Back to top

The Concubine, the Princess, and the Teacher

The Concubine, the Princess, and the Teacher
Voices from the Ottoman Harem

Three women who lived in the Ottoman imperial harem between 1876 and 1924 describe the lifeways of the imperial family, dispelling Western stereotypes of harem debauchery.

November 2008
This is a print-on-demand title. Expedited shipping is not available.
Add to cart
324 pages | 6 x 9 | 28 b&w illus., 1 map |

In the Western imagination, the Middle Eastern harem was a place of sex, debauchery, slavery, miscegenation, power, riches, and sheer abandon. But for the women and children who actually inhabited this realm of the imperial palace, the reality was vastly different. In this collection of translated memoirs, three women who lived in the Ottoman imperial harem in Istanbul between 1876 and 1924 offer a fascinating glimpse "behind the veil" into the lives of Muslim palace women of the late nineteenth and early twentieth centuries.

The memoirists are Filizten, concubine to Sultan Murad V; Princess Ayse, daughter of Sultan Abdulhamid II; and Safiye, a schoolteacher who instructed the grandchildren and harem ladies of Sultan Mehmed V. Their recollections of the Ottoman harem reveal the rigid protocol and hierarchy that governed the lives of the imperial family and concubines, as well as the hundreds of slave women and black eunuchs in service to them. The memoirists show that, far from being a place of debauchery, the harem was a family home in which polite and refined behavior prevailed. Douglas Brookes explains the social structure of the nineteenth-century Ottoman palace harem in his introduction.

These three memoirs, written across a half century and by women of differing social classes, offer a fuller and richer portrait of the Ottoman imperial harem than has ever before been available in English.

  • Acknowledgments
  • Introduction
  • Part One. The Concubine Filizten
  • Part Two. The Princess Ayse
  • Part Three. The Teacher Safiye
  • Conclusion
  • Notes
  • Glossary of Names
  • Glossary of Terms and Places
  • Bibliography
  • Index

Douglas Scott Brookes holds a Ph.D. in Near Eastern Studies from the University of California, Berkeley, and teaches courses in Ottoman and Middle Eastern history and culture. He lives in Oakland, California.


[UTP note: Some diacritics have been omitted from this online excerpt but are present in the book.]


The imperial harem of the Ottoman Sultans has long fascinated outsiders as a mélange of sex, debauchery, slavery, power, riches, and sheer abandon—in short, the incarnation of the most attractive vices. Concealed behind its own veil of circumspection, the imperial harem formed an object of mystery even in Ottoman culture, as decorum demanded respect for the privacy of the institution whose name means "forbidden, prohibited, sacred" in its Arabic original.


Penetrating behind that veil, and beyond the image of the harem in the public's fancy, this book reveals everyday life in the Ottoman imperial harem through the memoirs of three women whose home it was between 1876 and the dispersal of the harem in 1924. The memoirs are those of the concubine Filizten in Ziya Sakir's biography Çiragan Sarayinda 28 Sene: Besinci Murad'in Hayati [Twenty-Eight Years in Çiragan Palace: The Life of Murad V]; Princess Ayse Osmanoglu in her reminiscences, Babam Sultan Abdülhamid [My Father, Sultan Abdülhamid]; and schoolteacher Safiye Ünüvar in her record of life in palace employment, Saray Hatiralarim [My Palace Memories]. These works were selected for this translation because of the rich detail each provides, but also because of the divergent positions in the harem the authors possessed, which allows us to construct a portrait of the harem from diverse yet complementary sources.


The importance of the memoirs lies in their revelations of life in an institution still largely misunderstood today. For the veil that insulated the imperial harem during its existence has survived nearly intact, despite the death of the harem six years after World War One. Exploring the origins and persistently robust health of the harem myths lies outside the purview of this book, but surely an explanation of their persistence must cite the dearth of reliable accounts of life within the harem. For of Ottoman women familiar with the institution, whether as residents or visitors, not one would have considered writing of her life as long as a Sultan occupied the Ottoman throne. As a result, the few accounts of harem life by Ottoman females who knew it well appeared only after the fall of the empire and the concomitant destruction of the social system in which the harem thrived.


Then, of the exceedingly few foreign female visitors to the harem who wrote of their visits, to our knowledge none spoke Turkish or witnessed harem life for more than a short social call. This left them poorly positioned to reliably understand and interpret what they witnessed. For a similar reason, the accounts we have of the harem by male authors—all foreigners—may safely be treated with skepticism, or even discounted outright, since Ottoman court practice barred foreign men from entering the harem at all (with the possible exception of physicians). Their accounts are second-hand at best, and prone to fancy.


