Gendered and Ethnic Captivity and Slavery in Safavid Persia: A Literature Review

: The Safavid society’s approach to sexuality and gender has made it a reference for the “pre-modern” discourse, in which gender and sexuality manifest—in contemporary terminology— queerness and ﬂuidity. While it is important not to romanticize the image of the Orient as a queer heaven, it is possible to consider Safavid society as an important site of inquiry that offers valuable insights on pre-colonial gender and sexuality. A less discussed topic in Safavid literature on gender and sexuality is gendered and sexual slavery. This study conducts a review of primary and secondary literature on the Safavid period, including Western travelogues. The paper aims to outline the relations between existing forms of captivity and factors such as gender and ethnic background. The research explores the prevalence of slavery reported in Safavid literature and how slaves’ positions were deﬁned within social hierarchies.


Captivity during the Safavid Period
Human captivity and slavery as a social, economic, political, and cultural institution, has been present in many countries. Local histories of slavery have been a constant source of anxiety and embarrassment for all societies that have practiced it, including some Muslim contexts (Clarence-Smith 2006). As discussed by Babaie et al. (2004), in the Islamic world, slavery is a process that begins with the enslavement of non-Muslims and then converting them and recommending their subsequent manumission. In the legal discourse, it is religion rather than ethnic affiliation that justifies enslavement (Babaie et al. 2004). Slaves in interpretations of Islamic texts were defined carefully and were either born as slaves or had been enslaved in battles and other "rightful" practices. Despite this seemingly clear legal definition, slaves were allocated a bewildering variety of social roles with different social values, from emirs to outcasts (Clarence-Smith 2006). Besides religion as a crucial element, gender and ethnicity have also been suggested to contribute to captivity conditions in Muslim contexts. In existing interpretations of Islamic texts, the "master" is granted access to the bodies of the female slaves. The access to and the usage of the female slave's body was a prominent pattern throughout the centuries (Gordon 2017), and the overlap of gender and "feminine" slavery with complex forms of sexual relationships is historically traceable (Gordon 2017). Religion might be one of the factors that were used to legitimize slavery, but the terms for slavery that exist in the Arabic Language imply gender, function, ethnicity, and social status of the slaves (Nielson 2017), pointing to the diversity of factors affecting captivity.
In Iran, slavery is a highly neglected topic (Khosronejad 2016), and its existence and significance in Iranian history are sometimes downplayed in the literature (e.g., Seydiye 2008). Consequently, the historical practice of slavery is rarely known, discussed, or acknowledged by the Iranian public. There is ample and continuous evidence of "Habashi", "Zangi", "Indian", and "Caucasian" sex slaves, military slaves, and slave servants in brothels, teahouses, harems, and in the middle and upper-class Iranian families and among pastoral clans from the early 1500s to the beginning of the twentieth century (Clarence-Smith 2006;Khosronejad 2016). There are, however, relatively few research articles or dissertations on the subject of slavery, slave trade, trade routes, collection stations, creditors, or slavery for the medieval, early modern, or modern periods of Iranian history (Ricks 2001). This could be interpreted as both a lack of interest and a disregard of slavery in the studies of Iranian history.
During the more than two centuries of its reign, the Persian Safavid dynasty established Shia as the state's official religion and stabilized the empire's borders. The Safavid society's approach to sexuality and gender has made it a reference for the "pre-modern" discourse. This approach and the general discourses of sexuality in pre-modern or early modern Persia have been discussed as examples of classic discourses of sexuality where heteronormativity and heterosociality were not the only socially established norms (Najmabadi 2005;Rahbari 2009Rahbari , 2018. During this period, gender expressions and performance went beyond the dichotomic notions and presented aspects of what in contemporary terminology would be considered gender queerness or sexual fluidity (e.g., Babayan 2008;Losensky 2009;Rahbari 2009Rahbari , 2018. In this essay, I do not use "queer" as an adjective to refer to individual sexual orientation or gender identity. I use it instead to refer to the dominant discourses of gender and sexual fluidity without that fluidity being exclusively "non-heterosexual". Employing these queer cultural aspects, scholars have challenged the Orientalist narratives of sexuality during the period to expose the bias in the studies of the history of the Orient. While this line of study has produced valuable work (e.g., Cronin 2015), it is crucial to avoid the romanization of the pre-modern Orient as a queer and free heaven. On the other hand, while the studies of Oriental history have confronted Orientalist interpretations, it is possible to consider Safavid society as an important site of inquiry that offers valuable insights on pre-colonial gender and sexuality.
