Al-ʿUṣūr al-Wusṭā 31 (2023): 96-126

Abstract
Female domestic slaves designated as jawārī (sing. jāriya) featured in a range of medieval Arabic sources, 
including treatises on the mechanical arts. They appeared, for example, as liquid-serving devices and 
timekeepers. Scholarship on automated jawārī, however, has been scant; little, in fact, has been written on 
gender, slavery, and technology in the medieval Middle East. Generally, figurative machines have been framed 
as either practical, proto-robotic forms or wondrous, elite contrivances. Though seemingly innocuous, these 
approaches are at risk of reinscribing modern biases. In particular, the utilitarian discourse that sees machines 
as neutral, useful objects has been shown to manifest a position of mastery, reinforcing a slippage between 
worker and tool. Yet the hypothesis that the automata were objectifying, framing the jāriya as the object of 
a patron’s viewpoint, might also partake in the binaries of subject and object, master and slave. To attend to 
the gendered and class-based politics of automated jawārī requires that we forgo common assumptions about 
both technology and objectification. The notion of instrumentality—here defined as the quality of serving as an 
instrument (āla) in a process of carrying and transmitting—may prove helpful in this regard. Jawārī, it turns out, 
were portrayed as instrumental figures in various other domains, including in domestic and spiritual contexts. 
Ṣūfī saints’ encounters with inspired jawārī, for example, foregrounded the female servant as a vector of divine 
wisdom. Like the engineer’s prototypes, these narratives were the product of a male, patriarchal viewpoint, 
oscillating between demeaning and valorizing effects; as such, they fortified some of the norms that made jawārī 
representable, notably as hypervisible, mediating agents. At the same time, these representations expose the 
limits of utilitarian, apolitical approaches to technology while positing jawārī as vehicles—more than objects—
of instrumentality.

Female domestic slaves characterized as jawārī (sing. jāriya) are mentioned in a variety 
of medieval Arabic sources. Although the term could also designate a free young woman, in 
Arabic jāriya was most often employed for a female slave, regardless of age; preferred to less 
ambivalent denominations such as ama, it was the word of choice in the slave trade’s written 
record, including in bills of sale for slaves and slave-buying advice manuals, as well as in the 

Instrumental Jawārī:  
On Gender, Slavery, and Technology                                            

in Medieval Arabic Sources*

Lamia BaLafrej
University of California, Los Angeles

(lbalafrej@humnet.ucla.edu)

© 2023 Lamia Balafrej.  This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 
License, which allows users to copy and distribute the material in any medium or format in unadapted form only, for noncommercial purposes only, and 
only so long as attribution is given to the original authors and source.

* I am grateful to the three anonymous reviewers and to Zayde Antrim for providing careful, generous 
comments, which have significantly improved this essay. Research for this study—and the book project it 
relates to—has been supported by a 2023 Rome Prize in Medieval Studies from the American Academy in Rome 
and a 2023 Getty Scholar Grant from the Getty Research Institute in Los Angeles. I thank both institutions, as 
well as the fellows, whose feedback and encouragement have been invaluable.

mailto:lbalafrej%40humnet.ucla.edu?subject=


97  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

literary and historiographical canon. The term is also used in treatises on the mechanical 
arts, where jawārī appear, for example, as liquid-serving devices and timekeepers. The 
machines themselves confirm the link between the jawārī they represent and bonded labor, 
specifically domestic slavery. This is signified through the functions they replicate, such as 
the serving of liquids, and through the positions of the figures, which are often placed at 
the threshold, as though coming in and out of a house—a level of visibility that was often 
associated with jawārī’s enslaved status. 

Predominantly feminized in both historical and historiographical sources, household 
slavery was a common practice in many parts of the premodern Middle East, present 
in both courtly and nonelite contexts and across various communities, Muslim and 
non-Muslim alike.1 Although a lot of scholarly attention has been given to elite female 
slaves who sometimes gained access to wealth, prestige, and power, the largest proportion 
of enslaved women was likely employed for domestic and sexual labor in more anonymous, 
urban settings, including the homes of merchants, scholars, and artisans.2 Given that 
foreignness was one of the main criteria for enslavement, most of these women, like 
their male counterparts, were brought from the peripheries of the Islamic world—mainly 
through war in early Islam, then also through trade—though some were also born within 
its confines.3 As a result of their presence in urban and courtly milieus, jawārī permeated a 
broad representational repertoire, a phenomenon that has yet to receive full consideration. 
They appeared in sources not often studied in relation to slavery, including the mechanical 
treatises and literary accounts this essay explores. The circulation of jawārī as a motif in 
so many genres fortified, in turn, some of the perceptions and rules that regulated jawārī’s 
roles, including the expectation that female slaves were and ought to be more visible than 
free women due to their status as commodities and due to their activities, such as serving as 
their owners’ proxies. 

1.  On domestic slavery in the medieval Middle East in general, see Jan Hinrich Hagedorn, Domestic Slavery 
in Syria and Egypt, 1200–1500 (Bonn: Bonn University Press, 2020); Shaun E. Marmon, “Domestic Slavery in 
the Mamluk Empire: A Preliminary Sketch,” in Slavery in the Islamic Middle East, ed. Shaun E. Marmon, 1–23 
(Princeton, NJ: M. Wiener, 1999). For documents of the slave trade, see the deeds of sale of slaves published 
in Yusuf Ragib, Actes de vente d’esclaves et d’animaux d’Egypte médiévale, vol. 1 (Cairo: Institut français 
d’archéologie orientale, 2006). 