On quite the other hand, the memoirs translated here issue from ladies intimately familiar with the harem because they resided within it. Presented chronologically, each translation is an excerpt of the sections in the memoirs that directly recount life in the harem or that describe inhabitants of the palace. In order to provide as much detail of the harem as possible from all three ladies, yet produce a book manageable in size, the translations omit sections of the memoirs that do not directly concern harem life, primarily personal opinions of political events or statesmen of the day and events that transpired after each author left the imperial harem.


The Structure of the Ottoman Palace Harem of the Late Nineteenth and Early Twentieth Centuries


Although any discussion of the Ottoman palace harem must acknowledge that our understanding of the institution is limited by the relative paucity of studies of it, nevertheless certain conclusions can be drawn based on scholarly work to date as well as on the memoirs of our three ladies. To be sure, the brief analysis here applies to the imperial harem only in the last hundred or so years of its existence, as an institution such as the harem that endured for centuries would certainly have evolved a very great deal indeed over the long course of its existence.


Seemingly authenticating the popular fantasy of harem life is that, in theory, the Ottoman Sultan might choose any number of concubines from among the slave women in imperial service, the only restriction being the canonical injunction that he could name but four of them as his official wives. But in practice the Prince's sexuality was restrained by the fact that each maiden he took to bed had to be provided a rank, her own suite of rooms and servants, and the right to advance in the concubine hierarchy. In other words, she had claims on the Prince that he must honor.


As a result of these restraints, as nearly as we can determine for the monarchs of the period of these memoirs, Sultan Murad V had sexual relations with nine ladies, Abdülhamid II with thirteen ladies, and Resad and Vahideddin with five ladies each—more than the vast majority of their subjects, to be sure, who took one wife, but not the stuff of legend. From their multiple ladies, Murad sired seven children, Abdülhamid seventeen, Resad and Vahideddin four each.


In consequence of the practical restraints on the number of concubines the Sultan might take, combined with extremely high infant mortality rates, and the custom since the early seventeenth century of limiting contenders to the throne by forbidding a Prince to father children until he became Sultan, by November 1808 the Ottoman dynasty had dwindled to but one male, Sultan Mahmud II. And so the imperial family most assuredly did not number in the thousands by 1876, when our memoirs begin.


The restriction among male members of the dynasty that allowed only the monarch to sire children began to give way in the last years of the reign of Sultan Abdülmecid. In 1857 his brother and heir Prince Abdülaziz fathered a son, although the child's existence was kept secret until Abdülaziz ascended the throne in 1861.


The Entourage System


Given the presence of multiple concubines, royal children, and slaves resident in it, quite understandably court practice divided the Ottoman harem into entourages, or households.


The Sultan oversaw the largest entourage. His serving women and eunuchs ranked as the most revered in the palace. Then, each adult Prince and Princess oversaw his or her own entourage, as did each concubine and the highest-ranking officeholders in the harem. Special harem residents might receive their own small entourage—for example, our schoolteacher. Each entourage occupied its own set of apartments, physically separate from the other entourages. Each possessed its own cadre of female slaves and eunuchs in service to it.


If she were living, the Sultan's mother constituted the chief authority over the harem, so that only the more senior and experienced slave women served in her entourage. If she were deceased, authority over the harem devolved to the Sultan's Senior Consort, a position appointed by the monarch for life.


The entourage system flourished in miniature in the villas of the imperial family throughout Istanbul. Married Princesses had always lived in their own villas with their husbands. As an innovation of the nineteenth century, in imitation of European practice, adult Princes also received permission to move out of the palace to their own villas around the city.


The Imperial Family: Members and Adjuncts


In the Ottoman monarchy of our period, succession to the throne passed not to the monarch's eldest son but to the eldest male member of the dynasty. In order to limit claims to the throne, and to clarify rank and precedence, the House of Osman defined a member of the dynasty as the son or daughter of a Sultan or of a Prince. In other words, descent through males alone imparted royal rank. For this reason, children of Princesses were not considered members of the dynasty but rather mensuplar, adjuncts of it. A Princess's grandchildren were commoners, albeit prestigious ones.


Concubines too, even the mother of the Sultan, were held to be but adjuncts of the dynasty, not members of it. In this way the imperial family distinguished itself from these slave women in service to it, even though these slaves provided the family with its future members. Thus a concubine's children outranked her. However, from the memoirs it is clear that, rather than slighted, these women felt vastly honored to attain the distinction of adjunct to the imperial house. The distinction worked in their favor in 1924, when the law exiling members of the imperial family excluded the adjuncts. The son or daughter had to leave, but the mother could stay, and several did, including the mother of our Princess Ayse.


