Introduction

In her feminist polemic al-Sufūr wa-l-ḥijāb (Unveiling and Veiling, 1346/1928), the Lebanese Druze activist Naẓīra Zayn al-Dīn pauses to recount a list of discriminatory measures imposed on non-Muslim residents of the Muslim polity. Her source is the Damascene Ḥanafī jurisconsult Ibn ʿĀbidīn (d. 1258/1842).³ Among other ghiyār (distinctions) pertaining to dhimmīs, she notes the prohibitions on building lofty residences, ownership of slaves and slave-concubines, and the wearing of expensive garments.⁴ The relevance of these remarks to her larger argument about women quickly becomes apparent. “These sartorial restrictions (quyūd al-malābis)… [facilitate] the identification of non-Muslims…so people can treat them accordingly. This is naught but an innovation (bidʿa), like the innovation of veiling women—those poor ones subject to humiliation and abasement (ṣaghār).”⁵ According to Zayn al-Dīn, the ḥijāb is a pernicious custom (much like the ghiyār), one that degrades women and has no basis in the Qurʾān or Sunna—a critique shared by other reformists of the period.⁶ By “ḥijāb,” she refers to the veiling of the face and the accompanying system of sexual segregation, the standard referent of the term in the early fourteenth/twentieth century Arabophone world.⁷ The rationale for the practice, she argues, is evident in the tafsīr (Qurʾān commentary) literature, where one finds that it distinguishes free women from enslaved believing women.⁸ “This rationale (qaṣd) is confirmed by … [the account in which] a slave woman [whose dress] resembled that of a free woman passed by ʿUmar b. al-Khaṭṭāb…who reproached her, exclaiming ‘O base one (yā lukāʿ), are you trying to imitate free women?’”⁹

Zayn al-Dīn was not the first person to invoke the encounter between the Caliph ʿUmar and a slave woman to make a point about status distinction and sartorial practices in Islamic law. In this case, as miriam cooke explains, Zayn al-Dīn deploys it to argue that “since slavery was a practice of the past, the instrumental function of the veil no longer pertained.”¹⁰ If slavery no longer exists, in other words, the rationale for and the obligation of veiling disappear. The ʿUmar report—and its numerous variants—was subject to centuries of debate by jurists. The common core of the narrative is a confrontation between the Caliph and another person (or persons) about the dress of a slave (or slaves) in Medina, dress that visibly subverts the distinction between free and enslaved women.

In this article, I trace the various transmissions of the ʿUmar report and aspects of its citation in a range of literary genres—without adopting a position on its historicity—to shed light on the question of the veiling of slaves in Islamic law.¹¹ I hope to make a number of key interventions. Firstly, I attempt to characterize the relevant qurʾānic data, drawing on the tafsīr literature; secondly, I map juristic opinion on the question of veiling, highlighting the increased emphasis on the fear of fitna (sexual temptation) in early post-formative (i.e. third-early fifth/ninth-early eleventh centuries) discussions of sartorial regulation. Finally, I highlight juristic disagreements over modesty norms as they apply to slave women. In this context, I make some brief observations about opinions in later periods, down to the thirteenth/nineteenth centuries. My larger argument relates to the study of Islamic norms in their Ancient and Late Antique Near Eastern contexts, building on existing scholarship on veiling practices.¹² For example, Hina Azam has written about the distinction between “proprietary” and “theocentric” systems of sexual ethics, an important conceptual framework that will undergird the analysis below.¹³ Given the importance of this framework, I shall summarize some of its key features as they relate to my argument.

In systems of proprietary sexual ethics, Azam informs us, a woman’s sexuality is a commodity subject to sale and purchase. The “theft” of that sexuality through rape, for example, is punished like other property crimes, including financial compensation to the owner (in the case of a virgin, her father), among other measures. In this view, God or the gods have little concern with sexuality, and illicit sexual relations are less an affront to the deity than an infringement of someone’s entitlement to financial reward for the sale of sexual access.¹⁴ By contrast, the qurʾānic intervention is decidedly theocentric, although it retains elements of the proprietary ethic. Illicit sex, as in the Hebrew Bible, is a moral crime, a sin against an all-knowing and ever-watchful God. Those guilty of it are subject to punishment by scourging. Financial compensation plays a secondary role or no role at all in this scheme, according to Muslim jurists.¹⁵ In proprietary terms, however, the Qurʾān retains the practice of mahr (dower, e.g. Q. 4:24), which was conceptualized by jurists as the “vulva’s price” (thaman al-buḍʿ), clearly a reminder of earlier legal systems. Modern scholarship has debated the extent to which marriage resembles a contract of sale in Islamic law, but the term thaman al-buḍʿ does help explain key features of juristic reasoning on marriage, as Kecia Ali has argued.¹⁶ In this article I propose that veiling practices should be read in the context of competing systems of sexual ethics. I will argue that veiling, along with other elements of Islamic modesty norms, should be seen as a combination of proprietary and theocentric values. The distinctions between free and enslaved Muslim women—noted and sometimes critiqued by classical Muslim jurists—highlight these tensions. I should note that I foreground a particular aspect of Azam’s argument in my analysis. For the purposes of this article, the proprietary ethic is not so much about the ownership claims of masters or mistresses over their enslaved persons, and the corresponding control of their slaves’ (commodified) sexuality. In this article, I am more interested in the implications of the commodification of slaves and their sexual capacity for modesty norms themselves. In other words, how and to what extent does the fact that slaves are considered property nuance the theocentric dimensions of veiling?