2.  On enslaved women in particular, see Craig Perry, “Slavery and Agency in the Middle Ages,” in The 
Cambridge World History of Slavery, AD 500–AD 1420, ed. Craig Perry, David Eltis, Stanley Engerman, and David 
Richardson, 240–67 (Cambridge: Cambridge University Press, 2021), esp. 245–53; Khalil ʿAthamina, “How Did 
Islam Contribute to Change the Legal Status of Women: The Case of the Jawārī, or the Female Slaves,” Al-Qanṭara 
28 (2007): 383–408; S. D. Goitein, “Slaves and Slave Girls in the Cairo Geniza Records,” Arabica 9 (1962): 1–20; 
Matthew Gordon and Kathryn Hain, eds., Concubines and Courtesans: Women and Slavery in Islamic History 
(Oxford: Oxford University Press, 2017); Elizabeth Urban, Conquered Populations in Early Islam (Edinburgh: 
Edinburgh University Press, 2020).

3.  For an overview of the slave trade and its ideological justifications in medieval Islam, see Craig Perry, 
“Historicizing Slavery in the Medieval Islamic World,” International Journal of Middle East Studies 49, no. 1 
(2017): 133–38; Hannah Barker, That Most Precious Merchandise: The Mediterranean Trade in Black Sea Slaves, 
1260–1500 (Philadelphia: University of Pennsylvania Press, 2019).



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   98

The automated jāriya was a rather frequent motif, evident in poetic, narrative, and 
technical writings, yet it has garnered only limited attention. There has been, in fact, little 
research conducted on gender and slavery in the field of medieval Islamic technology 
studies. Generally, the technical and the social have been kept separate. Social historians of 
the medieval Middle East have mostly ignored the mechanical arts. Meanwhile, specialists 
in Islamic science and technology have mainly looked at machines as universal, proto-
robotic forms, free of social and political issues. Another common assumption, especially 
among historians of Islamic art, is that figurative machines were wondrous, courtly luxuries, 
reflective of a patron’s glory but with no implications for broader social categories. The split 
between technology and society is due to a host of reasons, including the prevailing view 
that technical artifacts are passive, neutral means, belonging to the sphere of capability, not 
the realm of social relations. But automated jawārī, I argue, pose a serious challenge to the 
utilitarian approach, refocusing attention from sheer mechanism and innocuous wonder to 
technology’s politics. 

Given the lack of research on gender, slavery, and technology in medieval Islam, this 
essay cannot be comprehensive; rather, its goal is to lay some ground for future research 
by defining a corpus, discussing its contours, and raising historiographical and conceptual 
issues. First, I address some of the reasons for the paucity of studies on automated jawārī. 
Special emphasis will be given to entrenched, yet inconspicuous, misconceptions that are 
not specific to Middle Eastern studies, such as the utilitarian definition of technology, or 
what I call functionalism: the discourse that approaches technical objects as a transparent 
means to an end, framing them as predictable, compliant objects. By projecting servility 
onto technology, functionalism relies on and reproduces a slippage between slave and 
machine. As feminist and Black studies have shown, the notion of equipment as a useful, 
efficient extension of human activity, though seemingly universal and benevolent, evinces 
a position of mastery, especially in its desire for dutiful service. By adding to the technical 
a representational level that centers the body of the enslaved, automated jawārī, in a way, 
unsettle the myth of technical neutrality, bringing to the fore the gender and class dynamics 
that functionalism has served to repress.

Second, I provide an overview of the main types of automated jawārī I have found so far, 
based on Arabic treatises on the mechanical arts and their manuscripts, which span the 
thirteenth and fourteenth centuries. Two categories stand out: the automated jāriya, acting 
as a liquid dispenser, and the timekeeping jāriya, often seen throwing pellets from her 
mouth as a way of marking the passage of the hours. In some cases, the mechanical jāriya 
crosses a threshold, suggesting mobility across various social spheres. These performances 
of visibility and serviceability enhance some of the norms that were used to represent and 
control enslaved women in medieval Islam; as such, they highlight jawārī’s availability and 
fungibility. Yet the hypothesis that the automata were objectifying, framing the jāriya as 
the object of a subject’s viewpoint, remains tied to utilitarian understandings of technology; 
it may be just as fraught as the functionalist attitude, for it partakes in the same dyads of 
subject and object, master and slave. 



99  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

To begin to render automated jawārī with any complexity requires that we step away 
from modern assumptions about both technology and objectification, especially if they imply 
a subject’s domination over a fungible, servile object. A focus on automated jawārī alone, 
severed from other discursive genres, would, moreover, reinforce the modern disciplinary 
boundaries that have separated technology and politics in Islamic studies. Looking at a 
wider archive through a critical, comparative approach in fact reveals a cross-genre link 
between jawārī and instrumentality, here defined as the quality of serving as an instrument 
(āla) in a process of carrying and transmitting. Notably, jawārī appeared in mystical stories 
as divinely inspired intercessors, delivering mystical guidance to a male Ṣūfī aspirant. Like 
the engineer’s machines, these narratives were the product of a male, patriarchal viewpoint, 
oscillating between subjugating and valorizing effects. This perspective, however, should 
not be confused with the modern, dichotomous form of objectification that opposes subject 
and object. Rather, the quality of being an instrument may point to a mixed, or vectorizing, 
process, framing jawārī as vehicles—more than objects—of instrumentality.