Concubines consisted of three ranks. The highest level belonged to the four consorts of the Sultan, ranked in precedence as Senior Consort, Second Consort, Third Consort, and Fourth Consort. The ambiguous term consort rather than wife is deliberate here to translate their title of Kadin (Lady), for it is unclear whether an Ottoman Prince underwent a wedding ceremony with a woman he took as a concubine or, if he acceded to the throne, whom he elevated to the rank of Kadin. The subject of weddings with imperial concubines requires further study, but given the present state of our knowledge we can say that until the mid-nineteenth century, such marriages formed the exceedingly rare exception in the House of Osman. For concubines were slaves, and Islamic religious law does not require a master to marry his slave in order to have sexual relations with her. Princess Ayse's story that her grandfather Sultan Abdülmecid promised his aunt he would actually wed her servant girl Perestu if she presented the girl to him bears witness to this custom.


After the mid-nineteenth century, however, marriages with concubines became more frequent as part of the Ottoman monarchy's increasing adoption of European royal practice, and Princess Ayse tells us that her father married her mother. Still, marriages that did occur remained brief and discreet events inside the harem, an expression of esteem on the part of the royal male. Not affairs of state, they went unreported in the press, unlike the lavish weddings of Princesses, which very much served the state as a means to honor high-ranking men of the government and military by linking their families in loyalty to the imperial house.


Below the four consorts ranked the two levels of concubines who were in theory not limited in number. The higher-ranking bore the title Ikbal (fortunate one), and like the consorts were ordered by seniority as "Senior Ikbal, Second Ikbal," etc. The lower-ranking bore the title Gözde (chosen one), but were not ordered by seniority.


The Female Slaves


Insofar as we can tell, nearly all the female slaves in the imperial harem in the era of our memoirs were ethnically Circassian.


Over the course of the nineteenth century the Russians had largely expelled the Circassian peoples, Muslims, from their native Caucasus. Most fled into the Ottoman Empire, at the time an impoverished society hardly in a position to absorb the large numbers of refugees streaming into it. One way in which uprooted Circassian families could realize a large sum of money, however, was to sell their daughters into slavery in the harems of the Ottoman elite. For Circassian girls, especially if blonde, blue-eyed, and fair of complexion, were considered highly desirable beauties. The fact that Islamic law forbade the enslavement of fellow Muslims was overlooked, given the economic and social advantages the trade offered for seller, buyer, and slave. Even after the imperial decree of 1854 that abolished slavery, the government turned a blind eye to the continued market in slaves, both for the revenue it generated and on the general principle that an institution as ingrained in Ottoman elite society as slavery could not realistically be banished overnight.


And so the trade in females went on, in both white Caucasian and black African slaves destined for the mansions of the elite and the palaces and villas of the imperial family. While the black female slaves in general entered into manual labor, the Circassian girls were preferred for more genteel service in the harems and as potential sources of concubines for the male members of the family. Indeed, purchasing young Circassian girls, teaching them refined manners and accomplishments, and then reselling them to wealthy families or the imperial court when they reached adulthood, constituted an honored profession for a select number of women well placed for access to the upper reaches of Ottoman society.


Though the traffic in humans continued, the Ottoman slave trade began to collapse in the era of our memoirs. Reflecting prevailing tastes worldwide, Ottoman society had come to regard the old institution with disfavor, but the decisive factor in the decline of slavery was economic. As Circassian refugee families assimilated into society, the supply of their daughters diminished, causing steep price increases that few families in this chronically impoverished economy could afford. By the end of the era we are considering, 1924, practically the only household still engaging slave labor was the imperial palace itself.


Once in the palace, a novice slave girl received a new name from the Sultan, by tradition a name Persian in origin that might describe her looks or personality. As the concubine tells us, a novice girl underwent training under the supervision of her elders in service until she had mastered the Turkish language, the standards of comportment, and the skills of the job. Her training complete, she received assigned duties within her entourage. After years of service, a deserving novice could be elevated to the rank of kalfa (lit., master assistant; overseer). A prettier kalfa might be chosen as concubine to a Prince. Even were she not so selected, however, if the Prince she served succeeded to the throne, a kalfa could still hope to attain senior status in the harem through appointment by the monarch to the supervisory rank of hazinedar.


An unknown number of the senior female and male slaves in palace service were manumitted, as the schoolteacher mentions, but chose to remain in palace service. Nor were slave women who were unhappy in service forced to remain in the palace until the end of their days, for as both our Princess and the schoolteacher tell us, they possessed the right to ask to leave service for marriage to a man outside the palace. In the era of our memoirs this request appears to have been always granted.


The Eunuchs


We do not know a good deal about the origin of the black African eunuchs at the Ottoman court. Captured by slave raiders, they were castrated shortly afterward by removing, it would appear, both penis and testicles. For the few who survived this ordeal, the physical repercussions of their trauma included testosterone deficiency that resulted in misshapen bodies, high-pitched voices, and lack of a beard. The emotional repercussions await in-depth study, but at the time were considered to include eccentricity, meanness, and the potential desire to avenge oneself on society. Our memoirs include examples of this behavior, but also of kindness, trustworthiness, and loyalty on the part of the eunuchs in palace service.


