1 Introduction

As is well known, one of the most important engines of social and religious change in the early Islamic period were unions between Muslim men and non-Muslim women. According to the Qurʾan (cf. Q. 5.5) and early Islamic law, Muslim men were entitled to take up to four wives, including Jews and Christians.² Muslim men were also entitled to take as many concubines as they pleased and could afford. In the early period, most of these were captives from non-Arabian, non-Muslim backgrounds, namely women who were rounded up during the conquest of new territories and then converted to Islam when they joined Muslim households.³ In both contexts, that of marriage and slavery, Muslim men and non-Muslim women often had children together. Though officially Muslims under the law, these children were raised between religions and cultures, with a foot in the worlds of both their fathers and their mothers to varying degrees. Thus, within the first few generations after the Prophet’s death, marriage and concubinage helped build cultural and social bridges between the Muslim rulers and their far more numerous non-Muslim subjects.

This essay explores a little-known dimension of this story, namely family ties between elite Muslim men and elite women from the non-Muslim conquered populations. By and large, these women were members of the aristocracy of late antique empires and kingdoms that had been defeated and overturned by the new Islamic caliphate. While their male relatives were either killed or assimilated into the new Muslim ruling class, these women faced a more ambiguous fate that can be hard to study on the basis of the surviving literary sources, late, legendary, and tendentious as they tend to be.

For our purposes, they are interesting for three main reasons: First, these women tell us something important about how Muslims related to the pre-conquest elite of the territories they governed. At times, as new rulers trying to expand their base of power, they used elite women to cement alliances with the indigenous rulers of the lands they now controlled. But at other times, as vengeful conquerors trying to humiliate vanquished enemies, they used elite women to abase the pre-existing aristocracy.

Second, whatever their basis in reality, stories about high-status non-Muslim women who entered Muslim households often became extremely elaborate over time, even legendary. As such, they tell us something significant about how aristocratic, non-Muslim lineage was remembered and manipulated by later generations of Muslim writers, including by those who claimed to possess such prestigious lineages themselves.

Third, stories of high-ranking non-Muslim women demonstrate that early Muslim society considered maternal lineage to be very important. While it is true that social standing was largely based on the father, the mother’s lineage was also very important; this is reflected in the manner in which high-status Muslim men pursued unions with elite non-Muslim women and later commemorated in these medieval texts. In this respect, the present article builds on the work of Asad Q. Ahmed, who has argued that maternal descent was a key factor in shaping kinship, social status, and political power within the early Muslim elite.⁴

This article consists of a series of case studies from between the seventh and ninth centuries CE, that is, the era of the Islamic conquests and the Umayyad and early Ê¿Abbasid caliphates. They come from a variety of geographical settings, including the Iberian Peninsula, North Africa, the Fertile Crescent, and Iran. They are not meant to be comprehensive, but rather, indicative of a broader phenomenon that has not been studied before, at least explicitly. Indeed, I suspect more thorough research would turn up other examples from the period, not to mention parallels from later moments in Islamic history as well as completely different cultural and geographic contexts.⁵ Because a number of these case studies are poorly known and may be of interest to non-specialists, I will provide translations of key passages from Arabic and New Persian texts where appropriate.

2 Sara the Goth: a Visigothic Princess among the Early Muslim Elite

When it comes to marriage with elite non-Muslim women – as opposed to concubinage – perhaps the finest example in early Islamic history is Sara the Goth, a Christian and one of the outstanding characters from the conquest of al-Andalus (modern-day Spain and Portugal).⁶ We know about Sara mainly through the work of her descendant, the historian Ibn al-QÅ«á¹­iyya (d. 367/977, “the son of the Gothic woman”), who devoted several pages to her life at the start of his famous account of the arrival of Islam in the Iberian Peninsula:⁷

[The Muslim conquerors] reached al-Andalus, and their situation remained so [i.e., in an alliance with the below-mentioned Visigoths, whose ownership of certain specific estates the conquerors had earlier affirmed] until the death of Almund [a Visigothic prince], who appointed as his inheritors his daughter, Sara the Goth, along with his two younger sons, the bishop (al-maṭrān) in Seville, and ʿAbbās [Lat. Oppas], who died in Galicia. Arṭabāsh [their uncle, hailing from a rival branch of the Visigothic royal family] extended his grasp over their lands, seizing them for his own. This occurred at the beginning of the reign of the caliph Hishām b. ʿAbd al-Malik [r. 105–25/724–43]. Therefore, [Sara] readied a ship in Seville.

Her father, Almund, had preferred living in Seville. His estate included a thousand [smaller] estates in the west of al-Andalus, while Arṭabāsh possessed something similar in the center [of the country], and he remained living in Cordoba. Among his descendants was Abū Saʿīd al-Qūmis [cf. Latin comes, “count”].

[…] At this point, Sara headed for Syria by ship along with her two brothers, until she alighted at Ascalon. She then hurried, halting at the gate of the caliph Hishām b. ʿAbd al-Malik. She relayed her story, as well as the pact which had been made for her father under al-Walīd [concerning the right of his children to inherit his estates], and how she had been wronged by her uncle Arṭabāsh. Therefore, [the caliph] had her escorted into his presence. There, before the caliph, she saw ʿAbd al-Raḥmān b. Muʿāwiya [that is, the first independent Umayyad ruler of al-Andalus, r. 138–72/755–88], who was then just a boy. ʿAbd al-Raḥmān used to mention it to her [years later when he was] in al-Andalus. Indeed, whenever she came to Cordoba, he granted her permission to enter the palace and to see the caliphal family.

Hishām wrote on her behalf to Ḥanẓala b. Ṣafwān al-Kalbī, the governor of Ifrīqiya, ordering him to enforce the pact of al-Walīd b. ʿAbd al-Malik through his governor [in al-Andalus] Ḥusām b. Ḍirār, and he is Abū ʾl-Khaṭṭāb (al-Khaṭṭār?) al-Kalbī, and that was carried out for her.

The caliph Hishām married her to ʿĪsā b. Muzāḥim [a mawlā of the Umayyad family] and he came with her to al-Andalus and recovered control of her estates. He is the progenitor (jadd) of the [Banū] ʾl-Qūṭiyya. She bore him two sons, Ibrāhīm and Isḥāq. He then died in the year in which ʿAbd al-Raḥmān b. Muʿāwiya entered al-Andalus [ca. 138/755].