Unlearning Functionalism

There are many possible explanations for the absence of gender-sensitive approaches to 
Islamicate technology, one of the most conspicuous being the profession’s own gendered 
dynamics. What concerns me here, however, is not so much the scarcity of research on 
mechanical jawārī per se, or even the biases of previous scholarship. It would have been 
disingenuous to expect anything else, for until feminist technology studies, which emerged 
only in the 1980s, the anthropology and history of technology had largely been unaware 
of gender, within and beyond Islamic studies.4 Rather, I would like to draw attention 
to some of the fundamental misconceptions that may continue to preserve uncritical, 
patriarchal representations, albeit in less discernable—and thus more insidious—forms. 
These include functionalism, the theory of the utilitarian machine that correlates form 
and function, means and ends in a direct, predictable way. The focus on mechanism has 
prevented scholars from analyzing technical objects as representations that may shed light 
on social and intellectual history. While seemingly harmless, functionalism is also at risk 
of reinforcing the dualisms of slave and master, body and mind, other and self, female and 
male, as feminist and Black studies have demonstrated. 

To begin with, a general reason for the rarity of discourse on the social and political 
dimensions of self-moving machines (ḥiyal or ālāt mutaḥarrika in Arabic) in medieval Islam, 
despite the fact that many represented lowly workers, is the assumption that technical 
treatises yield little insight into societal matters.5 This rests upon the larger belief that 
tools, instruments, and machines are practical, universal objects, unaffected by political 
change, having developed progressively, from the flint to the computer—a point I return 

4.  On the historical split between feminist studies and technology history, see Francesca Bray, “Gender and 
Technology,” Annual Review of Anthropology 36 (2007): 37–53.

5.  For a recent corrective and an exploration of slavery and automation in thirteenth-century Anatolia, see 
Lamia Balafrej, “Automated Slaves, Ambivalent Images, and Noneffective Machines in al-Jazarī’s Compendium 
of the Mechanical Arts, 1206,” 21: Inquiries into Art, History, and the Visual 3, no. 4 (2022): 737–74 .

https://doi.org/10.11588/xxi.2022.4.91685
https://doi.org/10.11588/xxi.2022.4.91685


Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   100

to below.6 Another reason for the dearth of critical analysis is the rhetoric of the golden 
age, the widespread impression that the premodern period constituted the peak of Islamic 
civilization.7 An aura has been attached to Islamicate automata, through the invocation of 
awe, pleasure, and innovation, framing instruments and machines as evidence of medieval 
Islam’s cultural and scientific efflorescence.8 To the extent to which there is any mention 
of politics, animated statues—trees filled with mechanical birds, monumental water clocks, 
automated guards, and metal armies—have been interpreted as spectacles of princely power, 
including for diplomatic purposes.9 This view, again, defines automata as exceptional, 
exclusive artifacts, crystallizing the (male) patron’s authority, with few implications for 
broader social groups.

Exceptionalism has had detrimental effects. Broadly, the “golden age” narrative has 
hindered critical engagements with the Islamic past while equating modern Islam with 
decline, thus playing into orientalist, colonial views. For a long time, it also buttressed the 
Eurocentric tendencies of Islamic technology studies. Indeed, the idea of medieval Islam’s 
superiority justified its study as a bridge between the ancient world and Europe’s so-called 
Renaissance.10 As a result, especially in technical and scientific studies, scholarly attention 
went to medieval Arabic sources insofar as they provided a mirror for understanding 
Western culture, either by preserving ancient Hellenistic thought or by anticipating early 
modern European developments or both.11 The search for continuities with Western 

6.  Though generally focused on mechanism and practical function more than on such political issues as gender 
and labor, surveys of tools and machines in medieval Islam have played an essential role in defining an archive 
for the study of technical and mechanical objects; see Francis Maddison and Emilie Savage-Smith, Science, Tools 
and Magic, part 1: Body and Spirit, Mapping the Universe (London: Nour Foundation with Azimuth Editions and 
Oxford University Press, 1997); David A. King, Islamic Astronomical Instruments (London: Variorum Reprints, 
1987); Siegfried Zielinski and Peter Weibel, eds., Allah’s Automata: Artifacts of the Arab-Islamic Renaissance 
(800–1200) (Ostfildern: Hatje Cantz, 2015). Surveys directed at the wider public include Aḥmad Yūsuf al-Ḥasan 
and Donald Hill, Islamic Technology: An Illustrated History (New York: Cambridge University Press, 1986). 

7.  For more references on the “golden age” discourse and its impact, notably on the study of slavery in 
Islamic art history, see Lamia Balafrej, “Domestic Slavery, Skin Color, and Image Dialectic in Thirteenth-Century 
Arabic Manuscripts,” Art History 44, no. 5 (2021): 1012–36, at 1014–15.

