
her” (Pardoe 1837, Vol. I, 114).
Marriage may not always be an existing motivation for purchasing and raising a
slave child but may instead be an applied “solution” to send the raised female slaves
out of the household. Melek Hanım, who was the wife of Mehmed Emin Pasha of
Cyprus (1813-1871), did precisely that. In all sincerity, Melek details a memory
of a Circassian slave girl she purchased when she was eight or nine years old. As
she notes, Melek Hanım nurtured and tutored her to become a governess for Melek
Hanım’s young daughter, Aisha, after the Circassian slave turned fourteen years old.
Eventually, she admits that she got jealous of the young Circassian girl’s presence
in the household. Melek Hanım was afraid that her husband, Mehmed Emin Pasha,
65
may want to marry her since it was permitted at the time to marry four women in
Islamic law. When her husband was assigned to a post far away from the capital,
Melek Hanım decided to take advantage of her husband’s absence and marry that
Circassian slave with a widowed soldier to send her away from the household (Melek
Hanım 1872, 102-103).
Some of the elites who raised child slaves married some of them off. However, why
were child slaves bought and raised in elite households preferred for marriage? Did
the elites gain anything in marrying the female slaves that they trained and raised
for years? It can be said that these unions might have brought some prestige and
respect to the elite owners because it could be significant in whose household and
by whom the wedded slave was raised. For example, when the Circassian female
slave was married off as Melek Hanım planned, the groom of the slave girl came to
Melek Hanım’s house to visit her right after the marriage day and kiss her hands in
order to express his gratitude towards her, saying “I am the husband of the young
lady who was brought up in your house” (Melek Hanım 1872, 111).
To what extent could the grooms have gained respect or prestige from the marriages
similarly to how the elite owners did? Now, contemporary accounts whose authors
were married to people who grew up in the mansions or palaces of some of the elites
should be mentioned. First, the marriage of Asçıdede Halil Ibrahim will be discussed,
followed by the marriage of Hagop Muntzuri’s friend. Asçıdede Halil Ibrahim held
many administration posts and was appointed to several different regions of the
Empire. One day, as he notes, his mother started looking for a girl to marry him
since he stayed single for a period of time that was considered long in her eyes.
“Even though my mother is getting old, she and her friends started looking
around for a (decent) person for me to marry. There was a Circassian
concubine, whom Emine Hanım, the wife of a merchant named Mehmed
Ali Aga, brought up from childhood, in a household opposite the deceased
Seyhülislam’s (the chief religious official in the Ottoman Empire)
household in Suleymaniye, below the current Seyhülislam’s household.
Emine Hanım’s mother was also Circassian, and she was a friend of ours.”
(Asçıdede Ibrahim 1960,28)4
As the passage indicates, it seemed essential for Asçıdede Ibrahim to provide the
4“Validem artık etegi elinde oldugu halde ahbapları ile beraber surada-burada kız aramaya basladılar.
Mukadderât-ı Ilâhî Süleymaniye’de Seyhülislam kapusunun alt tarafındaki eski merhum Seyhülislam’ın
konagının karsısında bir hanede Tüccerân-ı Mûteberândan Mehmed Ali Aga namında bir zatın haremi
Emine Hanım’ın çocukluktan tebriye ettigi bir Çerkes cariyesi varmıs. Emine Hanım’ın validesi büyük
hanım da Çerkes olup bizimkilerin ahbabıymıs.” (Asçıdede Halil Ibrahim: Hatıralar, 1960, 28). English
translation belongs to me.
66
location and details of the household in which his Circassian wife was brought up.
That household belonged to a merchant family located in Süleymaniye, and they
were neighbors with the former chief religious official in the Ottoman Empire (Seyhülislam).
Emine Hanım, the lady of the highest rank in the household, raised a
Circassian slave girl here, whose mother was also of Circassian origin. This part of
his memoir gives valuable insight into the social circles of the author. Within the
chapter, Halil Ibrahim informs his audience that his family was connected to Emine
Hanım’s mother in friendship. Emine Hanım perhaps invested in several other
Circassian girls like the one that Asçıdede Ibrahim got married to. Upon their marriage,
Emine Hanım could have gained more status and esteem in society. Asçıdede
perhaps thought of stressing this detail since his wife was not an ordinary Circassian
female slave, but the one nurtured in Emine Hanım’s house.
Just like Asçıdede, Hagop Mıntzuri (1886-1978)’s old friend had married a Circassian
girl who was brought up in the palace of Ibrahim Efendi. Mıntzuri was only ten
years old when he first stepped in Istanbul in the 1890s, where he remained for three
quarters of a century. Unlike his counterparts, he refrains from writing much about
what took place within the mansions of Istanbul. Instead, he provides insight into
the districts, people, and scenery of Istanbul.5 One day, he met a childhood friend
named Selahattin, and he included in his own memoir what was discussed at the end
of that encounter. It was indeed worth mentioning. As the author writes, Hagop and
Selahattin, who have not seen each other for a long time—around 43 years—had
a small talk on what had changed since they last saw each other, realizing that
they both got married and had children. Hagop asked his friend where his wife was
from, to which Selahattin replied with a laugh that his wife was Circassian and from
Dagestan. Selahattin continues, “My mother married me. She was from the palace
of Ibrahim Efendi of the (imperial) dynasty” (Mıntzuri 2008, 76).6
It is known that this event took place in 1940 since the author himself states that
when he calculated how many years it had been since the last time these two friends
saw each other.7 On top of that, it can further be interpreted that the Circassian
wife of Selahattin might have been one of the slaves raised and educated in Ibrahim
Efendi’s household before he married her.
5In this present thesis, I benefit from the following edition: Hagop Mıntzuri, Istanbul Anıları, 1897-1940,
trans. by Silva Kuyumcuyan, ed. by Necdet Sakaoglu. (Istanbul: Tarih Vakfı Yurt Yayınları, 2008)
6“Müsaadenle simdi Ihsaniye’ye esimin teyzesine gidecegim. - Esin nereli? .. Çerkez. Dagıstanlı. (Gülerek)
Saraylıdır, annem evlendirdi, o buldu. Hanedan’dan Ibrahim Efendi’nin sarayından” (Mıntzuri 2008, 76).
7“1897’den 1940’a hesapça kırk üç yıl olmus” (Mıntzuri 2008,76).
67
5.3.3.2 A Tool of Investment
As noted above, slave children were possessed for many purposes by some elites
of nineteenth-century Istanbul. They were likely carefully selected amongst many
other child slaves at the age of seven or eight and raised in households with special
training. So much attention possibly was paid to them that they were not given
any classical domestic tasks that could cause physical harm. Some children were
acquired and raised for arranged marriages, while some others were to be sold at
higher amounts in the future. Through the slave dealings, some elite slave owners
might have made financial profits. In this part of the present thesis, I will discuss
the reasons why some elite owners sold their slaves with the intention of making
profits, as well as who the slave children’s potential buyers were.
Selling the slave children when they grow up could have possibly appeared when
the owner (1) did not require the child slave’s servitude anymore, (2) got jealous
and suspicious of child slave’s existence around their husbands, (3) simply desired
to emancipate the slave by her own will, (4) decided to expand her network and
strengthen her social status, or lastly, (5) simply desired to make financial profits.
A contemporary observer, Lucy Garnett provides us with significant insight into
one of the reasons for high-ranking ladies’ interest in selling the slave girls. In the
following quotation, Garnett writes about a nineteenth-century Istanbulite female
elite, the wife of Minister Fuad Pasha named Behiyeh Hanım, and her desire to
invest in child slaves. As Garnett notes:
“The desire to possess an income independently of her husband had in
the first instance induced this lady to engage in this profession, and
her investments had turned out so well that in a few years, her house
was transformed into a vast training school of saleable maidens. Having
at one time a very large stock on hand Behieh Hanum became a little
anxious as to the possibility of finding a ready market for them all and
finally decided to have recourse to sorcery.” (Garnett 1891, 407)
If Garnett’s account is to be accepted, the exaggerated tone of the text should
be disregarded. This way, it suggests several significant insights. First, it points
out that some elite households were training their child slaves. They might have
presented an education service, and therefore can be qualified as a training place.
Second, it answers the questions mentioned above: why did some Ottoman elite
women desire to make profits? As Garnett notes, they might have wished to have
an income independently of that of their husbands.