At that point, Ḥaywa b. Mulāmis al-MadhḥijÄ« desired her greatly, and so did Ê¿Umayr b. SaÊ¿Ä«d al-LakhmÄ«. But ThaÊ¿laba b. Ê¿Ubayd al-JudhāmÄ« had allied himself with Ê¿Umayr b. SaÊ¿Ä«d in favor of Ê¿Abd al-Raḥmān b. Muʿāwiya [at the time of his arrival in al-Andalus], so he married him to her. She bore for him [a son named] ḤabÄ«b b. Ê¿Umayr, the progenitor of the BanÅ« Sayyid, the BanÅ« Ḥajjāj, the BanÅ« Maslama, and the BanÅ« Ḥajz al-Jurz, who constitute the noble descendants of Ê¿Umayr in Seville to this day. He also had children by another woman, but they do not surpass the nobility of these.⁸

The key to understanding Sara’s career is the identity of her father, Almund. This Almund was the son of one of the last Visigothic kings, Witiza (d. ca. 710), who reigned immediately before the Islamic conquest of al-Andalus (92/711). In the first part of this story, Ibn al-Qūṭiyya is describing a dispute among rival factions within the Visigothic elite, in which Almund and his children (including Sara) were pitted against his brother Arṭabāsh (their uncle) for control of land and power. The descendants of Witiza are portrayed as allies of the Muslims, going so far as to leverage their relationship with the conquerors to take revenge on Arṭabāsh and repossess their lands.

The centerpiece of Sara’s own story, of course, is her journey to Syria, then the seat of the Umayyad caliphate. Not only did Sara reportedly win the support of the caliph Hishām himself, but in so doing, she entered into a relationship with him like that of a vassal. This did not entail conversion – as was often customary with high-status vassals in this period – but Hishām sealed their alliance by arranging for her marriage to a Muslim, a mawlā (a non-Arab client) named ʿĪsā b. Muzāḥim, who accompanied her back to al-Andalus. There, they succeeded in reclaiming the lands that had been lost to Arṭabāsh – in effect, transferring Visigothic royal territory to the Umayyads, who were now bound to the descendants of Witiza through the ties of marriage. Ibn al-Qūṭiyya claimed descent from Sara via her children with the mawlā Ibn Muzāḥim, the progenitor of the eponymous Banū ʾl-Qūṭiyya; a separate branch of the family, however, was descended from Sara and her second husband, ʿUmayr b. Saʿīd, and they came to form an important part of the Muslim aristocracy of the city of Seville. In short, with the story of Sara the Goth, we are dealing with a foundation legend of Muslim society in al-Andalus, one that revolves around the theme of kinship.

It is hard to say how much of this story may be true. What is clear is that Ibn al-QÅ«á¹­iyya, who was a descendant of Sara, had a vested interest in glorifying the role of this Visigothic noblewoman in the conquest of al-Andalus. Specifically, he wished to show how, through her alliance with the caliph and her marriages to various elite Muslims, she bridged the divide between the old and new regimes, thereby guaranteeing the survival of the former within the latter (indeed, the very existence of the BanÅ« ʾl-QÅ«á¹­iyya as a group in AndalusÄ« society in the tenth century is clear proof of this⁹ ). We do not know whether Sara was actually significant in the eighth century when she reportedly lived. But as a dramatis persona in a tenth-century source reminiscing about the eighth century, Sara is a symbol of continuity, not of rupture, of the manner in which the indigenous population of al-Andalus, far from being marginalized by the arrival of Islam, allegedly consented to the conquest and reaped its rewards. She is a kind of mythological female ancestor we will encounter again in this article, who became important to posterity because she tied a later author – in this case, the tenth-century historian Ibn al-QÅ«á¹­iyya – to the formative events of Islamic history in al-Andalus.

So much for the literary significance of Sara the Goth. But if we also imagine her as a reflection of social realities in the eighth century itself, she seems to be emblematic of a type of pre-Islamic elite who profited by associating with the new Muslim ruling class, above all, by contracting marriage alliances with them. This was a mutually beneficial arrangement: as Simon Barton argued, it legitimized the act of conquest by embedding the new rulers within networks of old elites, and it kept old elites relevant by making them kinsmen of the new rulers of a given region.¹⁰ In this regard, indigenous aristocrats such as Sara – insofar as we can consider them real historical actors and not merely literary artifacts – were not passive participants in the conquest, manipulated by more powerful Muslims who “actually” ran the show. Rather, they seem to have exercised their own agency, forging strategic ties that enabled them to stay on top, albeit on radically altered terms.

3 Umm Jarīr, the Christian Mother of Khālid al-Qasrī: a Daughter of Kings?

The story of Sara the Goth is not mainly about religion; it is about elite politics and the creation of a hybrid aristocracy, part of which happened to be Christian, part of which happened to be Muslim. If we wish to understand the potential impact of high-status non-Muslim mothers on the religious outlook of their children, we should look east to the Fertile Crescent and the life of Khālid b. Ê¿Abdallāh al-QasrÄ« (d. 126/743–44), an Umayyad governor of Mecca and Iraq who served under the aforementioned caliph Hishām. Khālid’s mother was a Christian – on account of which he was known as “Ibn al-Naá¹£rāniyya,” or “the son of the Christian woman” – and she seems to have had noble blood.¹¹

In some respects, Khālid spent his career acting like a conventionally pious Muslim: according to early Arabic sources, he adorned the Kaʿba with gold, reorganized the rites of the Meccan pilgrimage, and busied himself suppressing “heretical” Muslim groups, including various Khārijīs and Shīʿīs. But in other respects, Khālid’s conduct was very unusual, indeed. He is said to have built a church for his mother behind the congregational mosque in Kūfa and to have invited a priest – not an imam – to bless the fountain in the courtyard of the mosque. He is also said to have favored Christians (and Zoroastrians) for government offices. Finally, he allegedly declared, “Their religion [i.e., Christianity] is better than ours.” The implication of these anecdotes, stylized and fictional though they may be, is that Khālid was deeply (and perniciously) influenced by his Christian mother.

We do not know much about the origins of Khālid’s mother, but we do know she was adamant about her religion, at one point defending her decision not to convert to Islam in a pointed letter she wrote to her son.¹² A rare hint of her background comes from a report in the early ḥadÄ«th collection of Ê¿Abd al-Razzāq al-á¹¢anʿānÄ« (d. 211/827) in a chapter about whether Muslims may take part in funeral processions for non-Muslims:

[Narrated from Muḥammad b. Rāshid]

The mother of Khālid ibn ʿAbdallāh al-Qasrī died and she was a Christian. Therefore, Khālid summoned the Christian bishops (asāqifa) in Damascus [where Khālid is known to have owned a home], saying to them, “Do for her what you do for the daughters of your kings (bi-banāt mulūkikum)! After all, she is one of the daughters of the kings.”