8.  Laura Lee Schmidt has analyzed this celebratory discourse in “Islamic Automata in the Absence of Wonder” 
(SM thesis, Massachusetts Institute of Technology, 2010). For an overview of the medieval Islamic imaginary of 
the automaton, see Anna Caiozzo, “Entre prouesse technique, cosmologie et magie: L’automate dans l’imaginaire 
de l’Orient médieval,” in La fabrique du corps humain: La machine modèle du vivant, ed. Véronique Adam and 
Anna Caiozzo (Grenoble: MSH-Alpes, 2010), 43–79. For a transhistorical approach to Islamicate automata, one 
also shaped by media studies, see Zielinski and Weibel, Allah’s Automata.

9.  T. M. P. Duggan, “Diplomatic Shock and Awe: Moving, Sometimes Speaking, Islamic Sculptures,” Al-Masāq 
21 (2009): 229–67; Joel Pattison, “A Golden Tree in the ‘Garden of Pages’: The Genoese Embassy to Morocco of 
1292,” Journal of Medieval Worlds 1, no. 4 (2019): 1–9; Nahid Norozi, “The ‘Metal Army’ of Alexander in the War 
against Indian King Porus in Three Persian Alexander Books (Tenth-Fourteenth Centuries),” Iranian Studies 52 
(2019): 903–22.

10.  For a nuanced account of this narrative, see George Saliba, Islamic Science and the Making of the 
European Renaissance (Cambridge, MA: MIT Press, 2007).

11.  This is true for other aspects of Islamic culture, including the study of the sciences; see Ahmed Ragab, 
“Making History: Identity, Progress and the Modern-Science Archive,” Journal of Early Modern History 21 

https://dspace.mit.edu/handle/1721.1/59207


101  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

technology may explain the lack of interest in automata’s figurative aspects, as opposed 
to their inner workings, for although the latter echoed ancient mechanisms, the former, 
particularly in their insistence on humble figures, must have seemed peculiar to modern, 
positivist scholars while also running counter to celebratory accounts of technology. This 
is not to say that animated figures of slaves were unique to medieval Arabic treatises (they 
could be found in sources from the Mediterranean to China) or that they did not anticipate 
modern technology (they do foreshadow aspects of robotics, including its conceptual 
linkage to slavery).12 Rather, it is to suggest the possible role of Eurocentric, modernist 
biases—including congratulatory, teleological views of technology as neutral, efficient, 
and progress-driven—in obscuring automata’s social dimensions, in particular their link to 
forced labor.

The scarcity of attention to automated jawārī may also have to do with the historiography 
of gender and slavery in medieval Islam, given its spotty engagement with mundane forms 
of slavery, as well as its emphasis on literary and normative sources, as opposed to more 
oblique forms of representation. Scholarly disinterest also appears through such elusive 
yet ingrained practices as translation habits, which tend to minimize the roles of nonelite, 
ordinary female slaves and servants. In English translations of Arabic treatises of the 
mechanical arts, jāriya tends to be rendered as “maiden” and “slave-girl”; because they 
foreground young age, both terms are infantilizing.13 Surely, jāriya could designate a young 
woman, but this was not its only meaning, especially in texts that often provided no clues 
as to the jāriya’s age. One could retort that “slave-girl” translates historical slaveholders’ 
propensity to employ euphemistic, juvenilizing language for slaves—this phenomenon 
is well known for ghulām or “boy” and also applied to jāriya.14 But this carries the risk 
of strengthening the owner’s perspective.15 Moreover, translating jāriya as “maiden” 
and “girl” may be sexualizing, as both terms imply unmarried standing and thus sexual 
availability. In addition, these terms romanticize the function of service while downplaying 
the jāriya’s possible status as a slave. As scholars of Arabic literature have pinpointed, 
such mistranslations testify to the subtle, sometimes unthought, mechanisms by which 

(2017): 433–44. A similar point—about the selective, Eurocentric approach to medieval Arabic sources—is made 
about al-Andalus in Maribel Fierro, “The ‘Bestsellers’ of al-Andalus,” in Artistic and Cultural Dialogues in the 
Late Medieval Mediterranean, ed. María Marcos Cobaleda, 31–56 (Cham: Palgrave-MacMillan, 2021), at 32–33. 

12.  On automated slaves and subalterns in the medieval mechanical arts, see Balafrej, “Automated Slaves.” 
13.  To translate jāriya, Donald Hill used “slave-girl” in his 1974 translation of Ibn al-Razzāz al-Jazarī’s 1206 

treatise al-Jāmiʿ bayn al-ʿilm wa-l-ʿamal al-nāfiʿ fī ṣināʿat al-ḥiyal; Ahmed Ragad employed “maiden” in his 2008 
translation of Ibn Khalaf al-Murādī’s eleventh-century manual Kitāb al-Asrār fī natāʾij al-afkār. Full references 
to these books and their translations will be given in this essay’s second section.

14.  Both ghulām and jāriya were preferred, respectively, to the less ambivalent ʿabd and ama, which tended 
to be used more as metaphors, for example to express one’s humility. On the euphemistic use of ghulām, see 
Goitein, “Slaves and Slave Girls,” 2. About the figurative use of mamlūka or ʿabda in, for example, legal petitions, 
see Geoffrey Khan, “A Petition to the Fāṭimid Caliph al-’Āmir,” Journal of the Royal Asiatic Society of Great Britain 
and Ireland 122, no. 1 (1990): 44–54.