This paper aims to explore the relation between captivity and the impact of social factors such as gender and ethnic background in slavery and captivity conditions. It is documented that during the Safavid period, hundreds of thousands of people with different ethnicities and religious backgrounds were (most often forcibly) moved into the Persian midlands from other territories (Maeda 2004;Ricks 2001). The Safavid slave trade system is not extensively documented, and it is hard to assess the demographic changes it brought about (Maeda 2004;Ricks 2001). The very fact that Safavid was a period of mass movements, migrations, and forced displacements makes it very hard to distinguish the changes in the demographic, ethnic diversities thoroughly and to document their patterns of mobility. Many of the displaced populations successfully integrated into the new environment and obtained positions in different strata of the Safavid society. However, the courses of their lives, living conditions, and post-displacement prospective roles depended on factors such as, but not limited to, skin color, language, religion, and gender. This paper uses a bibliographic approach and reviews some primary and secondary sources to reveals some ethnic and gendered aspects of captivity and slavery during the centuries of the Safavid reign in Persia. The reviewed literature includes Western travelogues (e.g., Chardin; Della Valle; Figueroa; Grès; Kaempfer; Olearius; Sanson; Sherley, Sherley, and Sherley; Tavernier), and local resources and histories (e.g., Asef 1969;Eskandar Beig Monshi 1938;Khatunabadi 1708;Mirza Afandi 1995). The paper discusses the role gender played in captivity conditions for married and unmarried women in the following section. Then, it will turn to the significance of ethnic background, religion, and class. The discussion places the findings within the broader context of the Safavid society.

Gender and Captivity
Due to the tidal waves of bonded people that flowed into the Safavid empire, captured from the Caucasus or purchased from India and Africa-many of whom ended up as eunuchs, concubines, officials, or soldiers-the Safavid Shia state in Persia suffered from growing social inequality (Clarence-Smith 2006). Some Western accounts have viewed this demographic surplus of young, enslaved, and displaced foreign men and women, who lacked social support, as the reason behind what they considered a massive sex slavery industry (e.g., Figueroa 1984). Besides these displaced and enslaved populations, others have also considered Persian-born women as captive sex slaves, not due to their legal status as such, but due to their restricted social as well as living conditions and because they were believed to have had primarily served for reproduction purposes (Chardin 1993;Tavernier 2003). While it is widely accepted that the Safavid period was an era of general socio-cultural limitations for women, some scholars have explored the possibility of existing agentic narratives of women during the period (e.g., see Babayan 2008).
The ranks, services, and conditions of the slaves differed greatly. Gender, in particular, played an important role in the conditions of the captivity of the slave and the relationship between the slave and the master. Other studies on slavery in Muslim contexts have noted that, besides men, upper-class women also invested and were involved in the trade and practice of slavery (Myrne 2017). While this study did not find evidence of women's active engagement in the slave trade, both the gender of the slave and the gender of the person whom the slave served broadly defined the array of services they were expected to provide. Slavery was practiced mostly by the upper-class population but was largely accepted in the society. Due to the existing rank differences among the Safavid female slaves, some have discussed that perhaps the "elite" female slaves were better treated than the "common" slaves. While the author did not come across evidence to support or disprove this, there is evidence that in other contexts, such as Abbasid Iraq, studied by Myrne (2017), elite slavery did not secure a person; in fact, these slaves were probably also more exposed to punishments and restrictions. I will return to this topic in the following sections.