He said: Khālid ordered his female servants to take charge [of washing and preparing the body], while [the bishops] gave them instruction. He said: When they were finished, she was borne away. Khālid rode [alongside her], while the nobles of the city (wujūh al-nās) rode alongside him, coming to pay her respects.

When he arrived with her at the tomb, he turned the flank of his steed and said, “This is our final show of filial piety for Umm Jarīr [that is, Khālid’s mother].” Then he continued, saying, “But I have done nothing for her other than what ʿAbdallāh b. Abī Zakariyyāʾ [a Damascene ḥadīth transmitter, d. 117/735–36] did for his mother.”

Muḥammad [b. Rāshid] explained: Ê¿Abdallāh ibn AbÄ« Zakariyyāʾ was one of the saints (Ê¿ubbād) of the inhabitants of Syria, one of their jurists, and one of their most illustrious men. Makḥūl [a famous Damascus legal scholar and ḥadÄ«th transmitter, d. ca. 112–16/730–34] relayed traditions on his authority.¹³

As the report makes clear, Khālid’s mother was remembered as the daughter of kings. Her funeral in Damascus was apparently a major public event featuring bishops and the local aristocracy and was presided over by her powerful Muslim son. Facing scepticism that Muslims should take part in such funerals, the narrator Muḥammad b. Rāshid compared Khālid’s actions to those of the famous Ê¿Abdallāh b. AbÄ« Zakariyyāʾ, an esteemed Muslim scholar of the period, who, the implication seems to be, fulfilled his duties as a pious Muslim by also participating in his mother’s funeral procession.¹⁴ In other words, Muḥammad b. Rāshid was keen to portray Khālid’s devotion to his Christian mother as fully “kosher.”

It is hard to say whether Umm JarÄ«r, Khālid’s mother, was actually a royal woman. This may be an exaggeration designed to puff up Khālid’s own reputation in the sources. If there is any truth to it, it may mean that Umm JarÄ«r was originally a Byzantine, perhaps even a captive like so many other Christian girls who found themselves living in Muslim households during the Umayyad and early Ê¿Abbasid periods. If we accept her royal status as fact, it could also mean she belonged to one of the Christian Arab tribes who had been allies of the Byzantines before Islam, including the Ghassānids, who claimed royal authority and remained a powerful force throughout the Umayyad period. Indeed, decades earlier, the caliph Muʿāwiya had married a Christian woman from the tribe of Kalb named MaysÅ«n. They had a son, the future caliph YazÄ«d I, who himself had relations with two Ghassānid princesses.¹⁵ Umm JarÄ«r may have been a latter-day version of these high-status Arab Christian women. Regardless, if we are to believe the sources, it is clear that she influenced her son’s religious outlook, and we have to imagine that similar dynamics shaped other religiously-mixed households which included elite non-Muslim mothers.

4 The Anonymity of the Typical Female Slave

The most common way in which elite non-Muslim women appear in the sources is as captives and concubines – not necessarily as wives. Here it is important to remember that medieval Islamic sources usually took, at best, a superficial interest in the backgrounds of concubines, elite or otherwise. This is because, once they entered Muslim households, slave women tended to adopt the religion, language, and culture of their Muslim masters, regardless of what they had been before. In other words, their old selves often experienced social death, and they were reborn with new identities, now as slaves within a new Muslim society.

We can catch a glimpse of this in two medieval works which reminisce about the high-status concubines of the ʿAbbasid period. Both are admittedly late texts, both highly romantic, and both interested in slaves with connections to the court. What is significant is the meagre information they relay about the women’s foreign origins. One assumes that most concubines – whether they belonged to caliphs or not – had little status or prestige before they entered captivity. Like the vast majority of people alive at the time, they had probably been peasants before being enslaved. Many had probably also been children when they were captured. What is key is how these women attained high status – and thus an identity in the eyes of later authors – as a direct consequence of joining the households of powerful Muslims; whatever happened before was more or less irrelevant.

For example, the littérateur Ibn al-Sāʿī (d. 674/1276) described the following concubines of foreign origin in his famous account of the consorts who were associated with caliphs and their ministers:

Al-Muʾnisa al-Maʾmūniyya, a Byzantine slave girl (jāriya rūmiyya). She was one of the concubines of al-Maʾmūn (r. 198–218/813–33) and part of his inner circle […] [No further information of note].

Banafshā the Byzantine, daughter of ʿAbdallāh [a typical convert name, suggesting that her father may have also been a slave of Byzantine origin], a dependent of the caliph al-Mustaḍīʾ (r. 556–75/1170–80) […] [No further information of note].

Ḥayāt Khātūn, slave of the caliph al-Ẓāhir (r. 622–23/1225–26) […] She was a slave of Turkic origin (jāriya turkiyyat al-jins), a favored and trusted concubine, and the mother of one of his sons. She was manumitted upon his death and became a free woman […] [No further information of note].

Shāhān, a dependent in the household of the caliph al-Mustaná¹£ir (r. 623–40/1226–42) […] She was a Byzantine slave belonging to Khatā KhātÅ«n. Khatā KhātÅ«n was the daughter of the commander Sunqur al-NāṣirÄ« the Tall […] Khatā KhātÅ«n took such care in Shāhān’s instruction and training […] When the caliph al-Mustaná¹£ir was given the oath of allegiance, Khatā KhātÅ«n presented Shāhān to him as a gift, as part of a group of slaves. Shāhān alone among them became his concubine and achieved a level of favor and intimacy that no one else could attain.¹⁶

Al-Suyūṭī (d. 911/1505) relays similar information in a separate work about famous slave women:

Qubayḥa the Byzantine, slave of the caliph al-Mutawakkil (r. 232–47/847–61), mother of the caliph al-Muʿtazz (r. 252–55/866–69) […] [No further information of note].

