:: Muwaṭṭaʾ Roundtable :: The Handmaiden’s Tale

By Kecia Ali (Boston University)

As someone who researches and teaches about early Islamic law, I have longed for a translation of the Muwaṭṭaʾ which renders legal terminology with consistency and precision, suitable for skimming, quoting, recommending to interested lay readers, and assigning to students. There have been two “nonacademic” English translations, primarily aimed at “pious Muslims” (2). The first, published in 1985, appeared in Pakistan; the second, by Ayesha Bewley in the U.K. in 1989. Neither is well-suited to classroom use. Of the two, I’ve had more recourse to Bewley’s version over the years. Though generally clear, it occasionally misrepresents points of law. This new translation involves Mālikī law experts and so avoids that pitfall. Yet it presents other challenges for scholars and teachers, especially in its euphemistic rendering of references to enslaved women and girls. In how it mediates among precision, contextualization, and accessibility, this version of the Muwaṭṭaʾ illustrates broader dynamics in current scholarly and popular conversations about slavery. My aim is not to criticize this translation, which is a considerable accomplishment, but to interrogate how we as scholars produce knowledge about the past that serves as a resource for scholarship and, at the same time, has religious and ethical valences for individual and communities.

Lately, people are talking a lot about slavery, concubinage, and consent within Muslim texts and history. Polemics and apologetics still abound, but there’s plenty of academic work too. At least three scholarly books addressing the premodern Muslim scholarly tradition on slavery and its implementation were published this year alone.[1] Academics, religious authority figures, and lay Muslims—overlapping categories which mask considerable variation in learning and thoughtfulness within each group—have weighed in on social media and blogs. Some offer whole theories about the need for enslaved women’s and girls’ consent to sex with their owners based on a decontextualized legal maxim or quotation—a purportedly pro-consent snippet from Shafiʿī has been making the rounds lately.[2] More substantive contributions consider not only the larger textual context for any given fragment but also the broader historical and intellectual framing that undergirds specific points of law. Most of the people engaging these debates, Muslim or not, readers of Arabic or not, have little familiarity with Islamic legal texts.

This translation aims to provide both a reliably precise English rendering and some historical background to situate the text in its historical and jurisprudential context for two distinct types of readers, whose needs stand in tension. Connell Monette, Mohammad Fadel, and their team of male translators imagine their audience as (a) academic but non-specialist readers, including “modern legal scholars who are not specialists in Islamic law,” and (b) “pious English-speaking Muslims.” (4)  Their rationale for consistently rendering legal terms and using “English technical terms” and “modern legal terminology, wherever possible” is to make the Muwaṭṭaʾ “accessible” to the first type of reader. (4) Yet their example of a technical term seems geared toward the latter: ṣalāṭ al-ṣubḥ, “Morning Prayer” (5-6). This obvious literal rendering hardly seems to qualify as technical terminology, especially since prayer isn’t a central concern of comparative legal study. While this example emphasizes ritual observance, the editors are at pains to note that the Muwaṭṭaʾ is by no means confined to “religious law” but also deals extensively with “‘secular’ law” (25-26), which they estimate comprises approximately half of the text—and in which enslaved people figure frequently.

Slavery is ubiquitous in premodern Muslim legal texts, with enslaved people appearing as both objects (e.g., of sales) and subjects (e.g., as ritual agents or as social actors in a divorce or commercial transaction).[3] A survey of the books of manumission (which comprises 7% of the Muwaṭṭaʾ) and “marriage, divorce, and fosterage by suckling” (9%) reveals an array of terms for enslaved people. Some of the terms used in translation de-emphasize the legal dimensions of enslavement, and soften its resonance, while playing up gendered distinctions. While ‘abd (slave, enslaved man) appears throughout as slave, the terms most often used for enslaved women and girls (ama and jariya) appear here as “handmaiden.”

The grammatical gender of Arabic nouns doesn’t convey easily into English, of course. The Library of Arabic Literature edition of Consorts of the Caliphs, which features many enslaved women, translates ama as slave rather than female slave; its translators assume, correctly, that context indicates gender, and favor the less clunky solution.[4] In the Muwaṭṭaʾ, where gender will not always be clear from context, one might argue for equal-opportunity clunkiness: rather than allow maleness to go unmarked, one might translate ‘abd as male slave (or enslaved man) and ama or jariyaas female slave (or enslaved woman).

The rendering of jariya and ama as “handmaiden” poses several problems. A minor consideration is that the translation renders the terms ama and jariya indistinguishable, but so would translating both as “enslaved woman.” More troubling, “handmaiden” masks or at the very least generates ambiguity about enslavement. English dictionaries emphasize the “servant” dimensions of “handmaiden.” Some offer handmaid and maidservant as alternatives; slave status is an afterthought, if mentioned at all. “Handmaiden” sounds archaic and perhaps courtly. Real talk, though: would we call Sally Hemings a handmaiden?

“Handmaiden” also appears here in the cumbersome translation of umm walad as “a handmaiden who has borne her master a child” (e.g., #1875) A couple of years ago on Twitter, I facetiously suggested that umm walad could be translated “baby mama.” There are numerous problems with that rendering—not least, its racialized connotations—but articulating why it isn’t appropriate forces us to think about what kinds of relationships our chosen terms conjure for readers.

The disjunction in the translations of terms for enslaved males and enslaved females creates and magnifies gender distinctions, even as it downplays women’s enslavement as in the case of “a handmaiden [who] is married to a slave” (#1772). Although that report goes on to refer to her manumission, which indicates her enslaved status, the English version sets up a lack of parity between the spouses not found in the original.