68
To whom were they sold at higher prices, and why? Leylâ Hanım speaks of the
prospective buyers, for whom these slave girls were prepared. These girls were sold
to slave merchants who came to the palace and “always looked for good material.” It
seems that there had been offers, perhaps even through a process of bargaining, since
the slave dealers “harassed the lady with all sorts of offers.” It can be understood
that the lady of the palace did not accept all of the slave dealers’ offers, since
Leylâ Hanım notes that “many times she made objections to the proposals” (Leylâ
Hanım 1994, 61). The prospective buyers of these well-educated and high-quality
slave girls were “always Egyptian princes, Tunisian Beys, rich moguls from India,
wealthy merchants from Baghdad or Yemen, and fine gentlemen from all the corners
of the world” (Leylâ Hanım 1994, 61). Just like that of Leylâ Hanım, Emine Tugay
Hanım’s account also gives insights on the customer profile of these raised slave girls.
“On reaching marriageable age, these slaves were sold at high prices
to the Egyptian princess, the Bey of Tunis, or millionaires in Muslim
countries. The price of these girls varied from several hundred to a
thousand gold pounds” (Tugay 1963, 309-310).
It could be said that some female elites’ desire to sell their slave children after
nurturing and educating them was indeed not restricted to the capital city. Still, its
impact extended throughout the Empire’s lands, perhaps connecting the periphery
to the center. It even sometimes exceeded the Iimperial territories. Here, it appears
that the slave merchant played a role in forging connections between sellers and
buyers, connecting households to households and empires to empires. They could
have acted as mediators, which allowed them to make profits from this business in
return.
5.4 Employment of Slave Children in Elite Households: Playmate-Gifts
Some well-to-do Istanbulite members of the Ottoman society obtained imported
slave children; they chose children instead of adults in the nineteenth century, and
the reasons for purchasing and employing the slaves in the households differed. As
noted earlier, some girls were given as “playmate-gifts for children of the prosperous
or daughter-attendants for the lady of the household” (Zilfi 2004, 7). Some were
purchased to be good and well-educated wives for high-rank ladies’ sons (White
1847, Vol II, 321), while some slave girls were acquired to be sold with high profits
in the future (Erdem 1996, 123).
69
Gifting child slaves to other children as playmates was not a practice implemented
only in the Ottoman Empire. Some child slaves were given to other children as
presents in some parts of America before emancipation. Unlike the Ottoman memoirs
on this subject, which help us draw a framework only at a certain level, retrospective
records of the adult slaves who experienced this when they were slave
children in America before emancipation has helped build a few stories. As a result,
a few theses, books, and articles have been produced on the subject.
Analyzing child slaves in the Ottoman context makes it possible to consider the
existence of child slaves in the household from many perspectives and contexts.
However, this study deals with child slaves in the context of domestic slavery. The
responsibilities and duties given to child slaves should undoubtedly have been different
from their adult counterparts. Beyond arguing or attempting to prove that child
slaves perform domestic services in the classical sense, such as cleaning or cooking,
this section will examine the tasks and employment examples they could fulfill as
children. In some elite households of nineteenth-century Istanbul, it can be observed
that child slaves were given as gifts to other children and in turn became their playmates.
In what follows, the study first problematizes the phenomenon and then
discusses the employment of some child slaves as playmate-gifts in the households.
When a well-to-do father, out of his “generosity,” gave a slave to his or her child to
be a playmate, sometimes as a birthday gift and other times for no reason, it could
be considered a complicated situation from many angles. From a legal perspective,
it meant that a father gave the property to his or her free child with no legal rights
since slaves were considered property. Second, children have been perceived as
the domain of others, particularly their parents, “because many children are often
treated de facto if not de jure, as the property of their parents, [thus] distinguishing
between the slave child and the free child can be a very difficult process” (Anna Dua
2017, 9). As a result, upon being delivered as a birthday present to another child,
the slave child’s “subordination and vulnerability are reinforced by becoming the
‘property’ of a person who was not capable legally of ownership independent from
her parents’ legal guardianship” (Sarah Winter 2017,54).
The free child’s first contact with the “slave friend” probably happened at the very
first moment when his father presented his “generous” gift to his child. It might
not have been noticed at first, but the free child would eventually realize that his
playmate-gift was different from the other slaves in the household, as this one belonged
to him or her. They would not be like siblings, but not fully like a master
and a slave either. The relationship would be different from the binary adult slavemaster
relationship, something specific to children. The child slave would have to
70
go if his master said come, and if the child master said let’s play, he or she would
have to play. Playing games may not seem like a very “heavy” and “serious” task
at first glance. However, it is questionable how much the slave children enjoyed the
games. Did the slave children have the right to win games, or be happy and rejoice
in the games they played with their child masters?
“In Bangladesh, an eight-year-old girl summed up the inequity governing
her relationship with her employer’s children: “When I play with
the master’s children, I must always let them win” (Blagbrough Craig
2017,159).
Does playing or being playmates constitute child slaves as a part of domestic slavery?
If a child is enslaved in a household and a task is assigned to him which requires
unfree labor, it could be considered a form of domestic slavery, especially if there is
no environment to allow the child slave to express he or she does not want to play.
What happened in a nineteenth-century Ottoman elite household when Halide Edip
was given a slave girl on her birthday by her father is worth mentioning. Born
in Istanbul, Halide Edip Adıvar was one of the most prominent female Turkish
literature writers and a distinguished social activist for women’s rights. One day,
Halide Edip was given a female slave as a birthday present from her father (Adıvar
1923, 166). That slave girl was somewhere between eight and ten years old and from
Yemen. Her name was Reshe. As Halide notes, Reshe seemed scared on the first
night she arrived. Halide, Reshe, and her sister slept the whole night in the same
room (Adıvar 1923, 166). Finally, her sister got scared of Reshe and asked Adıvar,
“What if Reshe was a cannibal and ate me?” Halide checked Reshe’s teeth and told
her that they looked the same as theirs. Next, her sister Nilüfer asked Halide to
check if she had a tail, but as Halide notes, she did not (Adıvar 1923, 167). The issue
of teaching Turkish should be noted. It is not difficult to guess that the slaves would
describe what happened to them or probably express their sadness about it upon
learning Turkish. How ready were the Istanbulite elite child slave owners who hoped
their slaves would speak the way they wanted, perhaps with an Istanbul accent, to
listen to what the child slaves went through? Did they expect this would happen?
Furthermore, it is predictable that there must have been some house rules that had to
be followed. How were these taught to a small slave boy or girl who had just arrived
and probably did not even know Turkish if there was no means of communication?
What punishments were given to child slaves, who likely sometimes did not follow
the rules or perhaps forgot? Unfortunately, neither Halide Edip’s account nor other
contemporary sources provide insights on this subject.
71
Both children, free and slaves, who were nurtured together in the same households,
could have “remained close as they matured. Usually these close bonds persisted
only between white children and slaves who became their personal servants. A
trusted playmate from childhood often became a trusted servant as an adult in the
household.” This is exactly what happened between Halide Edip and her playmate
Reshe when they grew up. When Reshe learned Turkish and explained everything
she went through, such as how she was kidnapped when she was playing, enslaved by
slave hunters, and shipped to Istanbul, subsequently, Halide promised Reshe they
could dress alike and that she would have a beautiful room. In addition, Halide
wrote her a “liberating paper,” hoping that this would mean that Reshe would no
longer technically belong to her (Adıvar 1923, 169). Even though Reshe was granted
a manumission paper, she remained in the same household. “When Halide married,
Reshe moved with her to her new home. They grew into women together, and Halide
kept her promise about the clothes” (Troutt-Powell 2012, 143).
5.5 Relationship Between Slave Children and Elite Owners
The degree of dependency in the domestic servitude that one slave might experience
also reveals differences in regard to where, when, and in whose household a slave
was employed. This also affected the intensity of the enslavement experience of
domestic slaves. Investigating the relationship between the African and Circassian
child slaves and their Istanbulite elite owners, I propose an asymmetric dependency
instead of a conventional and binary dependency model in which the children slaves
highly depended on their elite masters.8 In light of the sources discussed above, it
appears high-ranking elites of Istanbul highly profited from the African and Circassian
children’s enslavement. Those elite ladies who possessed Circassian children
slaves were primarily the wives of high-rank state officials or greatly respected and
wealthy merchants, but male elites played a significant role as well.