Zamurrud the Turk, mother of the caliph al-Nāṣir (r. 575–622/1180–1225) […] [No further information of note].¹⁷

To reiterate, the takeaway from these passages is how little information about the origins of slave women is contained in these texts. If they are to be trusted, the texts seem to indicate how little these women’s backgrounds affected their identities within Muslim society, aside from a general knowledge of where they had originated (remembering that most biographies of slave women omit this information altogether). Medieval Muslims were not indifferent to the ethnic or cultural origins of slaves: witness the numerous, often unflattering stereotypes that attached to slaves from different regions of the world, as recorded in various medieval Arabic and Persian texts.¹⁸ But by and large, their lives and histories prior to captivity were thought to matter little. I begin with these anecdotes in order to draw a contrast with a very different kind of enslaved woman: those who hailed from elite, non-Muslim backgrounds, who concern us here. These women did not endure social death like the typical slave, but preserved elements of their old identities and were valued precisely for their pre-Islamic social capital. Only by appreciating this contrast can we grasp what made them and their memorialization so unusual.

5 The Fate of the Daughters of Vanquished Enemies

Enslaving and exploiting the women of a vanquished foe has been a longstanding weapon of war and one that is sadly still with us today. Indeed, the humiliation of an enemy on the battlefield often continued with the humiliation of the women, including his wives and daughters, in the bedroom (as the epigraph from Genghis Khan at the start of this article makes clear).¹⁹ The Umayyad and early Ê¿Abbasid periods offer numerous examples of this, including of close female relatives of defeated elites who were captured and forced to join the harems of high-ranking Muslim men. In late-seventh/early-eighth-century North Africa, for example, the Berber chieftain Kusayla launched a famous revolt against the Muslim conquerors that led to the temporary establishment of Berber rule over the broader region.²⁰ Kusayla was originally a Christian but seems to have converted to Islam shortly before his uprising; some sources hint that he reverted to Christianity during the rebellion, but this is a matter of speculation. The Umayyads fought back, and according to some sources, Kusayla was eventually defeated in Morocco by Marwān, the son of the great Arab general MÅ«sā b. Nuá¹£ayr (d. 98/716–17). According to the account in the anonymous al-Imāma wa-ʾl-Siyāsa, written during the ninth or tenth centuries:

Their king Kusayla b. Lamzam was killed, and the number of their captives reached two hundred thousand. Among them were the daughters of Kusayla and their other chieftains (wa-banāt mulÅ«kihim). There were so many submissive women they could not be counted, [women who were so precious] they had no price or value. He said: When the daughters of the chieftains appeared before MÅ«sā, he said, “Bring me Marwān, my son!” He said: So he was brought to him, and he said to him, “O my boy, choose [whichever one you please]!” He said: So he chose a daughter of Kusayla, taking her captive. She became the mother of this Ê¿Abd al-Malik b. Marwān.²¹

This anecdote – which differs from most other traditions about the death of Kusayla, which place events in present-day eastern Algeria, not Morocco – is significant for several reasons. First, if it has any basis in reality, it suggests that Kusayla’s daughter became a victory trophy for the Muslim conquerors. In other words, they celebrated the defeat of the great Berber chieftain by pressing his daughter into the sexual service of the very man who had defeated him. Although the daughter’s religious identity is not foregrounded in the story, the fact that she was treated as a concubine implies that she was imagined to be a non-Muslim like her father, at least before his conversion to Islam and his rebellion. Second, although it is hard to look past the overriding message of Kusayla’s humiliation, the reality is that the union between Marwān b. Mūsā and Kusayla’s unnamed daughter meant that the bloodline of the famous Berber chieftain would survive within the new Muslim elite. Indeed, to claim descent from aristocratic Berbers on one side as well as high-status Muslims on the other was to enjoy a certain social prestige; it was not unlike the Muslims in the eastern part of the caliphate who boasted of being descended from Persian kings, princes, and princesses, as we shall see below.

Third and most interestingly, we know that the child of this union – Ê¿Abd al-Malik b. Marwān – would go onto have a distinguished career, serving as the last Umayyad governor of Egypt before the Ê¿Abbasid Revolution. This Ê¿Abd al-Malik had a grandson of his own, a historian and scholar named Muʿārik b. Marwān, who according to Mahmoud Makki, may have been the author of the anonymous al-Imāma wa-ʾl-Siyāsa, the very text in which this anecdote is preserved. If this is true, it may explain why the text devotes so much attention to the exploits of MÅ«sā b. Nuá¹£ayr – the aforementioned conqueror of the western Maghrib – as well as why the text homes in on the story of Marwān b. MÅ«sā, his union with Kusayla’s unnamed daughter, and the birth of their son Ê¿Abd al-Malik. In other words, this was the author’s own family history.²² This is another example of what we saw in the case of Ibn al-QÅ«á¹­iyya, namely, a later historian glorifying a distant non-Muslim female ancestor who served to connect him to the formative events of Islamic history; this, in turn, raises the question of whether we are looking at faithful depictions of real events or, more likely, later legends designed to serve later agendas.

Similar events unfolded in Iran during the early Ê¿Abbasid period. Ṭabaristān, for instance, was a mountainous region located along the southern shore of the Caspian Sea, which, partly by virtue of its geography, managed to resist annexation by the caliphs until 141–42/758–60, over a century after the conquest of the rest of Iran. Until this point, Ṭabaristān was ruled by a dynasty of local Zoroastrian princes known as the DābÅ«yids, who traced their ancestry back to aristocrats active during the late Sasanian period.²³ During the conquest of Ṭabaristān, Ê¿Abbasid armies are said to have captured several children belonging to the last DābÅ«yid iá¹£fahbad KhÅ«rshÄ«d – who committed suicide in the wake of his defeat. They also captured children associated with another local Zoroastrian lord, the Maá¹£mughān, whose base of power was to the west in neighboring Damāwand.²⁴ An important account of these events comes from Ibn Isfandiyār (fl. 7th/13th c.), the author of a local history of the region with a large amount of early and otherwise unattested information:

After [the capture of the iá¹£fahbad’s fortress, the Ê¿Abbasid army] brought the entire harem – magnificent and respected – to his eminence the caliph [that is, al-Manṣūr, r. 136–58/754–75], all the while preserving their seclusion and chastity. [The caliph] commanded ĀzarmÄ« Dukht and Warjama, saying, “Let them come under my charge so I may marry them (tā nikāḥ kunam),” but both refused. These daughters of KhÅ«rshÄ«d were as beautiful as the moon. [The caliph] gave one of them to Ê¿Abbās b. Muḥammad al-HāshimÄ«, who named her “Ummat al-Raḥmān” and from whom was born [his son] IbrāhÄ«m b. al-Ê¿Abbās. Both the wife and son outlived the husband. The other one the caliph took into his own possession.²⁵

The historian al-Ṭabarī (d. 310/923), himself a native of Ṭabaristān (but who spent most of his life in Baghdad), includes further information about these captives, seeming to describe them as concubines of the ʿAbbasids rather than as their wives:

The daughter [of the iṣbahbadh, sic] was taken captive and she later became the mother of Ibrāhīm b. al-ʿAbbās b. Muḥammad. The [ʿAbbasid] troops then turned towards the Maṣmughān, seizing him along with al-Baḥtariyya, who became the mother of Manṣūr b. al-Mahdī, and Ṣ-Y-M-R, who became the concubine (umm walad) of ʿAlī b. Rayṭa [a son of the caliph al-Mahdī], who was also [the Maṣmughān’s] daughter.