Marriage, Concubinage, and Motherhood
Some legitimized forms of marriage in Islam allow men to have four wives and an unlimited number of concubines simultaneously. In earlier Muslim contexts, the widespread practice of polygamy has sometimes been considered to be partly to blame for the prevalence of sexual slavery in the form of concubinage (Myrne 2017). As Afary (2009) has argued, in pre-modern Persia, some shunned marriage, believing that sexual relationships with women would diminish their physical prowess. This belief, among other cultural grounds, meant that the practice of non-heterosexual relationships was accepted and prevalent during the Safavid period. The eroticization of the male body and homoerotic love were trends in the medieval Islamic period and peaked in popularity in the Safavid era (Yaghoobi 2016). Yaghoobi (2016) analyzed Persian cultural productions, in which homoeroticism is depicted, and argues that these productions could illustrate either master-slave relationships, where slaves were sexually subjected to their masters, or the Sufi practice of Shahidbazi, where younger, beardless, adolescent boys were the object of an adult man's desire. Shahidbazi practice entailed spiritual, emotional, and sometimes lustful elements (for more on this, see Najmabadi 2005;Zargar 2013).
The existence of queer sexual relationships in Safavid Persia was reported by both local writers and Western travelers (Asef 1969;Olearius 2006;Sherley et al. 1983). The Western travelogues also documented the prevalence of male slaves in Safavid brothels (Afary 2009). The prevalence of "non-heterosexuality", however, did not hinder heterosexual marriage. Marriage and concubinage were integral for reproduction and were practiced widely. In this line, Chardin (1993) has emphasized the importance and value of childbearing for women. Procreation was so crucial that, as written by Western travelers, after women experienced menopause, the husbands would refer to younger and fertile slaves or concubines for their ability to bear children (Chardin 1993). For elite families, marriage served not only procreation but also the formation of socio-political bonds between the families.
Western travelers, such as Tavernier (2003) and Chardin (1993), wrote that Safavid marriage was comparable to the state of enslavement. Marriage in the early Islamic period was very much a family matter, and kin's involvement in arranging and concluding women's and girls' marriages was common (Ali 2010). This form of close kin involvement was also the case during the Safavid period (Rahbari 2009). The number of women acquired by buying, concubinage, and marriage varied, and reached hundreds in the case of elite men and royals. Shah Safi (r. 1629-1642) had three permanent marriages and three hundred concubines and slave girls (kaniz) in his harem (Olearius 2006). Shah Sultan Hossain's (r. 1694-1722) court was estimated to have five thousand male and female black and white slaves, including the one hundred black eunuchs who comprised the royal party (Ricks 2001). These numbers alone indicate that the marriages and concubinage did not always serve mere sexual and procreation functions.
The Safavid kings' wives were not always depicted as a victimized group. Some Western and local historians have presented Safavid harems as a joyful and plentiful place (e.g., Della Valle 1991;Kaempfer 1985). However, the fact remains that harem women, even the favored ones, were in danger of the kings' wrath. Shah Abbas II (r. 1642-1666) burned three of his wives alive because they refused to drink with him (Sherley et al. 1983), and burned another wife for lying about her menstruation period (Chardin 1993). Shah Safi (r. 1629-1642) stabbed his wife to death for disobedience (Sherley et al. 1983). The Sherleys and Chardin did not mention whether these women were permanent wives or concubines. Della Della Valle (1991) reported that the eunuchs were instructed to kill all the king's current wives if he was defeated in battle, in order to maintain the king's honor connected to his religiously wed wives. Among the ordinary people, adulterous women-or women who were perceived or rumored as such-were reportedly thrown into pits, thrown down high minarets, or burned (Sanson 1967;Sherley et al. 1983;Tavernier 2003). Olearius (2006) reported that a man killed his whole household, including his wife and all his children, to clear his name and restore his honor after rumors of his wife's unfaithfulness circulated. The notion of honor and its connection to the household's male head is highlighted explicitly in these narratives.