Some child slaves were owned at around seven or eight years old by some of Istanbul’s
elite and raised in their households. They were provided shelter, clothing,
food, healthcare, and education, along with a sense of belonging and identity. In
this regard, being brought up in an elite household could distinguish those Circas-
8I found inspiration to use the concept “asymmetrical dependency” from the studies, conferences, and
publications of Bonn Slavery and Dependency Studies at University of Bonn. I intended to adapt the
concept and the perspective to my research to assess the relationship between some Istanbulite elites
and their enslaved children. I had previously presented a paper in an e-symposium organized by BCDSS
in March 2021, where I attempted to analyze Circassian young slaves from “asymmetrical dependency”
perspective for the first time.
72
sian female slaves from other Circassian female slaves in the market or society and
provided an (in)visible status for those slaves, since the household the slave belonged
to possibly mattered in shaping their lives.
At first sight, it seems that only the child slaves were profoundly dependent on their
owners. However, some of their elite owners were similarly significantly dependent
on them. By possessing Circassian and African child slaves at young ages, it could
be said that they had initially acquired a social occupation for themselves: some
high-ranking ladies now had considerably significant roles. They perhaps believed
that they saved the lives of those little girls, raising them, teaching them housework,
and marrying them off. As the little slave girls, who they raised from childhood,
went out of the household by marriage, some elite ladies of Istanbul also saw their
courtesy, decency, and virtue leave the household. Thus, their elite status and
prestige in society could be increased. Not only that, but some slave owners also
sold those child slaves after raising them as some female elites used to purchase child
slaves at lower prices to sell them in the future at higher prices. Therefore, they
would make high profits. In this sense, it would not be wrong to propose that some
female elite of nineteenth-century Istanbul were investing in Circassian child slaves
brought from the Caucasus.
The relationship between Circassian and African child slaves and their female elite
masters could be interdependent. If so, the reciprocal dependency between them
would not be symmetrical since some owners had the right to control and make
decisions on behalf of their slaves. For example, if owners were to decide to sell
or marry them off or what and how to teach them. Also, they might have shown
motherliness and masterhood, but this depended on their will, again.
Needless to say, everything did not go well all the time. Some slaves who were
acquired to be raised for different purposes with different motivations could sometimes
attempt to escape and end the relationship with their owners. Since there
were newspapers and a culture of making missing persons reports in the Ottoman
Empire in the nineteenth century, some runaway child slave reports were able to be
detected. For example, Cerîde-i Havâdis (Journal of News) was one of the first newspapers
published, and it released several missing persons announcements, including
several written about slaves. On April 1, 1843, while walking around Rumelia, a
runaway ‘Arab’ (Afrikan/black) slave aged thirteen or fourteen was identified and
caught. He was being asked about his owner, but he could not reveal who the owner
was, either his name or his profession. In response, he was handed over to Tophane-i
Amire (The Ottoman Royal Cannon Foundry), where he stayed for more than fifteen
days, but still the owner was not found. Therefore, if any ‘Arab’ slave that got lost
73
matched the released description, they could visit the Tomruk Room in Tophane-i
Amire and try to identify the slave.9
In another report of a missing person in Cerîde-i Havâdis, an ‘Arab’ (Afrikan/black)
slave, estimated to be around ten or eleven years old, escaped from Istanbul on
December 16, 1843. While he wandered around the Rumeli Lighthouse, the officers
caught him and immediately took him to Tophane-i Amire. The owner had not
been revealed yet, as the child slave did not know the name and profession of the
owner. Those slave owners who had such a loss should refer to the Tomruk Room
in Tophane-i Amire.10
At this point, it should be noted that these child slaves’ escape attempts are known
not because they were successful, but because they failed. In light of these two
missing reports, it can be noted that the caught slaves were not held by the local
kadı as it was done during the classical age; rather, they were being held in Tophane-i
Amire (The Ottoman Royal Cannon Foundry) during the Tanzimat Era.
Some relationships continued even after the emancipation of the slaves. As it was
discussed above, the relationship between Halide Edip and her playmate-gift slave
girl named Reshe carried on even after Reshe was granted “liberating papers.”
5.6 Gender Aspect: Slave Girls, Female Owners, and Women Writers
In the Ottoman context, slaves were to be found in several aspects of life at any
given point in time (Toledano 2002, 57). In general, slave labor was practiced in
different areas and settings: domestic or outdoor tasks, respected or unrespected
works, duties they were qualified for or unqualified for (Zilfi 2018, 18). However,
what form of dependency and slavery one might have encountered in the Ottoman
realm was highly dependent on time, space, and context. For instance, the most
prominent “share of the demand for slaves in the Tanzimat period was generated
by the urban classes, and the bulk of slaves were employed as domestic servants”
9“13-14 yaslarında bir Arap Köle Rumeli taraflarında gezerken, fark edilerek yakalanıp sahibi sorulmus,
ancak sahibinin kim oldugunu, adını ve söhretini söyleyememistir. Tophane-i Amire’ye teslime edilerek
15 günden fazla kalmıs ancak sahibi çıkmamıstır. Bu esgale uygun Arap köle kaybeden varsa Tophanei
Amire’de Tomruk Odası’na ugrayarak bilgi sahibi olabilirler.” Ahmet Oguz. (1994). Ceride-i Havadis
Gazetesi’nde çıkan haber yazılar ve degerlendirmesi. MA Thesis. p.95. This reference in the MA thesis of
Ahmet Oguz was not found in the cited volume and number of Ceride-i Havadis newspaper.
10“Tahminen 10-11 yaslarında bir Arap köle Istanbul taraflarınan kaçarak Rumeli Feneri tarafında gezerken,
memurlar yakalayarak Tophane-i Amire’ye getirmislerdir. Sahibinin isim ve söhretini bilmediginden sahibi
henüz ortaya çıkmamıstır. Bu tür kaybı olanların Tophane-I Amire’de Tomruk Odasına basvurmaları gerekmektedir.”
Ahmet Oguz. (1994). Ceride-i Havadis Gazetesi’nde çıkan haber yazılar ve degerlendirmesi.
MA Thesis. p.95 This reference in the MA thesis of Ahmet Oguz was not found in the cited volume and
number of Ceride-i Havadis newspaper.
74
(Erdem 1996, 62). In the nineteenth century, most of these domestic servants were
female slaves predominantly of African, Circassian, and Georgian origin (Toledano
2016, 148). From a gendered perspective, the role of women in domestic servitude
gave the domestic slavery a distinguishing characteristic in the nineteenth century
(Zilfi 2004, 1).
When one looks at the available primary sources, the bulk of the travelogues or
memoirs shed light on the Ottoman domestic servitude in the imperial or elite households,
were penned by Ottoman, Ottoman-Turkish, and European women writers.
Concerning the usage of these sources as primary sources, one should filter them
from a critical eye and prove their findings and observations by comparing them to
one another, and even with archival sources if possible, since their narrations might
reflect their perspectives, preconceptions, and general subjective understanding of
what they saw.
During the mid-to-late nineteenth and early twentieth century a remarkable number
of European female travelers wrote of their adventures and encounters in the Ottoman
Empire. Why did harem and Imperial Households attract such extraordinary
attention? As Reina Lewis notes,
“There is no denying it–as a topic, the harem sold books from the eighteenth
century on, whether you wrote about living in one, visiting one,
or escaping from one, any book that had anything to do with the harem
sold. Publishers knew it, booksellers knew it, readers knew it, and authors
knew it. And women the world over who had lived within the
harem’s segregating system knew it too” (Lewis 2004, 12).
How and when did this trend emerge? Why were female writers interested in the
lives of those living in the harem to this extent? It can be said that “travel outside
Europe was a male experience” before the eighteenth century (Melman 1992, 10).
Lady Montagu’s Embassy Letters, written in the first half of the eighteenth century,
is the first example of secular work. Montagu’s letters played a crucial role in
making harem literature a tremendous and attractive area for women to venture into
intellectually. This new phenomenon, “curiosity about the Mediterranean Orient,
arose about the time of Napoleon’s conquest of Egypt in 1798” (Melman 1992,
10). By the second half of the nineteenth century, the harem theme was found
predominantly in the work of women who wanted to produce written work exploring
differences across cultures (Lewis 2004, 13).