[…] In the course of the appointed night, [the gate of the iá¹£bahbadh’s fortress] was opened for the caliph’s forces, and they killed any warriors inside the city, taking the children captive. Al-Baḥtariyya, who became the mother of Manṣūr b. al-MahdÄ«, was seized. Her mother was Bākand, daughter of the iá¹£bahbadh known as al-Aá¹£amm (“the deaf”) … They also seized Shakla, who became the mother of IbrāhÄ«m b. al-MahdÄ«. She was the daughter of KhÅ«nādān (?), al-Maá¹£mughān’s steward (qahramān).²⁶

There is considerable confusion in the sources about the identity of these women, though these details need not trouble us here. The key takeaway is that daughters and other close female relatives of the DābÅ«yid iá¹£fahbad and the Maá¹£mughān joined the Ê¿Abbasid harem and gave birth to sons who would later play important roles in the high politics of the Muslim empire. For example, the aforementioned al-Baḥtariyya became a concubine of the caliph al-MahdÄ« (r. 158–69/775–85) and gave birth to a son, Manṣūr. Decades later in 201/816–17, this Manṣūr would become the military commander of Baghdad, whose inhabitants – upset that the caliph al-MaʾmÅ«n had appointed the ShÄ«Ê¿Ä« imam Ê¿AlÄ« al-Riḍā as his heir – tried unsuccessfully to persuade Manṣūr to become a counter-caliph. Shakla, the daughter of the steward of the Maá¹£mughān (but elsewhere described as the daughter of an unnamed man from GÄ«lān, a province just to the west of Ṭabaristān²⁷ ), also became a concubine of al-MahdÄ« and with him had a son named IbrāhÄ«m. Unlike his half-brother, he eventually consented to becoming a counter-caliph during the factional violence that surrounded the Fourth Fitna.

There is a certain ambiguity in the fate of these Zoroastrian captives. On the one hand, they represented the vanquished rump of once-proud dynasties that had been lain low by the Ê¿Abbasid conquest. But on the other hand, as concubines, they joined the upper echelons of Ê¿Abbasid society and produced heirs for the Ê¿Abbasid elite. They were thus valued precisely as members of the indigenous Iranian elite, not as semi-random trophies of the battlefield. Therefore, like Sara the Goth and the unnamed daughter of Kusayla in North Africa, their story embodies themes of both defeat and survival.

This trend continued over the coming decades. In ca. 151/768, for example, the leader of a reformed Zoroastrian group known as UstādhsÄ«s, who had rebelled against the Ê¿Abbasids in BādghÄ«s (western Afghanistan) and SÄ«stān (eastern Iran), was defeated and brought to Baghdad in chains.²⁸ He was reportedly accompanied by his family, including a daughter (or daughters?) who were then placed in the Ê¿Abbasid harem. According to several traditions – whose reliability came under question from Wilferd Madelung²⁹ – one of these became the mother of the future caliph al-MaʾmÅ«n (r. 198–218/813–33). According to the account of the Persian historian GardÄ«zÄ« (fl. 5th/11th c.):

Marājil was the daughter of UstādhsÄ«s, who became the mother of the caliph al-MaʾmÅ«n. Ghālib was the son of UstādhsÄ«s and the maternal uncle of al-MaʾmÅ«n. He killed [the famous Ê¿Abbasid vizier] al-Faḍl b. Sahl at Sarakhs on the order of al-MaʾmÅ«n.³⁰

If true, this anecdote underlines the point made above, namely, placing the daughters of defeated rebels in the ʿAbbasid harem embodied the humiliation of their fathers but also expressed a certain high esteem for their bloodlines. Ustādhsīs was not a royal figure like Khūrshīd or the Maṣmughān, but he was a powerful leader in his own right, and this clearly counted for something in the eyes of the ʿAbbasids, who, if we believe the legends, may have used his daughter to produce one of their heirs. But even if the story is not true, such narratives prove that these genealogical claims carried weight and were considered salient in the eyes of elite Muslims at the time.

When it comes to Iranian upstarts, this pattern persisted well into the ninth century. The caliph al-Muʿtaṣim (r. 218–27/833–42), for instance, reportedly kept the daughters of several vanquished enemies in his harem. These included the daughters of Bābak, the notorious head of a Khurramī uprising in Azerbaijan, and Māzyār b. Qārin, the last independent non-Muslim ruler of Ṭabaristān, who carried the flame of “Zoroastrian resistance” in the region for decades after the defeat of the Dābūyids. Niẓām al-Mulk (d. 485/1092) reports the following adab-style anecdote about al-Muʿtaṣim:

One day al-Muʿtaṣim was sitting at a drinking party, and the judge Yaḥyā b. Aktham was present. Al-Muʿtaṣim got up and left the party and went into a room. After some time, he came out and drank some wine. Again, he got up and went into another room; yet again he got up and went into a third. After some time, he came out and went to the bathroom and did a major ablution. He soon emerged and called for his prayer mat; he performed two rakʿas of prayer and went back to the party.

He said to the judge Yaḥyā, “Do you know why I said these prayers?” He said in reply, “No.” He said, “It was a prayer of thanksgiving for one of God’s benefits – to Him be power and glory – which he vouchsafed me today.” Yaḥyā said, “O Commander of the Faithful, what benefit was that?” He said, “In this hour, I have deflowered the maidenhood of three maidens, all of them were daughters of my three enemies; one was the daughter of the king of Byzantium, the second the daughter of Bābak, and the third the daughter of Māzyār the Zoroastrian (gabr).”³¹

The report is so outrageous and its ideological motivations so transparent – to dramatically symbolize the humiliation of al-Muʿtaṣim’s enemies – that one should take it with a truckload of salt. But based on what we have seen above, it is not unreasonable to suspect that the daughters of some of these non-Muslim leaders did end up as concubines of the caliph and that they were forced to satisfy the sexual needs of the very man who had killed their fathers.