It seems that this notion of family honor (or perhaps, male kin honor) was not ubiquitously applied in the Safavid family. Figueroa (1984) documented married women who were performers and sex workers, but there are no detailed explanations in regard to the division of the sex worker's time and labor between the two. There are other narratives of female sex workers who, after accumulating wealth, got married and practiced repentance (in Persian, Towbeh) (e.g., Chardin 1993;Eskandar Beig Monshi 1938).
It is documented that gifting women, both concubines and female slaves, was practiced; if not common in society, it was at least practiced in the royal house. Due to the constant surplus of the king's wives-who were recruited continuously among younger enslaved and free women-women were gifted to the royal house's employees and other elites (Chardin 1993;Della Valle 1991;Olearius 2006). Hamid (2017) had also reported concubinage in the case of Timurid Persia (1370-1507), when the concubines were not always slaves, but women born free. The conditions of slave concubines and concubine free women were, thus at times, not so different.
It seems that in some Muslim contexts, concubinage was an exclusive role for enslaved women, but this was not the case during the Safavid period in Persia. While many concubines were slaves, other concubines were born free and were gifted to the royal court by the women's families or forcibly taken by the royal court. The alliances made through marriages served social and political goals. Gifting attractive women to the king and positioning women in the royal harem were ways to solve the problems of uncertain political competitions and ensured loyalties (Hamid 2017). The inclusion of free women among the ranks of royal concubines, despite the ready availability of female slaves, suggests that concubinage served the political characteristics of familial bonds (Hamid 2017) and was practiced for more than dynastic continuity and procreation. This does not mean that these women were not sexual slaves. Whether they engaged in sexual relations, women were, because of their gender, gifted, wed, and marketed.
Although they were victims of sexual and social abuse and suffered under captivity, slave women, in broader Islamic history, attempted, sometimes successfully, to gain social mobility and reach higher status and power positions. In the Abbasid period, female concubines used sexuality and sensuality to conquer men, who preferred them to free women due to their obedience (Myrne 2017). According to Myrne (2017), the concubines of Abbasid Iraq were even freer than free-born women in some respects. They were allowed more mobility, both symbolically and geographically. They acquired education supported by their masters since their masters gained prestige through their performances as skilled singers and poets. Another way to gain status and freedom was through procreation and motherhood. A man's marriage to a slave could change the woman's status legally and socially (Sharlet 2017). While the bearing of a child did not always lead to women's freedom, it did make them the mother of the master's child, for which they could no more be sold, and she would be freed by the master's death (Urban 2017), if not by his will. The evidence shows that the role of female slaves, in early Islamic history, was connected to motherhood and the value of childbearing (Robinson 2017). Mariya, the Prophet's slave concubine, was made free by giving birth to a son (Ibrahim), as declared by the Prophet himself (Urban 2017). Besides the possibility of social mobility, this example also indicates the social value attributed to motherhood. However, the children of the slave concubines from otherized ethnic backgrounds were a source of great consternation to those who wished to preserve unadulterated ethnic purity and supremacy (Dann 2017).
During the Safavid period, women's ethnic background did not become irrelevant after giving birth, becoming mothers, or even gaining one's freedom. While it seems that the idea of ethnic purity was not central in Safavid ideology, Shah Ismail II (r. 1576-1577) once stated that the kaniz-the enslaved girls-did not deserve to share a bed with him (Taheri 1970). Since the kaniz were sometimes forcibly turned to Shia, Shah Ismail II might have referred to their class and social status or ethnic background. The supposed "acceptance" of the slave women as concubines, or freeing them through motherhood, could also be viewed from a lens that reduces them to vessels of male lineage. Inclusion of the children born to them in the father's lineage could have meant that women's genealogical contributions to the children's identity were overlooked (Urban 2017). Such identities were not entirely forgotten, but were resisted under political and patrilineal suppression.