Conveyed from Africa to the slave markets in Egypt, Arabia, and Yemen, the newly created eunuchs were purchased primarily by dealers from Istanbul. While the imperial family acquired some of these eunuchs directly from slave dealers, most eunuchs came to the palace as gifts from wealthy officials seeking to ingratiate themselves at court through this exceedingly costly act. Once registered at the palace, eunuchs entered training in their expected duties as well as in proper comportment and etiquette. Only after completing this training were they assigned to duty within one of the entourages in the palace.


Princess Ayse notes in her memoir that the eunuchs' duties included the prosaic tasks of "locking the doors to the Imperial Harem each evening and unlocking them in the morning, standing watch at the doors, keeping an eye on those who entered and left, accompanying people to their carriages, escorting doctors in and out, and not leaving alone anyone who had come from outside the palace." But the highest-ranking eunuchs, reporting directly to the Sultan, also came to wield vast power as administrators of the palace and supervisors of wealthy pious foundations.


In terms of numbers, we know that as of 1903 a total of 194 black eunuchs served the imperial family, 31 at the Sultan's residence of Yildiz Palace, and the remaining 163 distributed in service to the other Princes and Princesses in their villas in and around Istanbul.


Norms Adopted in the Translations


Because the British royal court lacked slaves and harems, the English language lacks words to precisely translate a good number of the ranks and titles at the Ottoman court. Where English bears no suitable equivalent the term is left in its Turkish original.


Equerry designates a eunuch who possessed the high rank Musahib (companion). Eunuchs who ranked below this position are simply "eunuch" in the translation, to render the venerable title Aga that by the late nineteenth century at the Ottoman court nearly always denoted a eunuch.


The title of the monarch's mother, Valide Sultan (valide means "mother"), is translated as Princess Mother to approximate the Ottoman original, which sought to honor this lady as the only concubine at court permitted to bear the word sultan in her title. Apart from this lady, among women at court only Princesses bore the title sultan, which in their case followed the given name. The title sultana that one encounters in European texts for Ottoman Princesses is a strictly European invention, never used at the Ottoman court.


In Turkish, shortened versions of a monarch's name occur often; this variety has been retained. Thus Mecid, Aziz, and Hamid for the Sultans Abdülmecid, Abdülaziz, and Abdülhamid, respectively.


Ottoman Princes of this era usually received two given names, "Mehmed" followed by the personal name unique to that Prince and by which he was commonly known. For instance, at his accession Prince Mehmed Resad became commonly "Sultan Resad," although formally he was "Sultan Mehmed V." In the memoirs one encounters these Princes' names both with and without "Mehmed," and the translation follows this usage.


Calendar dates in the Islamic Hijri calendar or the Ottoman malî calendar have been converted to the Gregorian calendar.


For determining current-day values of sums of money mentioned in the text, the following comparison should prove most useful. At the time of our memoirs, the Ottoman lira or pound consisted of 100 kurus or piasters. In 1914, during the period of our memoirs, the average daily wage of a worker in Istanbul amounted to ten to twelve kurus, so that the average monthly wage of a worker in Istanbul totaled approximately three liras. The average annual income of a worker in Istanbul, then, came to around thirty-six liras.


Foreign words not found in a good English dictionary are italicized. Exception is made for the titles kadin, kadinefendi, kalfa, hazinedar, and usta, which recur so frequently that to avoid maddening the reader they are italicized only at first appearance. Spelling of Turkish words generally follows Modern Turkish orthography as in the New Redhouse Turkish-English Dictionary (Istanbul, 1974). Pronunciation of Turkish letters is as in English with the following exceptions:


  • c = j
  • ç = ch
  • ğ (g with breve) = not pronounced, simply extends the length of the preceding vowel
  • ı (undotted i) = the "i" in pick
  • i (including capitalized dotted i) = "ee"
  • ö = the "er" in her; same as French eu or German ö
  • s = sh
  • ü = "ee" pronounced with rounded lips; same as French u or German ü


Brackets indicate insertions by the translator, while parentheses indicate insertions by the authors in the original texts.


Footnotes followed by bracketed initials identify notes from the original Turkish texts that have been retained in the translation, as follows. Notes not followed by a bracketed initial are those of the translator.


  1. Concubine Filizten
    • [S] = Footnote by Ziya Sakir
  2. 2. Princess Ayse
    • [A] = Footnote by Princess Ayse
    • [ÖO] = Footnote by the Princess's sons Ömer and Osman in subsequent editions
  3. 3. Schoolteacher Safiye
    • [Ü] = Footnote by Safiye Ünüvar



Available for Kindle
Available on Google Play
Available on Kobo
Available for Nook
Available on the Apple Store

This book may also be available on the following library platforms; check with your local library:
3M Cloud Library/bibliotheca
UPCC/Project Muse