Variants of the ʿUmar Report

To better appreciate the implications of the ʿUmar report for the veiling of slaves, and the development of Islamic modesty norms more generally, I document and translate its variants below. I present these variants in chronological order, except when there are multiple transmissions of an earlier work, e.g. in the Muwaṭṭaʾ. Though the provenance of some of the texts below has been disputed, I do not discuss arguments for or against their authenticity.¹⁷ Even if they represent later texts back-projected onto more illustrious transmitters, this does not significantly undermine my argument, which is a cumulative one. For illustrative purposes, I attempt to characterize the transmitters who lived in the first two Islamic centuries in terms of garrison town. Frequently, such characterizations are less than adequate, as in the case of Ibn Shihāb al-Zuhrī (d. 124/742), whose rich and varied career as a traditionist, courtier and sometime tutor to Umayyad princes is not fully captured by his “Medinan” affiliation. Nonetheless, ḥadīth continued to be transmitted locally until the later years of the second/eighth century. In cases where the original texts provide isnāds, I present them here. For the sake of comprehensiveness, I include a number of citations of the report found outside of the usual ḥadīth and fiqh works in the first four Islamic centuries. This is because the transmitters of these texts cannot be assumed to be dependent on the earlier ḥadīth sources, which had not become “canonical.” Similarly, I present a number of later variants without isnāds when they feature interesting vocabulary or additional details not found in earlier attestations. In such cases, I include only the earliest attestation of these variants.