Nevertheless, not everyone was able to see the Imperial harem; as the fascination
around the harem grew, it became more challenging for passing travelers to earn
75
admittance to the harems every year (Lewis 2004, 14). Consequently, a particular
literature on the harem in the elite households emerged. That is why foreign women
began “to visit and report on Muslim households from less elevated social strata,
providing a view of harems of the middle classes rather than only the imperial elite”
(Lewis 2004, 14-15).
Regarding the conditions and the developments of the nineteenth century, there
could be other reasons and motivations behind the passion for the Mediterranean.
First of all, maritime and land transport improvements had raised women’s mobility,
and women could more easily travel by train. With these developments, travel
to the Mediterranean became more accessible and more affordable after the 1830s
(Melman 1992, 11). More importantly, European imperialism made several world
regions “safe and secure” for women travelers. “This meant that more women were
visiting the Middle East, still most often with male relatives but sometimes now
alone” (Lewis 2004, 14). Finally, the Orientalist literature production discussed
above made the East more “attractive and appealing” to visit in regards to women,
folklore, clothing, the harem, and social life. In the context and methodology of
Ottoman history, these texts are highly significant since they are the fundamental
source of information on harem life and domestic slavery in Ottoman households,
which were extremely difficult to obtain.
In the Ottoman context, most of these sources were written by female writers who
visited Istanbul and stayed for a specific time period for varying purposes. It should
be underscored that even though there are alone travelers, some of these European
women visited Istanbul with their husbands or any family members who were appointed
to specific diplomatic posts in Istanbul, such as ambassador. This aspect
also explains how these writers were able to observe some elite households’ harems,
so they could write about slave children as many times European visitors or high position
holders were invited to Ottoman statemen’s households. For example, French
historians Jean-Henri Abdolonyme Ubicini and Theophile Gautier, who were probably
in Istanbul more or less at the same time as Julia Pardoe, also might have
contacted and dined with some Ottoman state officials in the same way that Pardoe
did. Nevertheless, they do not provide any insight about the child slaves in the
households they visited unlike Julia Pardoe. Instead, they both provided similar
details that caught their attention, such as the structure, decoration, and visitors of
the household, the physical appearance of the people, and the taste of the food.
Still, Ubicini in particular paid attention to the forms and practices of slavery since
he provides significant information about the legal aspects of the practice of slavery,
such as the methods of manumission, the slave market, the different forms of slavery,
76
and details about Circassian, Georgian, and Abyssinian slaves in general (Ubicini
1856, 147-152). The reason he did not observe the slaves but rather wrote about the
institution of slavery is perhaps because he was not in the harem part of the mansion,
which was separated for only women, but in the selamlık section of the house, where
the part of a house was reserved for only men. If that is the case, then this is highly
related to why most of the available primary sources on domestic and child slavery
in the households are produced by women writers. In regards to the child slavery in
the late Ottoman Istanbul, the majority of the slaves were girls because the slavery
obtained a distinguishing characteristic in the nineteenth century. The owners who
nurtured and tutored them for specific employment were mostly female elite owners
of the slave girls. The authors who recorded the phenomenon of the high-ranking
female elites of Istanbul raising slave girls were foreign women. The women were
the ones to write about women.
5.7 Conclusion
This chapter dealt with Caucasian and African child slaves in some elite households
in Istanbul in the nineteenth century. In that time period, Circassian and Georgian
children from the Caucasus region and mostly Sudanese and Abyssinian children
from Africa were enslaved, and some of them were brought to Istanbul. By analyzing
local and foreign travel books and memoirs, this chapter attempted to touch upon
the employment areas in which child slaves could possibly be used in some elite
households in Istanbul, their differences compared to adult slaves, and the reasons
why the elites could prefer children. This chapter showed that some elites of Istanbul
in the nineteenth century were purchasing Circassian and African child slaves and
upbringing them in their households. After upbringing them, some owners were
later selling them with higher profits for varying intentions, including investment
purposes. Others were marrying them off with fine gentlemen inside and outside of
the Empire’s territories. As a result, they perhaps were able to increase their elite
status, prestige, and network. African slave girls were given as playmates by some
elites to their children as birthday presents. Also, it is argued that some elite owners
and child slaves had a mutually dependent relationship, that not only the children
were dependent on their masters, but also some of the elites were dependent on
child slaves to maintain their prestige and status. Last but not least, this chapter
discussed the gender aspect of this phenomenon and the fact that foreign female
authors were writing about the purchasing and upbringing of slave girls by female
elites of Istanbul. It argued that female travellers were able to enter harem parts
77
of the elite households, which is why they could observe and write about the slave
girls of the late Ottoman Istanbul.
78
6. CONCLUSION
Relying on some selected nineteenth-century travelogues and memoirs, not to mention
the pertaining secondary literature, this thesis examined the enslaved children
in some elite households of nineteenth-century Ottoman Istanbul. The existence of
child slaves in the elite households was not unearthed in this study for the first time;
they were already known. However, many aspects of their presence and employment
had not been investigated thoroughly before. This study intended to explore some
unrevealed aspects of the institution of child slavery in the late Ottoman Empire.
With this aim in mind, it primarily argued why some Istanbulite elites demanded
slave children instead of adults and the ways in which they used them in their households.
This inquiry consequently revealed some aspects regarding slave children in
some of the elite households of the Empire, such as the desired characteristics in
selecting the child slaves, the motivations of the elite slave buyers, and the ways
child slaves were employed in nineteenth-century Istanbul. It further aimed to examine
the relationship between the owners and their unfree children. It suggested
that not only the child slaves were dependent on their elite owners, contrary to what
might have been assumed at first sight, but also the owners were dependent on their
presence. Therefore, it was instead a mutually interdependent relationship. However,
this interdependent relationship between the two was not equal since the elites
that owned child slaves had the authority to make critical decisions over their slave
children. This thesis suggests that this relationship was most likely an asymmetrical
dependency. After analyzing that, this study finally discussed the link between slave
girls, female owners, and women writers through a gender perspective.
In order to contemplate the child slavery phenomenon in the nineteenth-century
Ottoman Istanbul thoroughly, this study first intended to address some of the classical
Ottoman slavery systems to discuss other forms of child enslavement had been
used in the Ottoman Empire before the nineteenth century. Before delving into
the enslaved children in the late Ottoman Istanbul, the thesis briefly discussed the
devshirme system in which some children living inside the borders of the Empire
79
had been levied for administrative-military purposes. Nevertheless, the devshirme
was not the first formula that the Ottomans had developed to recruit soldiers and
train loyal administrative officials. It was the pencik system that the Ottomans used
before, taking one-fifth of the war captives from its soldiers.
Using some fifteenth- and sixteenth-century Ottoman historical chronicles, the first
chapter discussed when the pencik system was first introduced, by whom it was
introduced, and the possible reasons behind this idea. In addition to that, the ways
in which the selected captives who were then named "pencik boys", were trained and
assigned afterward were also mentioned. The Ottoman state did not only collect
one of the five captives as taxes for itself. Soon after, the cash equivalent of the
captive’s value was introduced as an alternative tax option. This tax was also
named the pencik tax. Over time, this monetary pencik tax had been changed to
a general slave tax collected on all imported slaves. Thus, it would be determined
who paid the tax and did not pay the tax. A document called pencik was given for
the detection of who illegally brought slaves to the Ottoman lands. In short, the
term pencik and its meaning alters depending on the context.
With regards to the classical pencik system that called for the collection of one-fifth
war captives, the chapter demonstrated some possible problems that might have
been encountered while applying the pencik system and what might have been done
to eliminate those problems. Perhaps one of the most critical problems encountered
was that not all of the one in five captives collected were as the Imperial Household
wanted. The Imperial Household might not have been able to educate all of the
captives with various identities, from the ages of seven to seventy years old, and
may not have been able to upbring loyal statesmen amongst them as it desired.