Performers, Domestic Workers, and Sex Workers
Besides concubines, the courtesans, slaves or performers, and domestic workers of diverse genders were categories of captives prevalently present in Safavid spaces. The relation between captivity and musical entertainment is documented in Persian territories, and broadly in the other Muslim contexts and periods. Musical entertainment by slaves is reported to have been present in the Umayyad and Abbasid courts. During the Safavid period, dancing, playing music, and serving continued to be practiced by male and female slaves as well as free men and women (Figueroa 1984;Sanson 1967). Local writers and foreign travelers widely documented the performances of dancers, musicians, and other entertainers of different gender groups.
The performing women sometimes managed to subvert the social norms by their performances and shift the submissive role to the patron (Nielson 2017). In some cases, the performing artists gained fame and power by performing outstandingly, and standing out or attracting support and prestige through their masters. This route to relative power meant that some of the performers would eventually have a support system. Olearius (2006) reported that the Safavid men in power sometimes supported sex workers. He witnessed that a governor very seriously followed up a complaint by a sex worker who had not been paid by a client and made sure that the sex worker received payment for her services. Travelers and local historians have documented the popularity of sexual entertainment and treating guests with sexual services (Asef 1969;Olearius 2006;Rahbari 2018), including sex workers, or sometimes erotic performers available for sex after or during their performances. Sherley et al. (1983) have reported that, while sex workers of all genders were available and it was common practice to invite them to entertain elite men, non-consensual sexual advances towards sex workers and service members were not tolerated and, if revealed, were sometimes subject to punishments.
According to Clarence-Smith (2006), however, other activities of the slaves are overshadowed by sexual and domestic slavery in the literature on slavery in the Islamic world. He argues that female slaves performed many other tasks; for instance, the Inner Asian Turkmens had to watch the flocks, prepare the food, make felts, and weave carpets. Those belonging to South Asian Muslims performed a variety of outdoor tasks. I believe that this is the case in the literature on the Safavid period too. Not only is the literature predominantly limited to the urban areas, but it is also highly concentrated on elite and royal life, the somewhat imagined glamours of the harem. Therefore, other activities of the female slaves and workers may be missing from the literature that reduces women's work to sex or domestic and private-space activities.

Ethnic Background, Religion, and Class in Captivity
Anthropologist Pedram Khosronejad (2016) points out that the African slave trade in the Persian Gulf began before the Islamic period. Slaves and slave trading in Iran were recorded in written texts as early as the third century, AD (Ricks 2001). Medieval accounts refer sporadically to slaves working as household servants, bodyguards, militiamen, and sailors in the Persian Gulf, including, what is today southern Iran. The practice lasted and evolved through many centuries (Khosronejad 2016). Therefore, slavery in Iran did not start with Islam's presence in the region but did continue throughout centuries of Islamic Persian history.
The Prophet of Islam himself repeatedly rejected discrimination on the grounds of race and ethnic background. Black and Persian (freed) slaves were among his most prestigious companions (Clarence-Smith 2006). While he urged kindness toward the enslaved, the slaves remained at the mercy of their master. Sometimes, the slaves developed intimate ties with the masters (Babaie et al. 2004), but in general, the ethnic and racial stereotypes common in popular culture were akin to dynastic, sectarian, and nationalist loyalties (Clarence-Smith 2006). Persian scholars and religious figures played a role in the normalization of the practice of slavery, as well as its ethical and juristic conditions. Before the Safavid period, Ibn Sina (980-1137), the famous Persian physician and philosopher, defended the idea of natural servitude, further propounding that extremes of temperature produced temperaments suitable to slavery among the fairer and darker people, notably Slavs, Turks, and Africans (Clarence-Smith 2006). In another account, Nasir al-Din al-Tusi (1201-1274), a Persian Shia writer on ethics, described slaves as the gifts of a beneficent God to humankind. He admitted that some slaves were free "by nature", yet he recommended that they be well treated rather than liberated (Clarence-Smith 2006). These theories could have contributed to the naturalization of racial superiority of "Persians" and the normalization of slavery.