1. 1.Ibn Bukayr (d. 231/845) – Mālik b. Anas (Medinan, d. 179/795): ʿUmar b. al-Khaṭṭāb saw a slave woman belonging to ʿUbayd Allāh b. ʿUmar b. al-Khaṭṭāb mingling (tajūsu) with people and dressed in the apparel of free women (tahayyaʾat bi-hayʾat al-ḥarāʾir). He met with his daughter, Ḥafṣa, and asked, “Did I just see your brother’s slave woman going about [so dressed]?” ʿUmar b. al-Khaṭṭāb condemned this [behavior].¹⁸
2. 2.Yaḥyā b. Yaḥyā al-Laythī (d. 234/848) – Mālik b. Anas: ʿUmar b. al-Khaṭṭāb saw a slave woman belonging to ʿAbd Allāh b. ʿUmar b. al-Khaṭṭāb…ʿUmar condemned this [behavior].¹⁹
3. 3.Abū Muṣʿab al-Zuhrī (d. 242/856) – Mālik b. Anas: identical to report 1 above, except for the final mention of ‘Ibn al-Khaṭṭāb.’²⁰
4. 4.Yūsuf b. Abī Yūsuf (Kufan, d. c. 200/815–6) – al-Qāḍī Abū Yūsuf (Kufan, d. 182/789) – Abū Ḥanīfa (Kufan, d. 150/767) – Ḥammād b. Abī Sulaymān (Kufan, d. 120/738) – Ibrāhīm al-Nakhaʿī (Kufan, d. 96/714): Slave women are not obligated to veil (laysa ʿalā al-īmāʾ qināʿ) for prayer or otherwise. [People] used to consider it reprehensible (yukrah) that slave women veil, making them resemble free women.²¹
5. 5.Muḥammad b. al-Ḥasan (Kufan, d. 189/804) – Abū Ḥanīfa – Ḥammād b. Abī Sulaymān – Ibrāhīm al-Nakhaʿī: ʿUmar b. al-Khaṭṭāb—may God be pleased with him—used to strike slave women for veiling, saying, “Do not imitate free women (lā tatashabbahīna bi-l-ḥarāʾir)!”²²
6. 6.Yaḥyā b. Sallām (Basran, d. 200/815) – Saʿīd b. Bashīr (Basran, d. 168/784–5) – Qatāda b. Diʿāma (Basran, d. 117/735) – Anas b. Mālik (Basran, d. 93/712): ʿUmar b. al-Khaṭṭāb saw a slave woman wearing a veil and struck her with his switch. Saʿīd added, “He poked her with his switch, commanding ‘uncover your head!’” Saʿīd added, “Do not imitate free women.”²³
7. 7.ʿAbd al-Razzāq al-Ṣanʿānī (d. 211/826) – Ibn Jurayj (Meccan, d. 150/767) – ʿAṭāʾ b. Abī Rabāḥ (Meccan, d. 114/732): ʿUmar b. al-Khaṭṭāb used to forbid slave women from wearing robes (jalābīb) and imitating free women.²⁴
8. 8.ʿAbd al-Razzāq al-Ṣanʿānī – Ibn Jurayj: I was informed that ʿUmar b. al-Khaṭṭāb struck ʿAqīla, a slave woman belonging to Abū Mūsā al-Ashʿarī, for wearing a robe.²⁵
9. 9.ʿAbd al-Razzāq al-Ṣanʿānī – Maʿmar b. Rāshid (Basran, d. 153/770) – Ayyūb al-Sakhtiyānī (Basran, d. 131/748) – Nāfiʿ (Medinan, d. 117/735): ʿUmar saw a slave woman leave Ḥafṣa’s house—or perhaps it was a house of one of the wives of the Prophet—dressed in a robe. ʿUmar entered the house, asking “Who is this slave woman?” They answered, “A slave woman of ours”—or they said, “A slave woman belonging to the family of so-and-so”—whereupon ʿUmar became angry, asking, “Do you send your slave women out in all their finery, tempting people?”²⁶
10. 10.ʿAbd al-Razzāq al-Ṣanʿānī – Ibn Jurayj – Nāfiʿ – Ṣafiyya bt. Abī ʿUbayd (Medinan, d. after 73/692–3): When ʿUmar was preaching he saw a slave woman leaving the house of Ḥafṣa dressed as a free woman. When ʿUmar [finished preaching], he met with Ḥafṣa and asked, “Who is this woman who left your house, mingling with the menfolk?” She answered, “A slave woman belonging to [your son] ʿAbd al-Raḥmān.” [ʿUmar asked] “What led you to dress your brother’s slave woman in the dress of free women? I entered thinking she was a free woman, and wished to punish her.”²⁷
11. 11.ʿAbd al-Razzāq al-Ṣanʿānī – Maʿmar b. Rāshid – Qatāda b. Diʿāma – Anas b. Mālik: ʿUmar struck a slave woman belonging to the family of Anas for wearing a veil, saying, “Uncover your head! Do not imitate free women!”²⁸
12. 12.ʿAbd al-Razzāq al-Ṣanʿānī – Sufyān b. ʿUyayna (Kufan, d. 198/813–4) – ʿAmr b. Dīnār (Meccan, d. 126/744) – al-Ḥasan b. Muḥammad b. al-Ḥanafiyya (Medinan, d. 100/718–9): ʿUmar b. al-Khaṭṭāb used to forbid slave women from wearing robes.²⁹
13. 13.Saʿīd b. Manṣūr (d. 227/841–2) – Hushaym b. Bashīr (Baghdadian, d. 188/803) – Khālid b. Mihrān al-Ḥadhdhāʾ (Basran d. 141/758) – Abū Qilāba ʿAbd Allāh b. Zayd (Basran, d. 104/722–3): During his caliphate, ʿUmar b. al-Khaṭṭāb would not permit any slave women to go about veiled. He would say, “The veil is only for free women, so that they will not be harassed.”³⁰
14. 14.Ibn Abī Shayba (d. 235/849) – Wakīʿ b. al-Jarrāḥ (Kufan, d. 197/812–3) – Shuʿba b. al-Ḥajjāj (Basran, d. 160/777) – Qatāda b. Diʿāma – Anas b. Mālik: When ʿUmar saw my slave woman wearing a veil, he struck her, saying, “Do not imitate free women!”³¹
15. 15.Ibn Abī Shayba – ʿAbd al-Aʿlā b. ʿAbd al-Aʿlā (Basran, d. 189/805) – Maʿmar b. Rāshid – Ibn Shihāb al-Zuhrī (Medinan, d. 124/742) – Anas b. Mālik: When ʿUmar saw a slave woman wearing a veil he struck her, saying “Do not imitate free women!”³²
16. 16.Ibn Abī Shayba – ʿAlī b. Mushir (Kufan, d. 189/804–5) – al-Mukhtār b. Fulful (Kufan, d. c. 140/757–8) – Anas b. Mālik: ʿUmar b. al-Khaṭṭāb came across a slave woman whom he knew was owned by one of the Emigrants or Helpers. She veiled herself with [a loose end of] her robe. He asked her, “Have you been manumitted?” She answered in the negative. He asked, “Then why [are you wearing] a robe? Remove it from your head! The robe is only for free believing women!” When she dawdled (talakkaʾat), ʿUmar struck her with his switch until she dropped [the loose end of] her robe from her head.³³
17. 17.Ibn Abī Shayba – Hushaym b. Bashīr – Khālid b. Mihrān al-Ḥadhdhāʾ – Abū Qilāba ʿAbd Allāh b. Zayd: During his caliphate, ʿUmar b. al-Khaṭṭāb would not permit any slave women to veil. He would say, “The veil is only for free women, so that they will not be harassed.”³⁴
18. 18.ʿAbd al-Malik b. Ḥabīb (d. 238/852) – Anas b. Mālik: ʿUmar b. al-Khaṭṭāb saw a slave woman belonging to one of his companions wearing a head covering (mukhtamira). He asked her, “Has your master manumitted you?” She answered in the negative. He asked, “Then why are you wearing a robe? Put it aside (ḍaʿīh)!” When she refused, he hit her with his switch until she complied, proclaiming, “Do not imitate free women ever again!”³⁵
19. 19.ʿAbd al-Malik b. Ḥabīb – Anas b. Mālik: ʿUmar came to his family one day and noticed that there was a woman in his house wearing a robe. When he saw her, he left. When he returned and saw her again, he left (inṣaraf) again. He did this several times, until she departed. He then entered, asking, “Who is this who has been bothering us today?” His wife responded, “What do you have to do with her? She is so-and-so’s slave woman.” When [ʿUmar] left he proclaimed to the people, “Do not let slave women resemble their mistresses! Do not dress them in robes such that they resemble free, chaste women (al-ḥarāʾir wa-l-muḥṣanāt)! God the [most] exalted said (Q. 33:59) “…O Prophet, tell your wives, your daughters and believing women to cover themselves with their robes.”³⁶
20. 20.Hūd b. Muḥakkam (d. c. 290/903): ʿUmar b. al-Khaṭṭāb saw a slave woman wearing a veil, and he raised his switch against her, saying, “Uncover your head! Do not imitate free women!” They [viz. his sources among the learned] added that Anas b. Mālik reported, “The slave women of ʿUmar used to serve us with their heads uncovered, their breasts knocking together and their anklets exposed.”³⁷
21. 21.Al-Māturīdī (d. 333/944): It is reported that a slave woman passed by ʿUmar b. al-Khaṭṭāb—may God be pleased with him—wearing a veil. He struck her with his switch and said, “Remove your veil (ikshifī qināʿik) and do not imitate free women!”³⁸
22. 22.Al-Qāḍī al-Nuʿmān (d. 363/974) – Jaʿfar b. Muḥammad al-Ṣādiq (Medinan, d. 148/765)—upon whom be peace—was asked if slave women are required to veil during prayer. He answered, saying, “No, and when my father—may God be pleased with him—saw a slave woman praying in a veil, he would strike her, saying, ‘Do not imitate free women, you moron (lukaʿu)!’ This was [so] that free women would be distinguishable from slave women.”³⁹
23. 23.Al-Jaṣṣāṣ (d. 370/981): It is reported that ʿUmar used to strike slave women and order them to uncover their heads and not to imitate free women.⁴⁰
24. 24.Ibn Abī Zamanīn (d. 399/1008) – Yaḥyā b. Sallām – Saʿīd b. Bashīr and ʿUthmān b. Miqsam al-Burrī (Basran, d. 163/779–80) – Qatāda b. Diʿāma – Anas b. Mālik: ʿUmar saw a slave woman wearing a veil. He struck her with his switch—according to the version of Saʿīd—and in the version of ʿUthmān, “[ʿUmar] raised his switch against her and said, ‘Uncover your head!’” In the version of Saʿīd he said, “Do not imitate free women!”⁴¹
25. 25.Ibn Abī Zamanīn – Yaḥyā b. Sallām – Saʿīd b. Bashīr – Qatāda b. Diʿāma – Anas b. Mālik: When ʿUmar b. al-Khaṭṭāb saw a slave woman wearing a veil, he raised his switch against her, commanding, “Uncover your head and do not imitate free women!”⁴²
26. 26.Al-Qāḍī ʿAbd al-Wahhāb al-Mālikī (d. 422/1035): It is reported that ʿUmar b. al-Khaṭṭāb—may God be pleased with him—used to strike slave women if he saw them wearing a waist-cloth (izār), saying, “Do not imitate free women!” He rebuked his son, saying, “Did I not tell you that your slave woman was walking about in a waist-cloth imitating free women? If I come across her [doing so] I will administer a painful beating (la-awjaʿtuhā ḍarb^an).”⁴³
27. 27.Abū Isḥāq al-Thaʿlabī (d. 427/1035) – Anas b. Mālik: A veiled slave woman passed by ʿUmar b. al-Khaṭṭāb, who raised his switch against her, saying, “O base one (yā lukāʿ), are you trying to imitate free women? Put aside your veil!”⁴⁴
28. 28.Al-Bayhaqī (d. 458/1066) – Abū Saʿīd b. Abī ʿAmr (d. 421/1030)⁴⁵ – Abū l-ʿAbbās Muḥammad b. Yaʿqūb (d. 277/890–1) – Aḥmad b. ʿAbd al-Ḥamīd al-Ḥārithī (d. 269/883) – Abū Usāma Ḥammād b. Usāma (Kufan, d. 201/817) – al-Walīd b. Kathīr al-Makhzūmī (Medinan, d. 151/768–9) – Nāfiʿ – Ṣafiyya bt. Abī ʿUbayd: A robed and veiled woman was walking about [in Medina] and ʿUmar—may God be pleased with him—asked who she was. He was told that she was the slave woman of one of his sons. He sent for Ḥafṣa—may God be pleased with her—asking, “What made you veil and robe this slave woman, causing her to resemble free women, such that I thought of accosting her (an aqaʿu bihā), thinking she was free? Do not have your slave women resemble free women!”⁴⁶
29. 29.Al-Sarakhsī (d. 483/1090): ʿUmar—may God be pleased with him—would raise his switch against any slave woman who walked about veiled, saying, “Put your head-cover aside, oh odious one (yā difār)! Are you trying to resemble free women?”⁴⁷