15.  Shaun E. Marmon, “Intersections of Gender, Sex, and Slavery: Female Sexual Slavery,” in Perry et al., 
Cambridge World History of Slavery, 185–213, at 202 n. 27. 



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   102

translators, editors, and commentators have effaced, exoticized, or oversimplified issues of 
gender.16 

As mentioned earlier, more than with sexist, patronizing language, my concern is with 
broader, seemingly objective conceptions of technology that may perpetuate a masculinist 
perspective, even as they seem ungendered; such conceptions, it turns out, may be tied to the 
institution and ideology of slavery, a relationship that makes the critique of functionalism 
even more relevant to this essay. One common yet misleading belief has framed technology 
in a utilitarian fashion, with an emphasis on operation, function, and usability, in ways that 
reinscribe both heteropatriarchy and the master’s viewpoint, specifically the desire for 
obedient, machine-like workers. This could also be dubbed a positivist approach, considering 
that it represents technical artifacts as neutral, self-evident aids, external to the body and 
to the social realm. The result has been an uncoupling of fields between the technical and 
the social, through an emphasis on inner mechanism and practical use.17 This approach is 
by no means limited to Islamic technology studies and conveys a widespread, Promethean 
notion of the machine as efficient, useful, and civilizing.18 Progress is a corollary idea, a view 
of the history of technology as a steady, gradual process of complexification, leading to ever 
more sophisticated devices and systems.19 

But far from being apolitical, utilitarian definitions of tools and machines as predictable, 
unresisting, and as prostheses, expanding human action may actually betray a position of 
superiority, if not a language of mastery. Louis Chude-Sokei has pointed out the “sedimented 
racism of seemingly lifeless objects,” how the rhetoric of the inert tool has worked hand in 
hand in the modern period with colonial expansion, racial dominance, and the Atlantic slave 
trade, dehumanizing colonized and enslaved populations as “instruments of production,” in 
Aimé Césaire’s phrase.20 Bruno Latour repeatedly made a similar point, for example when 
he observed that the National Rifle Association’s approach to guns as passive tools was 
imbued with an illusion of control and with the supposition, contingent on the ideology of 
slavery, that the role of machines had been “rehearsed for centuries by obedient human 
servants.”21 Today the search for the subservient tool continues to inform the making of 

16.  Rachel Schine, “Translating Race in the Islamic Studies Classroom,” Al-ʿUṣūr al-Wusṭā 30 (2022): 320–83; 
Zayde Antrim, “Qamarayn: The Erotics of Sameness in the 1001 Nights,” Al-ʿUṣūr al-Wusṭā 28 (2020): 1–44, 
at 24–28; Christine Chism, “Lost Worlds: Encyclopaedism and Riddles in the Tale of Tawaddud/Theodor,” 
in Bestsellers and Masterpieces: The Changing Medieval Canon, ed. H. Blurton and D. F. Reynolds, 234–61 
(Manchester: Manchester University Press, 2022), at 235–36. 

17.  On the history of this split in the anthropology of technology, see Tim Ingold, “Eight Themes in the 
Anthropology of Technology,” Social Analysis 41, no. 1 (1997): 106–38.

18.  John M. Staudenmaier, Technology’s Storytellers: Reweaving the Human Fabric (Cambridge, MA: MIT 
Press, 1985). 

19.  Technological progress was also foundational to colonialism; see Michael Adas, Machines as the Measure 
of Men (Ithaca, NY: Cornell University Press, 1989).

20.  Louis Chude-Sokei, The Sound of Culture: Diaspora and Black Technopoetics (Middletown, CT: Wesleyan 
University Press, 2016), 82, 38. I thank Kodwo Eshun, Amanda Trager, and Erik Moskowitz for discussing this 
work with me.

21.  Bruno Latour, Pandora’s Hope: Essays on the Reality of Science Studies (Cambridge, MA: Harvard 
University Press, 1999), at 192–93, 207–8.

https://journals.library.columbia.edu/index.php/alusur/article/view/uw30schine
https://journals.library.columbia.edu/index.php/alusur/article/view/8287


103  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

technology, including algorithms, as exemplified by software such as Apple’s Siri, whose 
female gendered voice serves to highlight the algorithm’s function as a compliant personal 
assistant.22

The idea of technical objects as reincarnations of slaves is perhaps nowhere more 
obvious than in the imaginary of the robot. The very word “robot,” in fact, was coined 
in 1923 from the Czech word for “slave” or “serf” (robota) in Karel Čapek’s play R.U.R. 
(Rossum’s Universal Robots). The first robots, moreover, literally represented slaves. One 
example is the 1930 mechanical slave designed by the Westinghouse Electric Company 
to perform domestic tasks. Named Rastus and shaped as a Black man, the Westinghouse 
robot bears witness to the centrality of both race and slavery to modern automation in 
the United States, as Chude-Sokei has shown.23 The link between robotics and slavery was 
diffused among the general public as well. A 1957 article published in Mechanix Illustrated, 
an American consumer magazine, compared robots to “push-button servants,” bluntly 
announcing to its readers that by 1965 they would “have personal slaves again.”24 