It is important to note here that the sexual exploitation of daughters was only half the story. While elite non-Muslim women were often placed in the caliphal harem, their brothers (or other male relatives) were commonly enrolled in the caliphal army and/or administration. There are hints that this also happened, for example, with the sons of Kusayla and KhÅ«rshÄ«d,³² and it is clear that this happened with the sons of UstādhsÄ«s, Bābak, and Māzyar.³³ In the words of Patricia Crone, “The caliph would use the reproductive capacities of the daughters of defeated rebels for the procreation of children for his own family, and the muscle power of their sons for the killing of his own enemies. It comes across as the ultimate humiliation one could inflict on an enemy.”³⁴

6 The Myth of the Captive Sasanian Princess

From the realm of history, let us now step into the realm of myth and consider how social and political norms shaped legends about the elite non-Muslim mothers of prominent Muslim men. Before doing so, it is worth recalling that early Muslim society had conflicting – or perhaps evolving – attitudes towards the importance of matrilineal descent. For example, all of the first four caliphs and nearly all of the Umayyad caliphs were the sons of free-born Arab women, while nearly all of their early Ê¿Abbasid successors were the sons of non-Arab concubines.³⁵ It would be unwise to see this shift as necessarily reflecting changes within all of Muslim society as opposed to just the elite. That being said, it is significant for understanding how the rulers understood the importance of mothers, especially as contributing to the social and political status of their sons (or not, as the case may be).

We can see this difference play out in a famous exchange between Muḥammad al-Nafs al-Zakiyya, leader of a famous anti-Ê¿Abbasid Ê¿Alid uprising in Medina in 145/762, and his enemy, the caliph al-Manṣūr.³⁶ One way in which al-Nafs al-Zakiyya challenged al-Manṣūr’s claim to power was by boasting of his allegedly superior lineage. Not only was he descended on his father’s side from Ê¿AlÄ« b. AbÄ« Ṭālib – the Prophet’s cousin, son-in-law, and legal inheritor (al-waṣī) – but on his mother’s side, he was also descended from the Prophet’s grandmother Fāṭima bt. Ê¿Amr, his daughter Fāṭima al-Zahrāʾ, and his wife KhadÄ«ja bt. Khuywalid. A better bloodline one could not ask for. As al-Nafs al-Zakiyya put it in a letter to al-Manṣūr, as reported by al-ṬabarÄ«:

I am at the very center of the BanÅ« Hāshim’s kinship lines (wa-innÄ« awsaá¹­ banÄ« hāshim nasaban). My paternity is purest among them, unsullied by non-Arab blood (lam tuÊ¿arriq fÄ« ʾl-Ê¿ajam) and uncontested by concubine mothers (ummahat al-awlād) … In both the pre-Islamic and Islamic periods (fÄ« ʾl-jāhiliyya wa-ʾl-islām), God has never stopped electing for me fathers and mothers, even choosing for me in the Fire [i.e., Hell, where pre-Islamic pagans were believed to have gone after death]. I am descended from the person of highest rank in the Garden [that is, the Prophet Muḥammad] and the one with the lightest punishment in the Fire [probably AbÅ« Ṭālib, the Prophet’s beloved uncle and patron who died without formally professing Islam in ca. 619]. I am descended from the best of the good and the best of the bad.³⁷

Al-Nafs al-Zakiyya’s boasting about his maternal lineage was deliberate. His rival, al-Manṣūr, had illustrious forefathers within the family of al-ʿAbbās b. ʿAbd al-Muṭṭalib – the Prophet’s uncle, his oldest surviving male relative at the time of his death, and the progenitor of the ʿAbbasid line. But al-Manṣūr’s mother, Sallāma, was a Berber concubine from North Africa and therefore, bore the double-indignity of being a non-Arab as well as a slave. But this mattered little for al-Manṣūr, who, judging from his reply, had a very different conception of maternal lineage from al-Nafs al-Zakiyya:

My, how you pride yourself on kinship through women, as to delude the vulgar and the rabble! But God did not make women equal to uncles and fathers or [even] to paternal relations and guardians!³⁸

Al-Manṣūr, of course, had a vested interest in downplaying the importance of his mother’s bloodline. That being said, his words seem to reflect a new and different understanding of kinship that was then-ascendant among the elite of the mid-eighth century, namely the overriding importance of the father’s line and the decreasing significance of the mother’s. In light of this, it is no surprise that, going forward, Islamic history is filled with numerous men like al-Manṣūr: caliphs, emirs, sultans, and other powerful rulers who were born to slave mothers, with this having little noticeable impact on their status or influence.

The exchange between al-Nafs al-Zakiyya and al-Manṣūr provides a helpful insight into shifting attitudes towards mothers in the early ʿAbbasid period. But it ignores the fact that a compromise between the two positions was also possible. Along with freeborn women from the Arab elite and servile women from the conquered population, a third option existed, namely, concubines from the non-Muslim ruling class, especially members of pre-conquest royal families. We have already met several of these from North Africa and Iran, where the high status of captive mothers seems to have provided their sons with a certain social boost, even if their lineage was non-Arabian and non-Islamic.

These dynamics help explain an important body of legends, studied by Mohammad Ali Amir-Moezzi, which claim that a daughter of the last Sasanian king Yazdgerd III (d. 31/651) ended up the consort of the third ShÄ«Ê¿Ä« imam al-Ḥusayn b. Ê¿AlÄ«.³⁹ According to these stories, this daughter bore al-Ḥusayn a son, Ê¿AlÄ« b. al-Ḥusayn, better known as “Zayn al-ʿĀbidÄ«n” (ca. 38–94/658–712) who went on to became the fourth ShÄ«Ê¿Ä« imam. As Amir-Moezzi has shown, the earliest layer of Zayn al-ʿĀbidÄ«n’s biographical tradition claims that his mother was simply a slave girl who had been captured in Sindh or Sijistān in the east. In slightly later historical sources, however, his mother was portrayed as a Sasanian princess. The question is why?

According to Amir-Moezzi, the overriding concern was to construct a biography of Zayn al-ʿĀbidīn – and thus, his descendants among the imams – as the bearer of a “two-fold light”: that is, the light of walāya (the loyalty and support due to the imams, thought to be transmitted from one imam to the next via seminal fluid) and the light of glory (which was inherited from the kings of Iran). In other words, the story of the Sasanian princess was designed to show Zayn al-ʿĀbidīn and his successors as heirs of both prophetic and royal bloodlines.