During the Safavid period, the Shia doctrine played an even more significant role in defining the positions of the slaves in society and forming the Islamic Shia juristic viewpoint. Two particularly important scholarly Shia texts (fiqhi books) were written to directly address slavery upon Shah Sultan Hossain's (r. 1694-1722) request. One of these texts, written by Mohammad Saleh Khatunabadi (1648Khatunabadi ( -1714, a Shia faqih and religious scholar, provided an extensive delineation of who could be "legitimately" enslaved from regions surrounding the Persian Empire's borders. In this document, he referred to skin color and geographical location of the populations to describe them, but the final factor determining whether enslavement could be legitimately practiced was the religion of those captured, according to which, only Muslims, regardless of which branch they belonged to, could not be enslaved (Khatunabadi 1708, edited by Jafarian 2011. The other text, written by Mirza Abdollah Isfahani (known as Mirza Afandi, 1656-1718), discussed the conditions of freedom of the slaves. In response to the king's question whether castrating the slaves freed them automatically, Mirza Afandi wrote the document (Jafarian 2011) in which he rejected the idea, and thereby gave the king a "religious" basis to keep his castrated slaves (Mirza Afandi 1995). This latter document's legitimization of the dismemberment of the slaves is a reflection of the normalized and extreme violence towards enslaved bodies.
These ideological foundations led to the continuation of slavery in later Iranian history, where slaves continued to be an integral part of the elite households (Khosronejad 2016). Slave trading and slave labor contributed to the Safavid building construction, public services, road and bridge construction, and long-distance trade (Ricks 2001). Darkerskinned and black men continued working mostly as eunuched guardians inside the king's harem and houses, and darker-skinned and black women were servants to elite women (Khosronejad 2016). As seen above, ethnicity is a difficult case to look at since it did not directly relate to the Muslim legal doctrine, while other factors, such as gender and religious affiliation, were crucial to the rights and obligations of slaves (De La Puente 2017). However, there is enough evidence that skin color and perceived ethnic background did affect the conditions of captivity and the functions of the slaves.
In the studies of slavery during the Safavid period, Caucasian groups such as Circassians, Georgians, Russians, Armenians, and people from religious backgrounds such as Christians, were named as prominent groups that Safavid Persians kept captive as eunuchs, slaves, domestic workers (kaniz), wives, and concubines (mentioned in several travelogues, e.g., Della Valle 1991;Figueroa 1984;Grès 1973;Kaempfer 1985;Olearius 2006). Shah Abbas is reported to have enslaved up to 130,000 men and women in a battle with the Kingdom of Kartli-Kakheti (Ghaffarifard 2005). Persians relied heavily on Caucasus slaves, capturing 15,000 Georgians in a single campaign in 1795 (Clarence-Smith 2006). Figueroa (1984) has written about the slave trade and market. He discussed that due to the high demand in the market, Tartars and Lezgins, and sometimes fellow Circassians, snatched Circassians, men, and women and brought them to Persia to sell in the slave markets.
The tensions among the Shia and Sunni branches of Islam gave rise to Muslims' enslavement of members of other branches in some Muslim contexts. Despite Islam's ban on the enslavement of fellow Muslims, the Ottomans and the Uzbeks regarded it as legitimate to enslave Iranian Shia, considering them infidels (Blow 2014;Erdem 1996). However, there is evidence that Safavids did not enslave Sunni Muslims. According to the Safavid historian Eskandar Beig Torkman (known as Beig Monshi), the Safavid authority treated their Muslim hostages "better", by which he meant that, instead of enslaving them, the Safavid converted the hostages forcefully to Shia (Maeda 2004) or killed them if they refused conversion.