Functionalism may also bespeak heteropatriarchal worldviews. There is a vast literature 
on how the utilitarian machine intersects with the perception of the woman’s body as a 
reproductive tool and, more generally, with the gendered dichotomy that opposes passive, 
sustaining matter—the feminine—and active, rational form—the masculine.25 The machine’s 
status of practical prosthesis overlaps with patriarchal models of representation because 
it relies on a teleology of form and function that has also shaped biological essentialism, 
the idea that gender difference is determined by the form of sexual organs. As a mode 
of organization of production relations, the gender system that correlates biological 
form and social function intensified in modern times, when it was imposed on colonized 
populations, consolidating global colonial capitalism, as María Lugones has argued. For 
Lugones, heteropatriarchy is not an ahistorical, transcultural force, but was propelled by 
what she calls “the modern/colonial gender system,” a Western, hegemonic model that 
naturalized racial, gender, and sexual differences as hierarchical divisions while mapping 
these differences onto the split of physical and intellectual labor.26 

22.  Thao Phan, “The Materiality of the Digital and the Gendered Voice of Siri,” Transformations 49 (2017): 
23–33.

23.  On Čapek’s play, Westinghouse’s robot, and the entanglement of slavery, racism, and robotics, see Chude-
Sokei, Sound of Culture, esp. chapter 1. Also see Ruha Benjamin, Race after Technology: Abolitionist Tools for 
the New Jim Code (Medford, MA: Polity Press, 2019), esp. chapter 1.

24.  O. O. Binder, “You’ll Own ‘Slaves’ by 1965,” Mechanix Illustrated, January 1957, 62–65. For recent, critical 
takes on automation’s false promises, see Luke Munn, Automation Is a Myth (Stanford, CA: Stanford University 
Press, 2022); Jason Resnikoff, “The Myth of Black Obsolescence,” International Labor and Working-Class History 
102 (2022): 124–45.

25.  For recent summaries of heteronormative conceptions of the machine, with forays into how the 
machinic can also perturb heteropatriarchy, see Neda Atanasoski and Kalindi Vora, Surrogate Humanity: Race, 
Robots, and the Politics of Technological Futures (Durham, NC: Duke University Press, 2019); Jack Halberstam, 
“Automating Gender: Postmodern Feminism in the Age of the Intelligent Machine,” Feminist Studies 17, no. 3 
(1991): 439–60; Luciana Parisi, Abstract Sex: Philosophy, Bio-Technology and the Mutations of Desire (London: 
Continuum, 2004), 7–13.

26.  María Lugones, “Heterosexualism and the Colonial/Modern Gender System,” Hypatia 22, no. 1 (2007): 



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   104

A critical approach to Islamicate technology, in sum, must abandon, or at least suspend, 
the rhetoric of the golden age as well as any seemingly utilitarian, apolitical understanding 
of technology, for both come laden with ideological biases—such as the teleology of Western 
scientific supremacy and the search for technological mastery—that have been connected 
to the institution of slavery and patriarchal norms. Attending to the intertwinement of 
technology and slavery is a particularly urgent, and rather obvious, necessity in the case 
of automated jawārī, given their literal conflation of female body, enslaved labor, and 
mechanized motion. Characterized as enslaved women, these self-moving devices challenge 
the paradigm of the useful, practical machine through figurative representation by drawing 
attention to its possible link to slavery. At the same time, it would not be entirely accurate 
to consider these machines simply as early robots, anticipating the mechanical slaves that 
were produced under colonial modernity and racial capitalism, for this would erase their 
historical specificities, including the question of whether the concept of technology—a 
modern invention, loaded with problematic assumptions—can be applied to premodern 
examples. This is, then, this essay’s challenge: pinpointing the politics of automated jawārī 
while raising the question of how to historicize both technology and its objectifying effects.

Automated Jawārī 

Across all of the medieval Arabic treatises I was able to consult, the only female figures 
represented in both texts and images are jawārī. Combined with aspects of posture and 
action, the systematic use of the term jāriya points to enslaved status. In other words, 
medieval automata foregrounded unfree, rather than free, women. In some instances, 
the jāriya was, in fact, the only human figure represented, often placed at the center 
of a singular, free-standing water clock. Two main types of automated jawārī can be 
distinguished in both technical and literary sources, sometimes accompanied with images: 
one appears in relation to timekeepers and the other in the form of liquid-dispensing 
devices.27 Because the technical literature has been comparatively neglected in relation to 
social history, especially forced labor, I focus here on treatises on the mechanical arts—their 
earliest surviving manuscripts, which date to the thirteenth and fourteenth centuries—with 
sporadic references to literary texts.28

The first category of automated jawārī relates to horology. In most cases, a statue 
shaped as a jāriya encased a mechanism for discharging weights, one that was commonly 
associated with clocks. Typically, a weight would move through a figurative effigy and 

186–209.
27.  To my knowledge, there has been only a handful of scholarly mentions of automated jawārī, all based 

on historiographical and poetic sources, rather than technical ones; see D. S. Rice, “A Drawing of the Fatimid 
Period,” Bulletin of the School of Oriental and African Studies 21 (1958): 31–39, at 37; Marius Canard, “Quelques 
aspects de la vie sociale en Syrie et Jazīra au dixième siècle d’après les poètes de la cour Ḥamdanide,” in Arabic 
and Islamic Studies in Honour of Hamilton A. R. Gibb, ed. G. Makdisi, 168–90 (Leiden: Brill, 1965), at 185.