View Table

Analysis of the Variants

The ʿUmar report achieved wide dissemination quickly, as the isnāds listed above suggest. A number of the variants are poorly transmitted by the standards of (later) classical ḥadīth criticism, with noticeable gaps in their routes of transmission (i.e. they are marāsīl). Nonetheless, the basic integrity of the ʿUmar report has been accepted by several exacting Muslim critics, including al-Bayhaqī (d. 458/1066), Ibn Ḥazm (d. 456/1064) and, more recently, Nāṣir al-Dīn al-Albānī (d. 1420/1999). The latter two, interestingly, accept the authenticity of the report while rejecting its probativeness.⁴⁸ These gaps in transmission already appear in our earliest attestation of the ʿUmar report, the Muwaṭṭaʾ. That multiple recensions of the Muwaṭṭaʾ all convey what is more or less the same version of the report speaks strongly to its inclusion in one or more “editions” published by the work’s reputed author-compiler, Mālik b. Anas. The ʿUmar report also appears in two early independent Kufan versions, both transmitted by students of Abū Ḥanīfa. Basran and Meccan isnāds also make an early appearance, in the late second Islamic century.

All of the reports describe an encounter between ʿUmar and one or more persons and involve an element of reproach. There is one exception, the report attributed by al-Qāḍī al-Nuʿmān to Muḥammad al-Bāqir (d. 114/732) via his son Jaʿfar al-Ṣādiq, the fifth and sixth imāms of the Twelver-Shiʿa and Ismāʿīlīs, respectively. In this account, al-Bāqir is equally protective of the distinction between free and enslaved women, employing the same insult (lukaʿu) to enforce it. This appears to be the earliest attribution of the report to someone other than ʿUmar, though one can safely assume that al-Qāḍī al-Nuʿmān drew it from an already existing Kufan Shiʿi corpus.⁴⁹ In most versions of the main narrative, however, it is ʿUmar who confronts the slave woman. In some variants, he reproaches Ḥafṣa, one of the Prophet’s wives, or the people of Medina generally. The question of the slave woman’s owner also gave rise to conflicting answers, which may stem from the notion found in several versions of the report that ʿUmar habitually confronted slave women over their problematic dress. Some accounts purport to convey information about the same episode, however. In these cases the narrators differ as to which of ʿUmar’s children is the guilty party: it is either ʿAbd Allāh, ʿUbayd Allāh, ʿAbd al-Raḥmān, or Ḥafṣa. In the other reports, Anas b. Mālik or someone in his household, Abū Mūsā al-Ashʿarī, or one of the wives of the Prophet is the culprit. This confusion over identity also pertains to the slave woman herself. In most variants, she is anonymous, in one she is identified as ʿĀqila. The variants also diverge on the nature of the confrontation. ʿUmar either struck her (in most versions) or he did not; and he either reprimanded her directly, or criticized her owner (or somebody who knew them). The slave woman’s response, and the insults used against her, vary accordingly. In one instance she is slow to comply, provoking ʿUmar to hit her with his switch.