Nevertheless, some problems may have been solved, but some structural problems
could not be sorted out like the one mentioned above. That is why the Ottoman
Imperial Household might have decided to introduce the devshirme system. In this
sense, this thesis briefly discussed the devshirme system. It attempted to reveal
which new features were brought up in this new formula, unlike the pencik. Certain
rules and procedures had been established and revised about the selection, education,
and employment process of the boys. In regards to the first one, this time only
mostly children of a specific age would be levied to increase the likelihood that they
would be educated and trained more easily. In addition, although it is not known
how much it could have been applied in practice, certain attention was probably paid
to the fact that the children of the devshirme were brought up from good families,
did not speak Turkish, and did not know the city of Istanbul. This thesis argued
that these characteristics might be significant as they could resemble those sought
80
by the nineteenth-century Istanbul elite in their choice of child slaves.
Having explained the classical Ottoman slavery systems to discuss one of the longstanding
form of child enslavement, the present thesis moved on to the nineteenth
century. To give a full picture of what the child slaves went through before they
were brought to Istanbul and sold in the nineteenth century entered some elite
households, it first addressed the ways in which some children were pulled into
slavery, were exported to Istanbul’s slave markets, and met their owners. Thus,
the study discussed the enslavement methods of children, nineteenth-century slave
trade routes, and their conditions. It appeared that the slave girls and boys found in
nineteenth-century Istanbul were primarily of Caucasian and African origin. At this
point, this study presented that child slaves were purchased for different purposes,
and that as a result, their employment areas also differed, as well as their futures.
Despite the ban on the African slave trade in 1857, the black slave children continued
to be imported into the capital city. The white slave trade’s volume was affected by
the Crimean War with Russia in 1853-1856 in that there was a considerable increase
in the export of Circassian children to Istanbul, something which carried on into the
following decades.
After the arrival of the children, some child slaves could be sold openly in the slave
market in Istanbul until its abolition in 1846. In accordance with the pre-existing
literature, this thesis showed that the sale of slave children was not restricted to the
market but carried on in several alternative venues. In nineteenth-century Istanbul,
the social group that mainly possessed child slaves was some of the elite subjects of
the city. Most of those who played an active role in the “investment” of child slaves
were female subjects. Having child slaves, raising them, and educating them could
have provided them with prestige, a social occupation and purpose, and a reliable
person to keep in their networks in the future. This study proposed to refer to the
fondness of some of Istanbul’s elites for child slaves, their future plans for them, and
all other aspects of this process as an “investment.” As this thesis presented, some
elites could soon sell the slaves they acquired at a young age for higher prices in the
future after giving them a good education, ensuring they spoke excellent Turkish
or another language too, as well as building up their talents through training them
to play an instrument, cook, or do embroidery work. In this way, this dissertation
showed that some elite women were able to make financial profits from selling or
marrying off their slaves after raising child slaves from a young age.
This thesis examined the ways in which enslaved children were employed in some
of the elite households of nineteenth-century Ottoman Istanbul. From a gendered
perspective, it can be seen that slave owners had different investment plans for girl
81
and boy slaves. According to the primary sources based on domestic and foreign
observations, which primarily were nineteenth-century memoirs and travelogues, this
thesis showed that some women with status residing in Istanbul did not only raise
and sell slaves at high prices but also married those female slaves off to individuals
from their own networks. In this sense, this dissertation showed that this practice
of arranging marriages indeed provided prestige and privilege to the slave owner, as
well as adding value to the slave.
This study additionally dealt with the relationship between some Istanbulite elites
and their slave children of Caucasian and African origin. At first sight, it appeared
that child slaves were highly dependent on their elite owners; however, this thesis
proposed an asymmetrical dependency between them, in which some elite owners,
too, were reliant on their child slaves rather than a traditional binary master-slave
relationship. In this interdependent relationship, some elite owners had a critical
degree of authority over the enslaved children, which enabled them to make critical
decisions in the children’s lives. Due to the uneven distribution of authority, the
proposed relationship is not symmetrical.
While studying the children of Caucasian and African origin in the households of
some elite families in late Ottoman Istanbul, this preliminary research benefited
from the Ottoman archives in Istanbul, the travelogues written by the visitors to
the Empire at that time, and the memoirs penned by Ottoman individuals. This
thesis did not give a voice to the voiceless; not a single picture of a slave has been
unearthed, as no memoir written by a slave or a former slave has been discovered
yet. In the future, there could be archival documents in different archives other than
those in Istanbul, perhaps in another language, or a memoir that may be discovered
to hear the individual voices of slave children and to better understand what they
endured from their own point of view.
82
BIBLIOGRAPHY
Primary Sources
Archival Sources
The Directorate of State Archives of the Presidency of the Republic of Turkey, Istanbul,
BOA):
BOA.IE.AS 26-2315 (22 Receb 1094) (17 JULY 1683)
BOA. C..ADL.9 - 587 (25 Receb 1134) (1 MAY 1722)
BOA. TS.MA.e 918-71. (30 Recep 1200) (29 MAY 1786)
BOA. A.}MKT.MVL. 74- 59 (07 Zilhicce 1271) (21 AUGUST 1855)
BOA, A.}MKT.DV.. 5 – 96. (06 Zilkade 1263) (21 SEPTEMBER 1842)
BOA, HR.MKT.161- 41 (30 Muharrem 1273) (30 SEPTEMBER 1856)
Published Primary Sources
Asçıdede Halil Ibrahim: Hatıralar (Istanbul: Istanbul Ansiklopedisi Kütüphanesi, 1960).
Âsıkpasazâde Tarihi (eds) Necdet Öztürk. (Istanbul: Bilge Kültür Sanat, 2013).
Charles White. Three Years in Constantinople, or Domestic Manners of the Turks in
1844. Second Edition, Vol II. (London: Henry Colburn Publisher, Great Marlborough
Street, 1846).
Emine Foat Tugay, Three Centuries: Family Chronicles of Turkey and Egypt. (London:
Oxford University Press, 1963).
Eremya Çelebi Kömürcüyan. Istanbul Tarihi – XVII. Asırda Istanbul. trans. by Hrand
D. Anreasyan, ed. by Kevork Pamukcuyan. (Istanbul: Eren, 1988). Evliyâ Çelebi
Seyahatnâmesi I. Kitap Topkapı Sarayı Kütüphanesi Bagdat 304 Numaralı Yazmanın
Transkripsiyonu – Dizini Ed. By Robert Dankoff - Seyit Ali Kahraman -
Yücel Daglı. (Istanbul: Yapı Kredi Yayınları, 1996).
Gelibolulu Mustafa ‘Âli, Mevâ’idü’n-Nefâis Fî-Kavâ’idi’l-Mecâlis, ed. Mehmed Seker.
(Ankara:Türk Tarih Kurumu,1997)
83
Hagop Mıntzuri, Istanbul Anıları, 1897-1940, trans. by Silva Kuyumcuyan, ed. by
Necdet Sakaoglu. (Istanbul: Tarih Vakfı Yurt Yayınları, 2008).
Halide Edip (Adıvar). Memoirs of Halide Edib. (London, John Murray Albemarle
Street, W., 1923).
Jean-Henri Abdolonyme Ubicini, Letters on Turkey: An Account of the Religious, Political,
Social, and Commercial Condition of the Ottoman Empire. Part I: Turkey and
the Turks. trans. by M.A Ubicini. (London: John Murrey Albemarle Street, 1856).
Julia Pardoe. The City Of The Sultan; And Domestic Manners Of The Turks, In 1836.
Vol.1, (London: Henry Colburn, Publisher, Great Marlborough Street, 1837).
Koçi Bey Risalesi (eds.) Seda Çakmakcıoglu. (Kabalcı Yayınevi, Istanbul, 2007)
Leila (Saz) Hanimefendi. The Imperial Harem of the Sultans: Memoirs of Leylâ (Saz)
Hanimefendi. (Istanbul: Peva Publications, 1994).
Lucy Mary Jane Garnett and Stuart-Glennie, John S., The Women of Turkey and Their
Folklore, Vol. I & II (London: David Nutt, 270-271, Strand, W.C, 1891).
Mehmed Nesrî, Kitâb-ı Cihannümâ, ed. Faik Resit Unat – Mehmed Köymen, vol.1
(Ankara: Türk Tarik Kurumu, 2014)
Melek Hanım. Thirty Years in the Harem: Or The Autobiography of Melek-Hanum,
Wife of H. H. Kibrizli-Mehemet-Pasha, Vol.1 (London: Chapman and Hall, 1872).