Eunuched and Elite Slaves
Eunuchs were the trusted guardians of the harems because their genitals had been partially or entirely removed. Babaie et al. (2004) discuss that all slaves were uprooted, but eunuchs were desexualized, and black eunuchs were at service in harems. Chardin (1993) reported a hierarchy among the eunuchs that was not only based on ethnic background but also skin color; in a way that the eunuchs' skin colors defined their positions as the guardians of the harem of the king. He reported that the darker-skinned eunuchs were the only ones allowed and trusted in the harem's inner spaces and around women, while the lighter-skinned eunuchs acted as the protectors of the outer ports (Chardin 1993). Kaempfer (1985) gave the same description and reported that only dark-skinned eunuchs could move freely around the harem and investigate it. The existence of different roles based on the skin color of the eunuchs indicates that if not ethnic background, skin color did, in such cases, make the conditions of slavery different. Chardin, Kaempfer, and other Western male travelers did not have access to the harem, and their accounts of its structure, as well as its imagery, should be approached with necessary skepticism. Marie Petit, the French female ambassador to Persia, did not document the harem life (Grès 1973). Additionally, the castration of the male slaves was not only for desexualizing them, and so to protect the harem women, but also so that they would have no heirs (Clarence-Smith 2006). Sanson (1967) stated that the Persian king was the eunuchs' heir, and if they accumulated wealth in their lifetime, it would belong to the king's treasury after their death.
Safavid rulers also imposed castration as a punishment (Clarence-Smith 2006), and there is evidence that the castrated free men were sold into slavery just as enslaved eunuchs were. This form of punishment happened for several reasons. Castration and slavery could sometimes be the punishment for a committed crime by a free man. Tavernier (2003) discussed that thieves, who could not afford to return the value of their stolen goods, were sold as slaves, and if that was not sufficient, their wives and children were sold by order until the debt was paid off. Chardin (1993) has also referred to political and personal quarrels and reported that a local governor sold a person he had a quarrel with as a slave, then castrated his sons and gave them to servitude, and sold his daughters into slavery. Castration, however, did not always lead to slavery. Saru Taqi, the grand vizier of the Safavid empire, was castrated due to his promiscuous behavior around young boys (Chardin 1993), but he maintained his position as an important political figure after castration. These examples infer that besides castration, social status and class were factors defining captivity's occurrence and conditions. Clarence-Smith (2006) has argued that Shah Abbas I (r. 1587-1629) appointed slave governors in lands he conquered and created elite slavery. An Armenian from Georgia served as the army's commander-in-chief for more than fifteen years (1595-1613) (Encyclopedia Iranica n.d.). There were estimated to be a thousand military slaves and three thousand eunuchs serving the Shah in Isfahan, some of whom were in power and held elite positions. This tremendous growth of elite slavery provoked a religious backlash. Shah Abbas II (r. 1642-1666) ordered Saru Taqi's murder in 1645 and undertook a return to Shia orthodoxy. The ulama objected strenuously to Georgian and Armenian converts' political dominance, perhaps because they doubted their sincerity. Military and administrative slavery thus declined, albeit without disappearing (Clarence-Smith 2006).
Scholars of Islamic history have documented the other aspects of the activities of the female slaves besides their captivity, and have shown that these women could have found trajectories into freedom and even higher status, and thus shape a particular elite group as well. Throughout Islamic history, educated female slaves were signs of status for the owners and were exchanged as gifts to maintain alliances (Sharlet 2017). A significant number of elite children were born to concubines, and their free male masters and childbearing could have affected the quality of the concubine's life. Some Shia Imams were born to concubine women as well (Gordon 2017). Educated slaves involved with poetry, the Quran, or music sometimes rose in power and position (Sharlet 2017). Singing girls also exerted considerable sway, even over caliphs. If a concubine bore a son for a mighty man, she could wield influence, especially after the death of her master. The women and eunuchs had such influence that the early seventeenth-century Ottoman Turkish empire was called the "sultanate of the women", or the "sultanate of the African eunuchs" (Clarence-Smith 2006). However, female slaves' avenues to social distinction were much narrower than their male counterparts (Dann 2017). Despite this, examples of female slaves' social mobility exist. For instance, under Shah Safi (r. 1629-1642), some concubines replaced princesses as the mothers of rulers, and hereditary families of elite administrative slaves also emerged. As limited as it is, this evidence shows that even the social structures based on established social markers of status were sometimes weakened and disturbed.