28.  A descriptive list of the medieval Arabic engineering manuals that have survived is provided in Donald 
Hill, “Arabic Mechanical Engineering: Survey of the Historical Sources,” Arabic Sciences and Philosophy 1 
(1991): 167–86.



105  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

come out of its mouth; dropping in a container, it marked the passage of the hours 
through both animation and sound. The figures represented on such devices included 
animals, often birds, like the two ball-releasing falcons of the monumental twelfth-
century clock (no longer extant) of the Great Mosque of Damascus and the crows that 
feature in the timekeepers of the influential treatise of Pseudo-Archimedes, likely a 
medieval combination of short works in Greek, Persian, and Arabic.29 Equally common 
was the motif of the ball-throwing jāriya. An example from Malta was described by the 
cosmographer Zakariyyāʾ al-Qazwīnī (ca. 1203–83) in his Āthār al-bilād wa-akhbār al-ʿibād 
(Monuments of the lands and historical traditions about their peoples) as a “female slave 
(jāriya) that throws pellets (al-ṣanj).”30 Because Roger II, the twelfth-century Norman 
king of Sicily, had just been in Malta when he commissioned a water clock for his palace 
in Palermo, Michele Amari suggested that the Sicilian clock, too, may have represented a 
female figure, ejecting balls from her mouth whenever an hour had passed.31

More examples can be found in an eleventh-century Andalusian treatise, Kitāb al-Asrār 
fī natāʾij al-afkār (The book of secrets in the results of ideas), by Ibn Khalaf al-Murādī.32 
A third of al-Murādī’s contrivances involve one or two female personages characterized 
as jāriya in the text, and five of these devices are water clocks. In three timepieces, the 
jāriya throws a pebble (ḥaṣāt).33 As is often the case with al-Murādī’s machines, the first 
such specimen stages a little scene. At the top of the hour, an astrolabe holder (ṣāḥib 
al-asṭrulāb) turns his head toward the jāriya, whose jaws then open, dropping a weight. 
The next example boasts a more complex choreography, involving a third, mediating 
statue. Animation is again triggered by the motion of an astrolabe holder, who turns his 
head toward the generic statue (timthāl), which is standing next to a plate. The latter 
moves toward the female figure, and as soon as she ejects a pebble, he collects it with his 
hand, then brings it back to the plate. The third timekeeping device is the simplest of all 
three: when an hour has elapsed, a jāriya comes out of a door, expelling a pebble from her 
mouth.

Two other water clocks in al-Murādī’s treatise contain jawārī, but the figures materialize 
in pairs, and they do not emit weights.34 In one device, the gates open at every hour, 

29.  For the Damascus clock, see Finbarr Barry Flood, The Great Mosque of Damascus: Studies on the Makings 
of an Umayyad Visual Culture (Leiden: Brill, 2001), chapter 4. On the identification of “Archimedes” as “Pseudo-
Archimedes,” see Donald Hill, “The Pseudo-Archimedes Treatise,” in his Arabic Water-Clocks, 15–35 (Aleppo: 
Institute for the History of Arabic Science, 1981). For an English translation of the Pseudo-Archimedes, see 
Donald Hill, Kitāb Arshimīdas fī ʿamal al-binkamāt (London: Turner & Devereux, 1976). 

30.  Zakariyyāʾ b. Muḥammad al-Qazwīnī, Āthār al-bilād wa-akhbār al-ʿibād (Beirut: Dār Ṣādir, 1960), 557. 
31.  Michele Amari, Le epigrafe arabiche di Sicilia, trascritte, tradotte e illustrate (Palermo: L. Pedone-Lauriel, 

1971), 29–39.
32.  An Arabic edition and English translation, together with a facsimile of the only surviving manuscript, 

can be found in Ibn Khalaf al-Murādī, Kitāb al-Asrār fī natāʾij al-afkār (Milan: Leonardo3, 2018); the publication 
consists of three unnumbered volumes. 

33.  Devices 10, 11, and 20.
34.  Devices 12 and 13.



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   106

revealing two female characters (jāriyatān); as soon as they start moving forward, a Black 
man (aswad) with a knife emerges between them. In the other water clock, the two jawārī 
come out of the open entryway, this time accompanied by two snakes (ḥayyatān) throwing 
pebbles, while an astrolabe holder stands near them. Jawārī thus appear in water clocks 
even when they do not house the ball-releasing feature. This is known from literary sources 
as well. One example is the fourteenth-century Zayyanid clock at Tlemcen (Tilimsān) in 
modern Algeria that included a jāriya: every hour, the jāriya would emerge from a door 
with a piece of paper in her hand, presenting the number of the hours in versified form.35 

A jāriya also turns up in a short, unpublished treatise that was likely related to the 
Pseudo-Archimedes and that describes a water clock.36 Interestingly enough, although the 
text does not seem to mention a female personage, a diagram places a figure labeled jāriya 
in the center of the device, surrounded by three birds, each containing a ball-dropping 
mechanism (see Fig. 1). From the drawing, one may infer that the jāriya acted as an axle 
(described as a rotating plate in the manuscript, to the best of my understanding); she was 
placed at the center of a tangle of strings (sing. khayṭ), each passing over a pulley (bakra) 
that also connected to the head of a bird as well as to the mechanism, seen at the top, that 
released the pebbles (banādiq). Whatever the jāriya’s exact function, this example attests 
to the diffused, consistent association of jawārī with horological devices within and beyond 
the textual record and across the medieval Islamicate Mediterranean, from Muslim Spain to 
Malta to North Africa and the Middle East.