Oruc Beg, Tevârih-i Âl-i Osman, ed. by Necdet Öztürk. (Istanbul: Bilgi Kültür Sanat,
2014)
Secondary Sources
Abou El-Haj, Rifaat Ali. “The Ottoman Vezir and Pasha Households 1683 - 1703: A
Preliminary Report”, Journal of the American Oriental Society, 94/4, (1974), 438-47
Ágoston, Gábor and Bruce Alan Masters (eds.) Encyclopedia of the Ottoman Empire.
(New York, NY:Facts On File, 2009).
Araz, Yahya. 16. Yüzyıldan 19. Yüzyıl Baslarına: Osmanlı Toplumunda Çocuk Olmak,
(Istanbul: Kitap Yayınevi, 2017)
Argit, Betül Ipsirli. Life after the Harem: Female Palace Slaves, Patronage and the
Imperial Ottoman Court. Cambridge: Cambridge University Press, 2020.
Aydın, Mehmet Akif and Muhammed Hamidullah, ‘Köle’ Türkiye Diyanet Vakfı Islâm
Ansiklopedisi, (DIA), Ankara, 2002, pp. 237-246.
Aykan, Yavuz. (2017) “On Freedom, Kinship, And The Market Rethinking Property
And Law In The Ottoman Slave System”, Quaderni Storici, 52, 1, pp.13-39.
Bacharach, Jere L. “Bookreview: Race and Slavery in the Middle East: An Historical
Enquiry by Bernard Lewis” International Journal Of Middle East Studies 24,
No. 2 (1992): Pp.305-07.
84
Baer, Gabriel. “Slavery in Nineteenth Century Egypt.” The Journal of African History
8, no. 3 (1967): 417–41.
Barkan, Ömer Lütfi. “XV ve XVI. Asırlarda, Osmanlı Imparatorlugu’nda Toprak Isçiliginin
Organizasyonu Sekilleri, I, Kulluklar ve Ortakçı Kullar”, Iktisat Fakültesi Mecmuası
I, No: 1, Istanbul 1939.
Barkan, Ömer Lütfi. XV-XVI. Asırlarda Osmanlı Imparatorlugunda Zirai Ekonominin
Hukukî ve Malî Esasları, Kanunlar I, Istanbul 1943.
Barkan, Ömer Lütfi; “Türkiye’de “Servaj” Varmı idi.”, Belleten XX/78, Ankara 1956.
pp. 237-246.
Bekir, Onur. Türkiye’de Çocuklugun Tarihi: Çocuklugun Sosyo-Kültürel Tarihine Giris,
(Ankara: Imge, 2005).
Ben-Naeh, Y. Blond, tall, with honey-colored eyes: Jewish ownership of slaves in the
Ottoman Empire. Jew History 20, 315–332 (2006).
Blagbrough, Jonathan, and Gary Craig. “‘When I Play with the Master’s Children, I
Must Always Let Them Win’: Child Domestic Labor.” Chapter. In Child Slavery
before and after Emancipation: An Argument for Child-Centered Slavery Studies,
edited by Anna Mae Duane, 251–69. Slaveries since Emancipation. (Cambridge:
Cambridge University Press, 2017.)
Braude, Benjamin and Bernard Lewis (ed.) Christians and Jews in the Ottoman Empire.
2 vols. (New York and London: Holmes and Meier, inc., 1982)
Brunschvig, R. "’Abd," Encyclopedia Of Islam, New Edition (Ep) Vol. I (Leiden, 1960).
Cahen Claude, “Note sur l’esclavage musulman et le devshirrme Ottoman, à propos de
travaux récents”, JESHO, XIII/2 (1970), 211-218.
Campbell, Gwyn, Suzanne Miers and Joseph C. Miller (eds), Children in Slavery through
the Ages. (Athens OH: Ohio University Press, 2009.)
Canbakal, Hülya and Filiztekin, Alpay (2020) "Slavery and decline of slave-ownership
in Ottoman Bursa, 1460-1880", International Journal of Labour and Working Class
History, Vol.97, 57-80.
Crone Patricia. Slaves on Horses: The Evolution of the Islamic Polity. (Cambridge:
Cambridge University Press, 1980)
Davis, David Brion. Slavery and Human Progress. (New York: Oxford University Press,
1984)
Diptee, Audra A. “Notions of African Childhood in Abolitionist Discourses: Colonial
and Postcolonial Humanitarianism in the Fight Against Child Slavery.” pp. 208–30
in Child Slavery before and after Emancipation: An Argument for Child-Centered
Slavery Studies. Slaveries since Emancipation. ed. by Anne Duane. (Cambridge:
Cambridge University Press, 2017)
Drescher, Seymour. Abolition: A History of Slavery and Antislavery. (Cambridge:
Cambridge University Press, 2009.)
85
Duane, Anna Mae, ed. Child Slavery before and after Emancipation: An Argument
for Child-Centered Slavery Studies. Slaveries since Emancipation. (Cambridge:
Cambridge University Press, 2017)
Duindam, Jeroen, Tülay Artan, and Metin Kunt. Royal Courts in Dynastic States and
Empires: A Global Perspective. Leiden, The Netherlands: Brill.2011.
Erdem, Yusuf Hakan, “Kırım Savasında Karadeniz Beyaz Köle Ticareti”, Savastan
Barısa: 150.Yıldönümünde Kırım Savası ve Paris Antlasması (1853-1856), Istanbul:
Istanbul Üniversitesi Edebiyat Fakültesi Tarih Arastırma Merkezi, May 2007,
pp. 85-118.
Erdem, Yusuf Hakan, “Tanzimat Döneminde Kölelik, 1846-1876”, Tanzimat: Degisim
Sürecinde Osmanlı Imparatorlugu, Inalcık, Halil and Seyitdanlıoglu, Mehmet (eds.),
(Istanbul: Türkiye Is Bankası, 2011)
Erdem, Yusuf Hakan. (2010). “Magic, Theft, And Arson: The Life And Death Of
An Enslaved African Woman in Ottoman Izmit” In T. Walz And M. Cuno (Eds.),
Race And Slavery in The Middle East: Histories Of Trans-Saharan Africans in
Nineteenth-Century Egypt, Sudan, And The Ottoman Mediterranean, pp. 125-146.
Cairo: The American University Press.
Erdem, Yusuf Hakan. ‘Do not think of the Greeks as agricultural labourers’: Ottoman
responses to the Greek war of independence. In: Birtek, Faruk and Dragonas,
Thalia, (eds.) Citizenship and The Nation-State in Greece and Turkey. Social and
Historical Studies on Greece and Turkey Series. (Routledge, London 2005) pp. 67-
84.
Erdem, Yusuf Hakan. Slavery in the Ottoman Empire and Its Demise 1800-1900, (New
York: St. Martin’s Press, 1996).
Faroqhi, S. “Black Slaves and Freedmen Celebrating (Aydın, 1576),” Turcica 21 (1991),
pp. 206-215.
Faroqhi, Suraiya. (1994), “The Ruling Elite Between Politics And The Economy.” in
H. Inalcik and D. Quataert (eds), An Economic and Social History of the Ottoman
Empire 1300–1914, pp. 545–574. Cambridge.
Faroqhi, Suraiya. “Quis Custodiet Custodes? Controlling Slave Identities and Slave
Traders in Seventeenth- and Eighteenth-Century Istanbul,” in Faroqhi, Stories of
Ottoman Men and Women: Establishing Status, Establishing Control (Istanbul,
2002).
Faroqhi, Suraiya. Stories of Ottoman Men And Women. (Istanbul: Eren Publications,
2002)
Findley, Carter V. Bureaucratic Reform in the Ottoman Empire: The Sublime Porte,
1789-1922. (Princeton:Princeton University Press, 1980)
Finley, M.I. Ancient Slavery And Modern Ideology. New York: Viking Press, 1980.
Fisher, Alan (1980). “Chattel Slavery in The Ottoman Empire.” Slavery And Abolition,
I (1),pp.25-45.
86
Fisher, Alan (1980). “Studies in Ottoman Slavery and Slave Trade, II: Manumission.”
Journal Of Turkish Studies, 4, pp.49-56.