While the basic structure of the narrative is common to most of the accounts, in some variants ʿUmar reproaches the slave woman for covering her head, in others for wearing a robe. It is notable that it is the slave woman’s dressing as a free woman that is the source of temptation. The distinction suggests that sexual desire was attached to status—as though free women are inherently desirable and slave women are not. One version can be construed as making ʿUmar’s sexual needs explicit (an aqaʿu bihā, i.e., “Such that I wished to have sex with her”). The underlying proprietary ethic is clear: ʿUmar upholds the distinction between free and enslaved women, implying that the latter are subject to different modesty norms because of their status as property. This status should be clearly visible, and any attempts to subvert it should be treated with hostility.

It is only in isnād-less, later versions of the report that one finds colorful insults, including (in chronological order of their appearance) lukaʿu, lukāʿ and difār (versions 22, 27 and 29, respectively). These insults tend to appear in works of fiqh and tafsīr rather than compilations of ḥadīth, and one suspects that they represent the elaboration of a copyist rather than a transmitter. There is, of course, the possibility that I have not recorded all of the major variants, or that some of them are lost to us forever or remain in manuscript.⁵⁰ We are on firmer ground with speculation on the origins of the version adduced by al-Qāḍī ʿAbd al-Wahhāb, certainly a later interpolation. In this version, ʿUmar upbraids slave women for wearing an izār. This narrative makes no sense at all, as this item of clothing was worn by persons of all backgrounds, whether free or enslaved. Earlier authorites, such as the Mālikī Ibn Abī Zayd, explicitly mention slaves who wear the izār.⁵¹ In short, reproaching a slave woman for wearing an izār is meaningless.

Two transmissions of the ʿUmar report foreground what is only hinted at in the other versions: the veil/distinguishing dress item is for free women alone, “So that they will not be subject to harassment.” It is this statement (what classical uṣūlīs call mafhūm al-mukhālafa, counterimplication⁵²) that proved most problematic for some later jurists, as we shall see.

Veiling, Proprietary Sexual Ethics, and the Formation of Islamic Norms of Modesty

There are (at least) two approaches to the question of the influence of Near Eastern norms on the formation of Islamic law. One view posits the crystallisation of Islamic law in the wider Near Eastern context following the conquests.⁵³ The other approach emphasizes—often to very different effect—the ways in which Arabian custom already reflected these influences prior to the career of the Prophet.⁵⁴ These views are not mutually exclusive. This article is not the place to elaborate on the ways in which Islamic law was shaped by its Late Ancient Near Eastern context, or indeed to reflect more generally on the category of “influence”; others have already done this.⁵⁵

It is worth pointing out, however, that there is some (admittedly fragmentary) evidence indicating that the practice of veiling was by no means novel in Arabia in the lifetime of the Prophet. The Church Father Tertullian (d. 240 ce) mentions that the women of Arabia “cover not only the head, but the face also…with one eye free, to enjoy rather half the light than to prostitute the entire face.”⁵⁶ This practice—veiling the face with the exception of one eye—is endorsed in a number of early Qurʾān commentaries.⁵⁷ Similarly, other witnesses to pre-Islamic Arabian norms attest to the practice of veiling. In his massive study of women in jāhilī (pre-Islamic) poetry, Aḥmad al-Ḥūfī concludes that the evidence for veiling is extensive.⁵⁸ It includes the famous Tāʾiyya (poem rhyming in the letter “tāʾ”) of al-Shanfarā (d. c. 525 ce).⁵⁹ Since Muslim authors typically attributed loose sexual norms to jāhilī pagans, there were few incentives for them to “remember” veiling.⁶⁰ It is reasonable to suggest, therefore, that like other inhabitants of the Near East in Late Antiquity, pre-Islamic Arabians were familiar with veiling and at times practiced it.

The Qurʾān makes three major sartorial interventions. In Q. 7:31, “children of Adam” are encouraged to dress appropriately “at every place of prayer,” a verse most commentators understood as denouncing Arab pagans, who reportedly circumambulated the Kaʿba in a state of ritual nudity.⁶¹ Q. 24:31 instructs believing women

…not to show their beauty, except that which is apparent thereof (illā mā ẓahara minhā); let them conceal their bosom with their head-coverings (khumūrihinna). Let them not show their beauty, except to their husbands, their fathers, their fathers-in-law, their children, their step-children, their brothers, their brothers’ and sisters’ sons, their women (nisāʾihinna), their slave-folk (mā malakat aymānuhunna), those male attendants lacking sexual desire (ghayri ūlī l-irba min al-rijāli), and children not yet alive to it…

These instructions include both theocentric and proprietary elements. Firstly, husbands and relatives unlawful to marry are excepted from the general rule of concealment, along with “their” women (i.e. fellow Muslim women)⁶² and others who lack sexual desire (ghayri ūlī l-irba). There is a correlation between the implied absence of (illicit) temptation and the corresponding relaxation of modesty norms. The lowering of the gaze, commended in this and the previous verse, is said to be “purer” for believers, who are reminded (24:30) that “God is aware of all that they do.” Elsewhere (Q. 40:19), God “knows the treacheries of stolen glances (khāʾinat al-aʿyun) and what hearts conceal.” These references to God’s scrutiny and judgment are all theocentric. On the other hand, Q. 24:31 explicitly excludes slaves (mā malakat aymānuhunna), both male and female slaves, according to the majority of early authorities.⁶³ These elements of qurʾānic modesty norms are clearly “proprietary;” slaves are excepted from a series of ostensibly theocentric norms that apply to believers across the board, as though the normal calculus of sexual desire does not apply to them. Though slaves are not the addressees of this verse, their exclusion from the general rule has clear implications for modesty norms. It is in this sense that qurʾānic sexual ethics—particularly with respect to veiling—represents a combination of theocentric and proprietary values.