The fairly rare term mamlūk(a) appears consistently as “chattel slave,” a translation which emphasizes the element of ownership conveyed in the Arabic phrase. But this too can create confusion. One report on post-divorce maintenance contains the phrase “Neither a free man nor a slave who divorces a handmaiden who is a chattel slave (mamlūka)” (#1817). The Arabic simply has “mamlūka,” without any other noun; it could have been rendered “Neither a free man nor a slave who divorces a chattel slave.”[5] Presumably, the inclusion of handmaiden in such instances is to clarify gender. But by including then qualifying the originally absent term “handmaiden” (“a handmaiden who is a chattel slave”), the text seems to suggest some handmaidens aren’t chattel slaves. Which readers will know better? What sorts of not-enslaved maidservants might they imagine in relationships with and bearing children to men of varied social statuses?

Translations elide the vastly different social contexts in which all of these relationships appear. Some Arabic terms evoke both youth and enslavement. Ghulam (here: “slave-boy”) sometimes elsewhere merely means a male youth; jariya, often translated as slave-girl, can simply mean girl. The phrase “her people” (#1729) applied to the household of an enslaved woman reflects a sensibility in which enslaved people were part of (hierarchical) households/family structures. This point will escape non-specialist readers, who may be tempted to read “her people” not as her owners/captors/enslavers but rather her kin or family. Such terms invoke widely divergent social imaginaries for contemporary readers than they did for the Muwaṭṭaʾ’s original audience.

Other choices are less issues of translation than questions of best English usage. In recent scholarship and public commentary on United States slavery, there’s been a perceptible shift toward “enslaved person” rather than “slave.” As journalist Nikole Hannah-Jones explains, the term ‘slave’ suggests what someone is whereas the word ‘enslaved’ makes clear that they have been forced into servitude.[6] Having used the former for years, I now try to use the latter where possible. In rendering a text like the Muwaṭṭaʾ, one question is whether to reflect or push back against the sensibility of the original text.[7] If we use “slave,” are we giving a faithful version of the original for analysis or are we helping normalize the processes through which human beings become commodities? There is no perfect solution to any of these issues, but it is vital to grapple with them.

Discussing her recent translation of a canonical epic, classicist Emily Wilson observes “how much translators have worked to remove or reduce slavery from The Odyssey.” By rendering “the words for ‘slave’” in alternate ways, they produce “mistranslations, motivated presumably by a desire to idealize Homeric society and Homeric poetry and remove the troubling fact that it depicts a slave-owning society.”[8] Some years ago, translator Nuh Keller omitted from his English rendering the material on slavery and enslaved people in the Arabic text of Reliance of the Traveler. His translation was intended as guidance for contemporary believers, so he excised what he considered “matters now rare or non-existent” or simply “no longer current.”[9] This version of the Muwaṭṭaʾ thankfully makes no moves to bowdlerize the text or hide the presence of slavery. And yet, the vaguely biblical and deeply gendered rendering of terminology for enslaved people is something readers should reflect on—and those of us who want to teach with the text will need to address.

In other reflections on translating The Odyssey, Wilson has noted how much of the sensibilities of male translators showed up in their renderings.[10] I wonder how the final result might differ had women been involved in this long-term, transnational, complex project of editing and translating the Muwaṭṭaʾ. I bet we would have far fewer handmaidens.



[1] Mary Ann Fay, ed., Slavery in the Islamic World: Its Characteristics and Commonality (New York: Palgrave, 2019); Bernard K. Freamon, Possessed by the Right Hand: The Problem of Slavery in Islamic Law and Muslim Cultures(Leiden: Brill, 2019), and Jonathan A. C. Brown, Slavery and Islam (London: Oneworld, 2019).

[2] The phrase from the Umm is “فَأَمَّا الْجِمَاعُ فَمَوْضِعُ تَلَذُّذٍ وَلَا يُجْبَرُ أَحَدٌ عَلَيْهِ” (“However, intercourse is a matter of pleasure and no one is compelled to it”). For one recent take on the topic, see Abu Amina Elias, “Sexual Consent, Marriage, and Concubines in Islam.” https://abuaminaelias.com/consent-marriage-concubines/ (offline on 12/4/19). However, this passage, understood in its context, doesn’t speak to consent but rather asserts that men have no obligation to have sex equally with their wives. Kecia Ali, Marriage and Slavery in Early Islam (Cambridge, MA: Harvard University Press, 2010), 119.

[3] Kecia Ali, Marriage and Slavery in Early Islam (Cambridge, MA: Harvard University Press, 2010), 6-8.

[4] Ibn al-Sa‘i, Consorts of the Caliphs: Women and the Court of Baghdad, ed. Shawkat M. Toorawa (New York: New York University Press, 2015).

[5] A variant manuscript does include the noun ama, making it like #1795 where “a handmaiden who is a chattel slave” translates “ama mamlūka.”

[6] In this excerpt from a video interview on The Breakfast Club, journalist N. Hannah Jones discusses why she uses “enslaved” rather than “slave.” https://twitter.com/nickimayonews/status/1167407040702099456

[7] If someone argues that “enslaved person” is too unwieldy, given that “slave” is only one word in Arabic, they might be reminded that the Arabic ḥurr(a), a “free person,” requires two words to be comprehensible English.

[8] Fran Wilde, “The Voices of The Odyssey: Emily Wilson On Language, Translation, and Culture” May 3, 2018. https://www.tor.com/2018/05/03/all-the-voices-of-the-odyssey-emily-wilson-on-language-translation-and-culture/

[9] See Kecia Ali, Sexual Ethics and Islam: Feminist Reflections on Qur’an, Hadith, and Jurisprudence (London: Oneworld, 2016), 62-63.

[10] https://twitter.com/EmilyRCWilson/status/970406691375378432

Leave a Reply