Fisher, Alan W. ‘Chattel slavery in the Ottoman Empire’ in Slavery and Abolition, 5:1,
(Taylor & Francis Group, 1980)
Fisher, Alan. “ The Sale of Slaves in the Ottoman Empire: Markets and State Taxes,
Slave Sales, Some Preliminary Considerations,” Bogaziçi Üniversitesi Dergisi, Beseri
Bilimleri VI (1978: 149-179)
Fortna, Benjamin (2010) ‘Reading between Public and Private in the Late Ottoman
Empire and Early Turkish Republic.’ Comparative Studies of South Asia, Africa
and the Middle East, 30 (3). pp. 563-573.
Fortna, Benjamin C., Learning to Read in the Late Ottoman Empire and the Early
Turkish Republic, Basingstoke: Palgrave Macmillan, 2010.
Frenkel, Yehoshua. “Slavery in 17th-Century Ottoman Jerusalem in Light Of Several
Sharia Court Records” in Slaves and Slave Agency in the Ottoman Empire (ed) by
Stephan Conermann and Gül Sen. Bonn University Press, 2020.
Fynn-Paul, Jeffrey. (2009). “Empire, Monotheism and Slavery In The Greater Mediterranean
Region From Antiquity To The Early Modern Era.” Past & Present. 205.
pp. 3-40.
Genç, Mehmet. Osmanlı Imparatorlugunda Devlet ve Ekonomi, (Istanbul: Ötüken
Nesriyat, 2002).
Genç, Mehmet. ‘Malikane’ Türkiye Diyanet Vakfı Islam Ansiklopedisi, Vol 27, Ankara,
2003, pp. 516-518.
Göçek, Fatma Müge. Rise of the Bourgeoisie, Demise of Empire; Ottoman Westernization
and Social Change. (Oxford University Press, NY, 1996).
Hanna, Nelly. “Sources For The Study Of Slave Women And Concubines in Ottoman
Egypt,”In: Sonbol, Amira E. (ed.). Beyond the Exotic: Women’s Histories in Islamic
Societies. Syracuse, New York: Syracuse University Press, 2005. pp. 119-129.
Hathaway, Jane. The Politics of Households in Ottoman Egypt: The Rise of the Qazdaglıs.
Cambridge: Cambridge University Press, 2002.
Ilıcak, Hüseyin Sükrü. “A Radical Rethinking of Empire: Ottoman State and Society
during the Greek War of Independence (1821–1826).” PhD diss., Harvard University,
2011.
Inalcık, Halil (1979). “Servile Labor in The Ottoman Empire” in A. Ascher, T. Halasi
Kuhn, And B. K. Kiraly (Eds.), The Mutual Effects Of The Islam And The
Judaeo-Christian Worlds: The Eastern European Patterns (pp. 25-53). New York:
Columbia University Press.
Inalcık, Halil. “Timar,” The Encyclopaedia of Islam , Ed . , P. J. Bearman , Th .
Bianquis, C. E. Bosworth , E. Van Donzel , W. P. Heinrichs , X , (Leiden: Brill,
2000): pp. 501 – 507.
87
Inalcık, Halil. The Ottoman Empire The Classical Age 1300-1600. (Phoenix Press, GB,
1973).
Ipsirli-Argıt, Betül (2009). “Manumitted Female Slaves Of The Ottoman Imperial
Harem (Sarayîs) in Eighteenth- Century Istanbul.” (Unpublished Doctoral Dissertation).
Bogaziçi University, Istanbul.
Jennings, R. C. (1987). “Black Slaves and Free Blacks in Ottoman Cyprus, 1590-1640.”
Journal of the Economic and Social History of the Orient, 30 (3), pp. 286-302.
Karamürsel, Ceyda. “In The Age Of Freedom, In The Name Of Justice”: Slaves, Slaveholders,
And The State In The Late Ottoman Empire And Early Turkish Republic,
1857-1933.” (Unpublished Ph.D. Dissertation, University of Pennsylvania, 2015)
King, Wilma. Stolen Childhood: Slave Youth in Nineteenth-Century America. (Bloomington
and Indianapolis: Indiana University Press, 1995.)
Kunt, Metin. “Transormation of zimmî into askerî” in Christians and Jews in the Ottoman
Empire. Benjamin Braude and Bernard Lewis (ed.): 2 vols, I: The Central
Lands. (New York and London: Holmes and Meier, inc., 1982)
Lavina, Javier and Michael Zeuske. The Second Slavery: Mass Slaveries and Modernity
in the Americas and in the Atlantic Basin. (LIT Verlag Münster, 2014).
Lewis, Bernard. Orta Doguda Irk Kavramı ve Kölelik. tran. Özgür Balkılıç. (Akılçelen
Kitaplar, 2017)
Lewis, Bernard. Race And Color In Islam (Harper Torchbooks. Harper And Row,
Publishers, New York, Evanston, San Francisco, London, 1971)
Lewis, Bernard. Race and Slavery in the Middle East (Oxford, 1990)
Lewis, Bernard. Race Et Couleur En Pays d’Islam (Paris: Payot, 1982)
Lewis, Reina. Rethinking Orientalism: Women, Travel and the Ottoman Harem. Rutgers
University Press, 2004.
Maksudyan, Nazan. Orphans And Destitute Children in The Late Ottoman Empire.
(Syracuse, Ny: Syracuse University Press, 2014.)
Melman, Billie. Women’s Orients: English Women and the Middle East, 1718–1918:
Sexuality, Religion and Work. United Kingdom: Palgrave Macmillan UK, 2016.
Ménage, V.L. (1966) “Some Notes on the Devshirme”, Bulletin of the School of Oriental
and African Studies, University of London, 29(1), pp. 64-78.
Ménage, V.L. “Devshirme”, Encylopedia of Islam. Vol. II, E. J. (Brill, Leiden 1991.)
pp.210-213.
Ménage, V.L., “Devshirme”, IE.2, vol.II. (Leiden: Brill 2007) Montana, Ismael. The
Abolition Of Slavery in Ottoman Tunisia, (University Press Of Florida, 2013).
Morony, Michael G. “Bookreview: Race And Slavery in the Middle East: An Historical
Enquiry by Bernard Lewis” Journal Of Near Eastern Studies 54, No. 1 (1995):
pp.69-71.
88
Oguz, Ahmet. (1994). “Ceride-i Havadis Gazetesinde çıkan haber yazılar ve degerlendirmesi.
MA Thesis.”
Özbay, Ferhunde. Turkish Female Child Labor in Domestic Work: past and present.
Project Report submitted to ILO/IPEC. Bogaziçi University Press: Istanbul. 1999.
Özkoray, Hayri Göksin. "From Persecution to (Potential) Emancipation", Hawwa17, 2-3
(2019): 257-280.
Pakalın, Mehmet Zeki. Osmanlı Tarih Deyimleri ve Terimleri Sözlügü, (Istanbul:
MEB,1983.)
Palmer, J.A.B. “The Origin of the Janissaries”, Bulletin of the John Rylands Library,
XXXV (1953): 448-
Pargas, Damian Alan. (2011) “From the Cradle to the Fields: Slave Childcare and
Childhood in the Antebellum South”, Slavery & Abolition, 32:4, 477-493.
Patterson, Orlando. Slavery and Social Death: A Comparative Study. (Cambridge,
Mass: Harvard University Press, 1982.)
Peirce, Leslie. The Imperial Harem: Women And Sovereignty In The Ottoman Empire.
(Oxford University Press 1993).
Peirce, Leslie. Morality Tales: Law and Gender in the Ottoman Court of Aintab. (Berkeley:
University of California Press, 2003.)
Repp, R.C. “A Further Note on the Devshirme”, BSOAS, XXXI (1968): 137-9.
Richardson, Kristina. "Singing Slave Girls (qiyan) of the Abbasid Court" in Children in
Slavery Through the Ages ed. by Campbell, Gwyn, Suzanne Miers, and Joseph C.
Miller, Ohio University Press, 2009, pp. 105-118.
Rosen, David M. Armies of the Young: Child Soldiers in War and Terrorism.
(Brunswick, NJ: Rutgers University Press, 2005)
Rossi, Benedetta. “Gwyn Campbell, Suzanne Miers and Joseph C. Miller (eds), Children
in Slavery through the Ages. (Athens OH: Ohio University Press, 2009.) Africa 80,
no. 3 (2010): pp. 516–18.
Sahillioglu, Halil. (1985). “Slaves in The Social And Economic Life Of Bursa in The
Late 15th And Early 16th Centuries.” Turcica 17, pp.43–112.