The third sartorial intervention (Q. 33:59) was understood by commentators to enforce the distinction between the modesty norms expected of free and enslaved persons. The Prophet is commanded to “Tell your wives, your daughters and believing women generally (nisāʾ al-muʾminīn) to cover themselves in their robes (yudnīna ʿalayhinna min jalābībihinna); that is better, that they be known and not harassed (yuʾdhayna)…” Many commentators link this verse to episodes in which free women were mistaken for slaves and were subject to harassment,⁶⁴ usually during nocturnal excursions to the toilet,⁶⁵ still a concern in some parts of the world today.⁶⁶ Free believing women are thus instructed to “cover themselves in their robes” to distinguish themselves from slaves and to preclude unwanted attention. This interpretation—requiring free women to distinguish themselves from slaves—is widely attested, even though the verse categorically commands “believing women” to veil. Q. 33:59, according to its most widely-attested interpretation, is thus the supreme example of proprietary attitudes in Islamic modesty regulations: it enforces the free/slave binary.

Middle Assyrian law also makes a sartorial distinction between free and enslaved women. I should note that I am wary of asserting the unqualified “influence” of Assyrian law on Islamic law, not least because there are no obvious parallels to Islamic slave concubinage in the Late Antique Middle East (and thus, insufficient evidence of continuity).⁶⁷ In any case, in Middle Assyrian law:

Wives of a man…who go out into the main thoroughfare [shall not] have their heads [bare]…[they] shall be veiled…a concubine…is to be veiled…slave women shall not be veiled…if a man intends to veil his concubine, he shall assemble five or six companions and he shall veil her in their presence…⁶⁸

A number of early Muslim jurists reportedly supported the veiling of slave concubines (ummahāt al-awlād), or married slaves, in contradistinction to other enslaved women. But many early Muslim jurists limit the obligation (or preference) to ritual prayer. ʿAṭāʾ b. Abī Rabāḥ stipulated a khimār (head covering) for slave women married to other slaves, or to free men.⁶⁹ Al-Ḥasan al-Baṣrī (d. 110/728) upheld the same view for married slaves and slave concubines.⁷⁰ According to Saḥnūn (d. 240/854) Mālik b. Anas insisted that ummahāt al-awlād veil for prayer.⁷¹ Similar doctrines are found in contemporaneous Mālikī texts, including the recently edited Mukhtaṣar of Abū Muṣʿab al-Zuhrī, which singles out the umm al-walad among slave women for veiling. The author, an important transmitter of the Muwaṭṭaʾ, remarks that it is reprehensible (yukrah) for a slave concubine to “uncover her head in prayer,” a view shared by other fourth/tenth century Mālikīs.⁷² Aḥmad b. Ḥanbal (d. 241/855) reportedly preferred that slave women pray in a khimār (head covering), though prayer without it remains valid. However, once a female slave has been emancipated (even if in mid-prayer), veiling becomes mandatory.⁷³ Ibn Abī Shayba attributes contrasting views to earlier jurists, some of whom held that a slave woman, presumably including slave concubines, ought to “pray as she appears in public (tuṣallī al-ama kamā takhruj),” meaning unveiled.⁷⁴ Other jurists are more explicit: “a slave woman is not required to veil [in prayer] even if she bears her master a child.”⁷⁵

Most proto-Sunni jurists did not distinguish between slave concubines, married slaves and others for purposes of veiling, which they held to be neither recommended (mustaḥḥab) nor obligatory.⁷⁶ This view was shared by early Twelver-Shiʿa and Ismāʿīlīs, as well as Zaydīs.⁷⁷ We do know from one early non-Muslim account that veiling was the norm in Baghdad shortly after that city’s foundation. Though sensitive to the legal distinctions between slaves and free persons (both are obligated to pray), our source is silent on the question of whether slaves veiled or not.⁷⁸ There is, moreover, no suggestion in the sources of any equivalent to the Assyrian legal procedure for veiling slave concubines.

There is a difference between holding an act to be non-obligatory—as most early jurists did for the veiling of female slaves—and prohibiting it. The distance between the two positions, one might say, is the length of ʿUmar’s switch. Since this distinction seems to be lost already in most of our earliest sources, reviewed above, I shall explore the possible rationale for ʿUmar’s reported actions by drawing on recent scholarship.

Eli Alshech has argued that the assertion that the purpose of veiling (and related practices) was to minimize sexual temptation is not persuasive.⁷⁹ Sexual temptation was part of the impetus, but it is not the whole story. No less important was the need, inter alia, “to create a private sphere.”⁸⁰ The first of three examples adduced by Alshech to emphasize this point is the veiling of slaves: “…if preventing lust and sexual desire were their only concern, why did many scholars differentiate between the dress code of female slaves and that of free women?”⁸¹ Alshech observes that the equation of modesty laws with sexual morality began to receive greater emphasis in the “late classical period” (i.e. the fourth-eighth/tenth-fourteenth centuries).⁸² Similarly, Youshaa Patel calls attention to the multiple purposes of veiling norms in the formative period. He emphasizes their “function of materializing hierarchical social distinctions between women.”⁸³ Patel remarks that the ʿUmar report demonstrates how “clearly the aim of maintaining visible class distinctions between slavegirls and freewomen [sic] in public life trumps the aim of controlling male excess sexual desire.”⁸⁴ Both Alshech and Patel emphasize status considerations over other factors. That status trumped sexual temptation in the minds of early jurists is clear evidence of the continued importance of the proprietary ethic. In a purely theocentric system, there would be no room for the distinction between free and enslaved women. By contrast, Ze’ev Maghen has characterized the Islamic approach to the control of sexuality as belonging to the “building strong walls” approach. The “building strong walls” approach, in contrast to the “building strong men” approach, implies a less than sanguine view of human nature.⁸⁵ The ʿUmar report thus embodies what Maghen would recognize as an exception to the general rule. It does not fit clearly into either conceptual scheme.