An early version of this story appears in the work of the Shīʿī historian al-Yaʿqūbī (d. 284/897), who writes:

The mother [of Zayn al-ʿĀbidÄ«n] was Ḥarār, the daughter of Yazdgerd the Persian king. This was because when the caliph Ê¿Umar b. al-Khaṭṭāb [r. 13–23/634–44] brought the two daughters of Yazdgerd [in chains], he gave one of them to al-Ḥusayn b. Ê¿AlÄ«. He named her “Ghazāla.” One of the noblemen used to say when mentioning Ê¿AlÄ« b. al-Ḥusayn [that is, Zayn al-ʿĀbidÄ«n], “[Because of him] everyone would love it if their mothers were slave women (imāʾ)!” It is said that his mother was among the captives taken from Kabul.⁴⁰

A much more elaborate version is found in the work of another contemporary Shīʿī author, al-Ṣaffār al-Qummī (d. 290/902–3), who alludes to the notion of the “two-fold light” in his account of the arrival of this Persian princess in Medina:

[Narrated from Abū Jaʿfar Muḥammad al-Bāqir, the fifth imam and son of Zayn al-ʿĀbidīn]

During the reign of the caliph ʿUmar, a daughter of Yazdgerd was brought to Medina, and she was nobler than all the virgins of the city. The mosque [of the Prophet] was illuminated by the light of her face. When she entered the mosque and saw ʿUmar, she disguised her face and said [in Persian], “Āh bīrūj bādā hurmuz” [a garbled version of the phrase, Āh pīrūz bādhā hurmuz, “May Ohrmazd/God be victorious”]. [Not understanding this] ʿUmar became angry with her, saying, “You have just insulted me!” And he became fixated on her.

Then the Commander of the Faithful [that is, ʿAlī b. Abī Ṭālib] said to him, “This is not your concern. Let her choose a man from among the Muslims, then credit her to him as part of his conquest booty (bi-fayʾihi).” Therefore, ʿUmar said to her, “Choose!”

So she came forward until she placed her hand on the head of al-Ḥusayn b. Ê¿AlÄ«. The Commander of the Faithful said to her, “What is your name?” She said in reply, “Jahān-shāh.” He said, “Rather, [it shall be] Shahrbānuwayh [that is, Lady of the Land]!” Then he looked at al-Ḥusayn and said, “O AbÅ« Ê¿Abdallāh, may she bear you a son who is the best person in the world (ghulām khayr ahl al-arḍ)!”⁴¹

There is a wide variety of similar stories about the princess, including ones claiming she converted to Islam.

There is no consensus about the fate of Yazdgerd’s daughters after the conquest, but it seems that legends about the mother of Zayn al-ʿĀbidīn prompted rebuttals, including by Sunnīs. A good example of this comes from al-Ṭabarī, who notes the following in a report about the caliphate of ʿAlī ibn Abī Ṭālib, who dispatched a deputy named Jaʿda ibn Hubayra to Khurāsān on certain business:

[While in Khurāsān, Ibn Hubayra] obtained two slave girls from the Persian royal family (jāriyatayn min abnāʾ al-mulÅ«k) who had been granted safe-conduct and sent them to Ê¿AlÄ«. He proposed that they should convert to Islam and that he should marry them off. [The girls] said in reply, “Marry us to your two sons [i.e., al-Ḥasan and al-Ḥusayn, the second and third ShÄ«Ê¿Ä« imams],” but he refused that. One of the dihqāns [that is, a member of the former petty aristocracy of the Sasanian Empire] said to him, “Hand them over to me, for that would be an honor by which you would honor me.” Ê¿AlÄ« handed them over, and the two of them stayed with the dihqān, who spread out for them silken carpets and gave them food from golden vessels. Then they returned to Khurāsān.⁴²

Although the passage does not mention Zayn al-ʿĀbidīn, it is tempting to imagine the story of the refused betrothal to al-Ḥasan and al-Ḥusayn and the return of the Sasanian princesses to Khurāsān served as a riposte to Shīʿī claims that their imams were descended from such women.

The motif of elite non-Muslim mothers was not restricted to the fourth imam. Several other imams were reportedly born to slave mothers, and as Michael Dann has shown, medieval sources go to great lengths to portray these women as possessing noble blood. The mothers of the seventh and eighth imams, MÅ«sā al-Kāẓim (d. 183/799) and Ê¿AlÄ« al-Riḍā (d. 203/817), respectively, were reportedly from “the nobles of the non-Arabs.” The mother of the ninth imam, Muḥammad al-Jawād (d. 220/835), was allegedly a Nubian and a relative of the Prophet’s beloved concubine, Mary the Copt. The mother of the twelfth imam, Muḥammad al-MahdÄ«, who famously went into occultation in 260/874, was named NarjÄ«s and was allegedly a granddaughter of the Byzantine emperor (as well as a descendant of St. Peter!).⁴³

ShÄ«Ê¿Ä«s were not the only Muslims who profited by claiming aristocratic Persian lineage. Zooming out from the special case of women, a number of medieval Muslim dynasties claimed descent from Sasanian nobles, though on the male side. These included the Rustumids – a dynasty of Ibāḍī imams in North Africa who reigned between the mid-eighth and early tenth centuries – and the BÅ«yids – a family of Iranian condottieri from Daylam who famously subjugated the Ê¿Abbasid caliphs between the tenth and eleventh centuries.⁴⁴ When it comes to women, similar legends surrounded the Umayyad caliph YazÄ«d (III) b. al-WalÄ«d b. Ê¿Abd al-Malik (b. 86/705, r. 126/744), the son of a slave girl who was allegedly a descendant of the last Sasanian king. As the sources make clear, there is some confusion about the name of this woman, as well as whether she was a daughter or granddaughter of Yazdgerd III. Based on YazÄ«d’s dates, the latter seems much more likely. Indeed, if there is any reality to the story, it may be that her father was Yazdgerd’s son FÄ«rÅ«z, who is known to have tried to rally support from the Tang in China and to have sparked anti-Muslim revolts in eastern Iran.⁴⁵ Overall, however, the motivation for these Umayyad legends appears to be much the same as those about Zayn al-ʿĀbidÄ«n: that is, the glorification of a high-status Muslim by connecting him with a high-status non-Muslim mother. An early report about YazÄ«d’s mother, for example, comes from KhalÄ«fa b. Khayyāṭ (d. 240/854):