Sahillioglu, Halil. “On Besinci Yüzyıl Sonunda Bursa’da Is Ve Sanayi Hayatı Kölelikten
Patronluga”, Ömer Lütfi Barkan Armaganı, Paris 1980.
Sak, Izzet. (1989). “Konya’da Köleler” The Journal Of Ottoman Studies, IX, 159-197.
Seng, Yvonne J. "Fugitives and Factotums: Slaves in Early Sixteenth-Century Istanbul."
Journal of the Economic and Social History of the Orient 39, no. 2 (1996): 136-69.
Sobers-Khan, Nur. Slaves Without Shackles: Forced Labour and Manumission in the
Galata Court Registers, 1560-1572. (Berlin, Boston: De Gruyter, 2020).
89
Somel, Selçuk Aksin. Historical Dictionary of the Ottoman Empire. (Lanham, Maryland,
and Oxford: The Scarecrow Press, Inc. 2003).
Spivak, Gayatri Chakravorty. “Can The Subaltern Speak?” In Marxism And The Interpretation
Of Culture, Ed. Cary Nelson, And Lawrence Grossberg. (Urbana:
University Of Illinois Press, 1988) pp. 271–313.
Spyropoulos,Yannis. ‘Slaves and freedmen in 17th- and early 18th-century Ottoman
Crete,’ Turcica, 46, 2015, pp. 177-204.
Tezcan, Baki. The Second Ottoman Empire: Political and Social Transformation in the
Early Modern World. (Cambridge: Cambridge University Press, 2010)
Toledano, Ehud R. (1981) “Slave Dealers, Women, Pregnancy, And Abortion: The Story
of A Circassian Slave-Girl In Mid-Nineteenth Century Cairo,” Slavery & Abolition
2 (1), pp.53-68.
Toledano, Ehud R. (2002). Representing the slave’s body in Ottoman society. Slavery
& Abolition, 23 (2), pp. 57–74.
Toledano, Ehud R. “Enslavement in the Ottoman Empire in the Early Modern Period.”
in The Cambridge World History of Slavery. (eds) David Eltis and Stanley L.
Engerman, 3:25–46. (Cambridge: Cambridge University Press, 2011.)
Toledano, Ehud R. “Late Ottoman Concepts of Slavery (1830s-1880s).” Poetics Today
14, no. 3 (1993): pp. 477–506.
Toledano, Ehud R. “Ottoman and Islamic Societies: Were They ‘Slave Societies’?” in
What Is a Slave Society?: The Practice of Slavery in Global Perspective. (eds) Noel
Lenski and Catherine M. Cameron, pp. 360–82. (Cambridge: Cambridge University
Press, 2018)
Toledano, Ehud R. “Shemsigul: A Circassian Slave in Mid-Nineteenthcentury Cairo.”
In Struggle And Survival in The Modern Middle East, Ed. Edmund Burke Iıı, And
David Yaghoubian. (Berkeley, California: University Of California Press, 1993) pp.
59–74
Toledano, Ehud R. 2011. “Coping with Trauma in a NewWorld: Cultural Responses of
Enslaved Africans in the Ottoman Empire.” In African Communities in Asia and the
Mediterranean: Identities Between Integration and Conflict, ed. Ehud R Toledano.
Trenton, NJ: Africa World Press pp. 167–194.
Toledano, Ehud R. 2017. “Ottoman Elite Enslavement And ’Social Death’.” In On
Human Bondage: After Slavery And Social Death, Ed. John P Bodel, And Walter
Scheidel. Chichester, West Sussex: John Wiley Sons pp.136–150.
Toledano, Ehud R. Insight Turkey. Vol. 13, No. 3, Ak Party’s Electoral Hegemony The
Making Of A “New Middle East” Turkey’s Return To The Balkans (Summer 2011),
pp. 208-211.
Toledano, Ehud R. The Ottoman Slave Trade And Its Suppression, 1840- 1890, (Princeton:
Princeton University Press, 1982.)
90
Troutt-Powell, Eve M. (2006) “Will That Subaltern Ever Speak? Finding African Slaves
in The Historiography Of The Middle East.” in I. Gershoni, A. Singer And Y.
H. Erdem (Eds.), Middle East Historiographies: Narrating The Twentieth Century
(pp.242-260). Seattle: University Of Washington Press.
Troutt-Powell, Eve M. A Different Shade Of Colonialism: Egypt, Great Britain, And
The Mastery Of The Sudan (Berkeley: University Of California Press, 2003)
Troutt-Powell, Eve M. Tell This in My Memory: Stories of Enslavement from Egypt,
Sudan and the Ottoman Empire. (Stanford, CA: Stanford University Press, 2012.)
Uzunçarsılı, Ismail Hakkı. Kapıkulu Ocakları, Vol.1 (Ankara: Türk Tarih Kurumu, 1988)
Uzunçarsılı, Ismail Hakkı. Osmanlı Devletinin Merkez ve Bahriye Teskilatı, (Ankara:
TTK Basımevi, 2009).
Vryonis, Speros. “Isidore Glabas and the Turkish Devshirme”, Speculum, XXXI (1956),
pp. 433-443.
Wagner, Veruschka. (2020). “Speaking Property” With The Capacity To Act: Slave
Interagency In The 16th- And 17th-Century Istanbul Court Registers” in Slaves
and Slave Agency in the Ottoman Empire (ed) by Stephan Conermann and Gül
Sen. Bonn University Press, 2020.
Walz,Terence and Kenneth M. Cuno (eds): Race and Slavery in the Middle East: Histories
of Trans-Saharan Africans in Nineteenth-Century Egypt, Sudan, and the Ottoman
Mediterranean. (Cairo and New York: The American University in Cairo
Press, 2010)
Wilkins, Charles L. “Slavery and Household Formation in Ottoman Aleppo, 1640-1700”,
Journal of the Economic and Social History of the Orient 56, 3 (2013): 345-391.
Winter, Sarah. “The Slave Child as ‘Gift’: Involutions of Proprietary and Familial Relations
in the Slaveholding Household before Emancipation.” pp. 50–74 in Child
Slavery before and after Emancipation: An Argument for Child-Centered Slavery
Studies. Slaveries since Emancipation. ed. by Anne Duane. (Cambridge: Cambridge
University Press, 2017)
Wittek, Paul. “Devshirme and Shar¯ıa”,BSOAS, XVII (1955), s. 271-278.
Yasa, Fırat. “Between Life and Death: Slaves and Violence in Crimean Society in The
Last Quarter Of 17th Century”, Selçuk University Journal Of Studies Turcology, Vol
47 (2019), pp. 433-443.
Yasa, Fırat. “Desperation, Hopelessness, And Suicide: An Initial Consideration of Self-
Murder by Slaves in Seventeenth-Century Crimean Society”, Turkish Historical Review,
Vol. 9 (2018): pp. 198-211.
Yılmaz, Gülay. “The Devshirme System and the Levied Children of Bursa in 1603-4”,
BELLETEN, vol. 79, no. 286, pp. 901-930, Dec. 2015.
Yılmaz, Gülay. (2009). ‘Becoming a devsirme: the training of conscripted children in the
Ottoman Empire’ in Children in Slavery Through the Ages ed. by Campbell, Gwyn,
91
Suzanne Miers, and Joseph C. Miller, Ohio University Press, 2009, pp. 119-135.
Ze’evi, Dror. “Women in 17th-Century Jerusalem: Western and Indigenous Perspectives,”
International Journal of Middle East Studies 27/2 (1995): 157–73.
Zilfi, Madeline C. “Thoughts on Women and Slavery in the Ottoman Era and Historical
Sources.” In Beyond the Exotic: Women’s Histories in Islamic Societies, ed. Amira
El- Azhary Sonbol, pp.131–8. (Syracuse, NY: Syracuse University Press, 2005)
Zilfi, Madeline. “Servants, Slaves, And The Domestic Order In The Ottoman Middle
East,” HAWWA 2, 1 (2004): pp. 1-33.
Zilfi, Madeline. Osmanlı Imparatorlugunda Kadınlar ve Kölelik. trans. Ebru Kılıç.
(Istanbul: Is Bankası, 2018.)
Zilfi, Madeline. Women and Slavery in the Ottoman Empire, (Cambridge: Cambridge
University Press, 2010).
92