There is much evidence to suggest that no later the third Islamic century, the fear of fitna had begun to challenge the view that veiling maintained social hierarchies. The increased fear of fitna anticipates changes that, according to Alshech, took place in the late classical period (the fourth-eighth/tenth-fourteenth centuries). Asked if it is reprehensible (tukrah) for slave women to veil in public, Aḥmad b. Ḥanbal replied that a beautiful slave woman should in fact do so (idhā kānat jamīla tantaqib).⁸⁶ The main consideration here is the potential for temptation. Similarly, Mālik b. Anas is reported to have harangued the governor of Medina to prevent slave women from going about topless, albeit without success.⁸⁷ ʿAbd al-Malik b. Ḥabīb (d. 238/852) himself clearly experienced temptation in Medina, but nonetheless maintained the free/slave distinction.⁸⁸ By the end of the late classical period, the moralistic dimensions of veiling became more pronounced, as in the Madkhal of Ibn al-Ḥājj (d. 737/1336).⁸⁹

According to Alshech, the jurists’ equation of veiling with minimizing sexual temptation increases from the late classical period onwards (fourth-eighth/tenth-fourteenth centuries); the (theocentric) notion of curtailing fitna as the ratio legis of modesty norms becomes more prevalent. Postclassical jurists, and the moderns who follow their example—like al-Kawtharī—usually insist that free women conceal their faces, an opinion much less common in the formative period. The view that face-veiling is obligatory becomes the standard opinion in authoritative late works of the four Sunni legal schools. Only with the Ḥanbalīs does obligation represent continued fidelity to madhhab doctrine. The obligation is attested in the “canonical” works of the four Sunni schools authored immediately before the modern period.⁹⁰ The non-veiling of slaves, on the other hand, remains the standard position in the four Sunni legal schools. This distinction suggests an increasing divergence between the norms prescribed for free and enslaved women in the postclassical period,⁹¹ a discussion that evidently continued into the early fourteenth/twentieth centuries. As the modesty zone of free women expanded to encompass the face and hands—owing no doubt to the “corruption of the present age—” the modesty zone remained as it was for enslaved women. In other words, according to the majority of Sunni jurists, the tensions between the theocentric and proprietary ethics remained in place. As the argument for concealing the bodies of free women became increasingly popular, there does not seem to have been a corresponding doctrinal shift in the case of slaves. The non-veiling of slave women was, however, challenged.

No later than the fifth/eleventh century, a minority of Muslim jurists began to insist that the same veiling norms apply to free and enslaved women, a position that represents the triumph of theocentrism. This seems to be a rare opinion, of which I have been able to locate only a handful of examples. Tentatively, therefore, I suggest that this insistence is found most commonly among jurists of a textualist bent, including Ibn Ḥazm, Ibn al-Qaṭṭān (d. 628/1231) and Abū Ḥayyān al-Gharnāṭī (d. 745/1344). Among other textualists, Ibn Taymiyya (d. 728/1328) and his disciple Ibn al-Qayyim (d. 751/1350) are more restrained, stipulating veiling for slaves only in cases of fitna.⁹² Among jurists of the postclassical period, al-Shawkānī (d. 1250/1834) summarizes the arguments of both sides of the debate without clearly committing himself either way, attributing the pro-veiling view exclusively to the Ẓāhirīs.⁹³

Among the jurists who explicitly express their support for the veiling of slaves, Ibn al-Qaṭṭān limits himself to a passing remark that the command in Q. 24:31 that women “not reveal their beauty, except what is apparent thereof” applies to slave women no less than to free ones.⁹⁴ Similarly, Abū Ḥayyān observes that the instruction that “believing women” should cover themselves with their robes (Q. 33:59) is directed at both groups equally; if anything “the fitna of slave women is greater owing to their going about freely (taṣarrufihinna), in contradistinction to free women; excepting [slave women] from [the category of] “women” generally requires clear proof (dalīl wāḍiḥ).”⁹⁵ The explicit emphasis on fitna here bears out the point made by Alshech. The same concern is reiterated by Ibn Ḥazm, who is by far the most strident (and eloquent) critic of the teachings of the madhhab-jurists on the question. In an extended discussion of their arguments, he skewers his opponents for their inconsistencies, lack of rigor and gross disregard for the wellbeing of enslaved Muslim women.⁹⁶ It is a searing, searching critique, teeming with righteous indignation, and it represents the exact opposite of the view Alshech attributes to early classical scholars. On the proper interpretation of Q. 24:31 (typically understood, as noted, as meaning that free women must distinguish themselves from slaves), Ibn Ḥazm exclaims:

We declare ourselves innocent before God of this pernicious interpretation (tafsīr fāsid), which is either the error of a learned one—combining virtue and obliviousness (fāḍil ghāfil)—or the fabrication of a dissolute liar. [This is] because [the non-veiling of slaves] suggests that God the most high unleashed the depraved (fussāq) against Muslim slave women, a terrible calamity! No two Muslims disagree that illicit sex is prohibited with both free women and slave women, or that the punishment for such relations is the same, whether committed with one or the other…for this and other reasons it is evident that no opinion of anyone after the Prophet—may God have mercy on him and grant him peace—can be accepted unless it is supported by a chain of narrators [directly] to the Prophet.⁹⁷

Ibn Ḥazm gives short shrift to the ʿUmar report. Because it is not a Prophetic ḥadīth, it is not probative, particularly given the alleged disagreement of early Muslims on the question.⁹⁸ He is more of a scripturalist than the proto-Sunnis, who granted non-scriptural sources a much more considerable place than Ibn Ḥazm did in their jurisprudence.⁹⁹ Ibn Ḥazm’s methods and conclusions are echoed by al-Albānī, who refers to numerous sources and presents what is, to the best of my knowledge, the most sophisticated argument that free and enslaved Muslim women are subject to the same modesty norms.¹⁰⁰ To come full circle, and to return to the point on which I began this article, al-Albānī refutes the claim of an anonymous contemporary that the ḥijāb is now obsolete on the grounds that veiling is premised (as in interpretations of Q. 33:59) on a free/slave binary that no longer exists.¹⁰¹ There is no evidence to suggest that al-Albānī was aware of the writings of Naẓīra Zayn al-Dīn, but it is unlikely that he would have been impressed by them.¹⁰² At least among the abovementioned textualist jurists, one finds that the tension between proprietary and theocentric ethics is fully resolved, unambiguously, in favor of the latter.