In this year [126/743–44], YazÄ«d b. al-WalÄ«d received the oath of allegiance in the beginning of the month of Rajab. His mother was the daughter of Yazdgerd son of Khusraw.⁴⁶

Al-Yaʿqūbī, writing slightly later, supplies the name of this woman:

Then there reigned YazÄ«d b. al-WalÄ«d b. Ê¿Abd al-Malik, and his mother was ShāhfarÄ«d, the daughter of FÄ«rÅ«z son of Khusraw.⁴⁷

Meanwhile, al-Masʿūdī (d. 345/956) specifies that this woman was of servile status:

The mother [of YazÄ«d] was a slave girl (umm walad) called Shāhfirand [sic], daughter of FÄ«rÅ«z son of Khusraw.⁴⁸

The most elaborate traditions about this Sasanian princess comes from al-Ṭabarī, who mentions her at two separate points in his history. In the first, he writes:

[Narrated from an old man from the Banū Sadūs from Ḥamza b. Bīḍ]

Qutayba [b. Muslim, the conqueror of Central Asia, d. 96/715] acquired in Khurāsān – in Soghdia – a slave girl (jāriya) who was one of the descendants of Yazdgerd. He said, “Do you think that the son of this [girl] will be hajÄ«n [that is, a “mongrel,” meaning a person of mixed Arab and non-Arab ancestry]?” They said, “Yes, but he will be hajÄ«n by virtue of his father [that is, the girl’s Persian ancestry was so pure that it was the inferior blood of her Arab master that risked spoiling her children’s lineage]. He sent her to al-Ḥajjāj [b. YÅ«suf, the governor of Iraq, d. 95/714], and al-Ḥajjāj sent her to [the caliph] al-WalÄ«d. She bore him YazÄ«d b. al-WalÄ«d.”⁴⁹

In the second, al-Ṭabarī writes:

The mother of Yazīd was a concubine (umm walad). Her name was Shāh Āfrīd the daughter of Fīrūz son of Yazdgerd son of Shahriyār son of Khusraw. Yazīd used to say [the following lines of poetry]:

I am the son of kisrā [the Arabic title of the Persian kings, that is, Khusraw]; my father is Marwān [progenitor of the ruling branch of the Umayyad family]

My grandfather is a qayá¹£ar [the Arabic title of the Roman emperor, that is, Caesar]; the other a khāqān [the ruler of the Turks]⁵⁰

YazÄ«d ruled at a time of extreme factionalism in the Umayyad caliphate. His enemies derided him for his mother’s servile status in contrast to the “pure” blood of his rivals, including his cousin al-Ḥakam b. al-WalÄ«d b. YazÄ«d, whose mother was a freeborn Arabian tribeswoman.⁵¹ To boast of having a Sasanian mother was to overcome this deficit, indeed, to exceed his rivals by claiming to be descended from one of the greatest families of the pre-Islamic age. This boasting extended to his alleged descent from other pre-Islamic kings, though this seems even more doubtful than his alleged Sasanian blood.

The symbolism of having a Sasanian princess as a mother – an idea first hatched in Muslim circles – soon spread to non-Muslim communities. This is a fascinating development, one that demonstrates the extent to which Muslim understandings of maternal lineage and legitimation were quickly adopted by the subject populations of the empire. For example, according to Jewish legends of the period, the first exilarch (that is, the secular leader of the Jewish community of Babylon, who claimed to be descended from King David) to rule after the Arab conquests was a man named Bustanay ben Kafnay (ca. 618–70).⁵² Information about this Bustanay differs from one source to the next, but all claim that his concubine was a daughter of the Sasanian king, a woman sometimes known as “Azdādwār,” who found her way into captivity during the reign of the caliph Ê¿Umar b. al-Khaṭṭāb. Ê¿Umar reportedly awarded her as a slave to Bustanay, whom he appointed as exilarch. Together, they had three sons.

The following version of the story comes from a Judeo-Arabic document discovered in the Cairo Geniza (British Library OR 5552.4–4a), which was studied by George Margoliouth many years ago:

The sultan at the time was Ê¿Umar b. al-Khaṭṭāb … He accorded [to Bustanay] the chieftaincy [that is, the role of exilarch] and granted him permission to be present at his dÄ«wān [that is, to be a member of his court]. He also gave him a daughter of kisrā, for Ê¿Umar b. al-Khaṭṭāb had sacked a city of the kisrā and killed him. And Bustanay of the Davidic line married the girl given to him by Ê¿Umar b. al-Khaṭṭāb, and there were children of the marriage. But he had not given her her freedom either through neglect or pride. And all of Bustany’s children were from this woman, for he married none other, and he had no other children. Amongst his descendants are Anan, and besides, Boaz, the sons of Zakkai, the exilarch in Baghdad, and a few persons in al-Andalus. All these trace their descent from Bustanay.⁵³

Like the other stories of Sasanian princesses, this one is clearly legendary. Yet it took on great significance among the geʾonim – the heads of the Talmudic academies of medieval Babylon – especially during the tenth century. The story became a test-case for determining the status of non-Jewish slaves who bore children for Jewish masters, as well as the legal status of these offspring once they became adults. It was also a point of contention between the exilarchs, who claimed descent from Bustanay and his royal Persian slave, and the geʾonim in the academy at Pumbedita, who harped on the story of the servile, infidel mother in order to undermine their rivals, the exilarchs, whom they challenged for leadership of the Jewish community. Once again, we see how stories of high-status non-Muslim (in this context, non-Jewish) concubines became a way of expressing important ideas about lineage, genealogy, and power in the early Islamic empire.

…

This essay has attempted to cast light on the neglected phenomenon of high-status non-Muslim women who became wives and concubines of high- status Muslims during the early centuries after the conquests. As historical artifacts of the seventh, eighth, and ninth centuries, they underscore the complicated relationship of the Muslim conquerors to the indigenous elites of the territories they ruled. These elites were sometimes partners in the task of building a new Islamic empire and legitimizing Muslim rule. But they were also sometimes symbols of an old order that needed to be defeated and overcome. This historical reality is also reflected in the literary traditions that sprang up around these women. Indeed, many seem to have been entirely fictional, having become characters in stories that were, in turn, used to promote the genealogical claims of elite Muslims in the past, not to mention their descendants in later periods. In this sense, many of the women discussed in this essay float tantalizingly between the realms of fact and fiction, often creations of later periods and reflecting later debates. Through it all, they remind us that women, far from being peripheral actors in the formation of early Muslim society, were key determinants of social status, political power, and prestige.⁵⁴