Discussion and Concluding Remarks
This paper reviewed some of the literature on the prevalence of slavery throughout the Safavid period and the impact of gender and ethnic background in the conditions of slavery. The paper showed that slave trading and slave labor contributed to Persia's political, economic, and social changes (as shown in other studies, e.g., Ricks 2001). The practice of slavery was prevalent in the form of forced migration of hundreds of thousands of people from different regions into mainland Persia. Movement was thus an unavoidable part of slaves' lives, as travel for new slaves in the Islamic world began with the captivity of the slaves after the battles, followed by transportation to slave markets (Tolmacheva 2017). Shah Abbas I (r. 1587-1629) expanded the earlier system of slave raids into the Christian, Georgian, and Armenian regions of the Caucasus and conducted mass resettlements of urban and rural people within Iran and along the empire's borders (Ricks 2001). There is evidence that Safavid slavery took place across gender, ethnic, and religious backgrounds (with the exception of Islam in the latter case). The reviewed literature also showed that gender, ethnic background, skin color, and class were factors that possibly affected a slave's position in the Safavid society.
Enslaved men and women slaves took on different roles in the Safavid society. Yet, they also engaged in some common activities as service and entertainment staff, and within the sex industry. Enslaved women and sex workers often traveled, as they accompanied the soldiers to the battle for moral support (Figueroa 1984). Besides, when the masters traveled, the female domestic slaves were sometimes taken along because the free women could avoid the danger and discomfort by declining to accompany the husband; but the concubine slave's status did not allow such agency (Tolmacheva 2017). Olearius (2006) reported that traders and elite men who traveled a lot also had several wives in different cities, where they stopped during their journeys. Moreover, since slave women were sometimes gifted, they were also exchanged during travel, and thus ended up in environments they were not familiar with (Tolmacheva 2017). This shows the precarity of the slaves' status that could again upset their lives and result in considerable changes in their living conditions.
There are many limitations to the study of Safavid history, including studies on captivity and slavery. One such limitation in the case of slavery is the scarcity of academic literature. As mentioned by Ali (2010), who studied early Islamic periods, another bias is the urban bias in the general studies of slavery in Muslim contexts, while a large number of slaves worked in rural sectors, such as agriculture and mines. This same bias exists in Safavid histories, as the rural Safavid life is not extensively documented; this is particularly the case for Safavid rural and everyday urban women's lives since most travelers were drawn into the royal court and cultural hubs in bigger cities, such as Isfahan. Additionally, the prevalence of queerness and its coexistence and/or conciliation with Islam in Iran's pre-modern history was rejected by the Iranian modernist, nationalist, and colonial thinkers who rejected both queerness and religion as signs of anti-progressiveness (Rahbari 2020). These perspectives "imagined" the nation of Iran as "authentically" non-Islamic, and void of queer and "oppressive" culture. This imagination has haunted popular and scholarly opinions on Iran's history and established narratives of the nation where slavery and queerness belonged to the "other", and hence, were not part of the Iranian nation's authentic identity.
Safavid slavery is, however, only a small part of the Iranian history of slavery and is a topic with many intricacies. As partially presented in this paper, not only gender and ethnicity but also other factors, such as age, class, social status, skills, and desired embodied characteristics, affected the conditions of slavery and the lives of slaves. The discourses of slavery have drawn on religious, moral, and cultural norms and beliefs. They have been justified by scholars and religious figures, and have served demographic, economic, and political goals. This exploratory study aimed to review some primary and secondary literature and has been very limited in its analytical scope. The essay thus merely scratched the surface in terms of the Safavid era's conditions of captivity and slavery, and their relation to the gender and ethnic background of the slaves. Further contextualized studies that include the viewpoints of all populations involved in the slave trade-whether as partakers or majorly enslaved or affected populations-are required to create a broader and more precise picture of slavery in Safavid Persia. This study substantiated that, most undeniably, slavery was practiced in Safavid Persia, and slaves have contributed in many ways to Persian society and history, a contribution that has largely remained unacknowledged.