Conclusion

The unveiling of slaves is grounded in the proprietary conception of sexual ethics. If juristic opinion had been guided entirely by theocentric norms—the fear of fitna—the distinction between slaves and free women would not have arisen. The increased importance of theocentric norms from the early to late classical periods suggests that less “permissive” attitudes towards sexual morality became the norm once the schools had more or less monopolized the production of Islamic law. Archaic fiqh data also points to the decreasing permissiveness noted by Alshech.¹⁰³ Though some jurists who dissented from the teachings of the legal schools extended the obligation of veiling to slaves, on theocentric grounds, this was a minority opinion that had only minimal influence on the school jurists. There are good reasons for this lack of influence, not least the presence of the free/slave binary in the Qurʾān, according to which slaves are exempted from the general rule (Q. 24:31) that women “conceal their beauty, except that which is apparent thereof.” Azam’s characterization of the tensions in Islamic sexual ethics is thus very appropriate.

Slaves are subject to the same moral imperatives as other human beings, with some exceptions (e.g. Q. 4:25); but they are also property, subject to most of the same transactional rules as other items of property. This ambivalence in the legal status of slaves is highlighted by Jalāl al-Dīn al-Maḥallī (d. 864/1459), who observes in the case of blood-money that

…there is indecision [as to] whether he [viz. the slave] is [analogous] to a free human—from the perspective that he is a human—or an animal—from the perspective that he is property. He resembles property more than he resembles a free human, through the proof that he is sold, can be inherited, or declared an endowment…[sic]¹⁰⁴

Al-Maḥallī’s opinion should not be taken at face value. As Behnam Sadeghi reminds us, the “legal reasons” a jurist assigns to particular rulings are fluid.¹⁰⁵ Al-Maḥallī defends the Shāfiʿī view against the Ḥanafīs, who held that the blood-money due for slaves is slightly less than the amount for free persons of equivalent status (e.g. religion). The Shāfiʿī doctrine is that the blood-money due for a slave is his or her sale price, however high this may be (though many set a maximum amount). In advocating the Shāfiʿī view, al-Maḥallī analogizes slaves to property rather than to human beings.¹⁰⁶

Notwithstanding these qualifications, the free/slave distinction remains meaningful. Jurists interested in expounding the wisdoms of the law similarly appealed to the status of slaves as property to explain their diminished modesty zone. Al-Qaffāl al-Kabīr (d. 365/976), commenting on the difference, writes that a slave woman is merely “one type of property (silʿa min al-silʿ); those who purchase her must turn her around [to view her from all angles].”¹⁰⁷ The same understanding is found in adab (belles-lettres) texts. In his (in)famous Epistle on Singing Girls, al-Jāḥiz (d. 255/868) saucily remarks that female slaves “are a variety of merchandise…and both vendor and purchaser need to examine the piece of goods carefully, and subject it to a close scrutiny [sic].”¹⁰⁸ Prurience aside, it is clear that slaves were subject to different standards of morality than free persons, and that these reduced expectations are based on the view that they are ultimately less than fully human. In Islamic law, the non-veiling of slaves is therefore a function of proprietary ethics: slaves are literally property and, as such, are subject to different rules than free persons.

The view that veiling is rooted in modesty norms that distinguish free women from slaves has serious moral implications that it would be remiss to leave unaddressed in this article. In her critical historiography of Islamic legal studies, Ayesha Chaudhry rightly emphasizes the need for scholarship to be morally responsible.¹⁰⁹ Academic research never takes place in a vacuum, and however much the scholar cultivates an attitude of disinterestedness, she or he can never entirely escape the political implications of their work. In his article on Orientalism and waqf (endowment) law in French Algeria and British India, David Powers has demonstrated that even otherwise sound academic research can have deleterious social consequences.¹¹⁰ Research can be instrumentalized for nefarious ends.¹¹¹

Reformists in the early twentieth century ce, such as Naẓīra Zayn al-Dīn, held that the ḥijāb is a hidebound and archaic institution. She and many of her contemporaries linked veiling to material and intellectual progress. “Those un-covered nations (al-umam al-sāfira),” she writes, “are the ones whose research has uncovered the secrets of nature and have subjugated the elements to their will… As for veiled nations (al-umam al-mutaḥajjiba)… They only sing of glories past.”¹¹² But the ḥijāb—or any sartorial or social practice—is amenable to multiple interpretations, as evidenced by a large body of scholarship on the varied significations of veiling in Muslim contexts. It is a point that bears repeating.¹¹³ In her Veil, Rafia Zakaria has discussed no fewer than five major resonances of the practice, ranging from “submission” and “purity” to “feminism,” “rebellion” and “subversion.”¹¹⁴ In A Quiet Revolution, Leila Ahmed writes about her realization that the ḥijāb can serve as a symbol of feminist empowerment, as on American college campuses.¹¹⁵ Just as Muslim jurists have attributed diverse rationales to discrete points of legal doctrine,¹¹⁶ Muslim women can and do exercise their agency in attributing diverse rationales to the ḥijāb.

Whatever the religious “authenticity” or contemporary resonances of Muslim veiling practices, the significance of veiling to Muslims has a history—subject to the constraints of the sources – that can be traced. I have sought to trace the earlier chapters of that history, without depriving modern Muslims of their agency to assign radically different meanings to the veil. That veiling symbolised the distinction between free and enslaved women, as exemplified by the ʿUmar report, and that later jurists generally upheld or even accentuated the distinction, are conclusions that may lead in any number of directions. It is clear, however, that whichever direction believers choose to pursue, they must take responsibility for their interpretive choices.¹¹⁷