The second type of automated jawārī consists of serving devices, as shown by at least 
two treatises. One is Ibn al-Razzāz al-Jazarī’s famous manual al-Jāmiʿ bayn al-ʿilm wa-l-
ʿamal al-nāfiʿ fī ṣināʿat al-ḥiyal (A compendium on the theory and useful practice for the 
fabrication of machines), completed in 1206 in Artuqid Anatolia.37 As I have discussed 
elsewhere, with its emphasis on automated slaves the compendium conveyed the scope and 
importance of courtly slavery under Artuqid rule while echoing a transregional, courtly 
imaginary of automated servants.38 Most of the mechanical attendants are free-standing, 
android-like male figures, but one exception is a wine-serving device representing a woman 
whom the text refers to as a jāriya.39 The machine consists of a cupboard with the statue 
of a jāriya inside it and a wine reservoir at the top (Fig. 2). Through a tipping bucket and a 
trough, wine flows from the reservoir into the glass the jāriya holds. As soon as the vessel 
fills up, the figure, who is mounted on wheels, rolls out of the cupboard, presenting wine to 
the royal patron.

35.  Flood, Great Mosque, 135–37.
36.  Oxford, Bodleian Library, MS Marsh 699, fols. 11v–13.
37.  For the Arabic edition, see Ibn al-Razzāz al-Jazarī, al-Jāmiʿ bayn al-ʿilm wa-l-ʿamal al-nāfiʿ fī ṣināʿat 

al-ḥiyal, ed. Aḥmad Yūsuf al-Ḥasan (Aleppo: Institute for the History of Arabic Science, 1979). For an English 
translation, see Ibn al-Razzāz al-Jazarī, The Book of Knowledge of Ingenious Mechanical Devices, trans. Donald 
Hill (Dordrecht: Reidel, 1974). For both text and images, I have relied on the earliest known manuscript copy, 
dated 1206 (Istanbul, Topkapı Palace Library, MS Ahmet III 3472).

38.  Balafrej, “Automated Slaves.”
39.  Al-Jazarī, Book of Knowledge, 125–26.



107  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

The second manual is the Arabic Philo, an Arabic translation of the Pneumatics of Philo 
of Byzantium that also contains medieval additions, including two chapters—numbered 
thirty and thirty-six—with jawārī.40 Most of the artifacts featured in the work are jars, cups, 
washstands, and water-lifting machines devoid of figurative motifs, which makes the devices 
with enslaved women further stand out. This was noted in the early twentieth century by 
Bernard Carra de Vaux, who suggested that chapter 30 might have been directly inspired 
by al-Jazarī’s treatise, a point recently repeated by Sylvia Berryman.41 Chapter 30 presents 
an anthropomorphic wine-serving machine characterized as a jāriya and represented with 
a jug in her right hand (Fig. 3).42 There is a split reservoir inside the figure’s chest, with 
water in one half and wine in the other; tubes connect both parts to the jug, while air pipes 
link the top of the reservoir to the left arm. As soon as one places a cup on the left palm, 
the hand descends, pushing air into the wine container and thus wine into the ewer. As the 
hand comes down further, opening up the air pipe of the water container, wine is replaced 
by water. 

Chapter 36 of the Arabic Philo features a vessel for ablutions with a figure in the middle 
that is called a jāriya in the text.43 In all three surviving medieval copies, the diagram 
follows a similar pattern: we see a crescent-shaped basin placed on top of a water reservoir 
with two embedded containers, the figure of the jāriya floating in the inner vessel.44 Water 
trickles down from the basin into the container, prompting the jāriya to rise with the level 
of the water and to come out of the vessel, pushing the lid. When the main container fills 
up, water is dispensed through the faucet at the bottom; as it subsides, the figure moves 
downward. Although the jāriya does not itself perform any action, functioning rather as a 
float, as in chapter 30 its inclusion here in a liquid-serving device was likely premised on 
jawārī’s association with housework.   

40.  At least three medieval manuscripts of the Arabic Philo have survived: Oxford, Bodleian Library, MS 
Marsh 669, and Istanbul, Süleymaniye Library, MSS Ayasofya 3713 and 2755. For an Arabic edition, accompanied 
by a French translation, see Philo of Byzantium, Le livre des appareils pneumatiques et machines hydrauliques 
par Philon de Byzance, ed. and trans. Bernard Carra de Vaux (Paris: Klincksieck, 1902). I follow Carra de Vaux’s 
numbering of the chapters. For an English translation, see Frank D. Prager, Philo of Byzantium, Pneumatica: 
The First Treatise on Experimental Physics, Western Version and Eastern Version (Wiesbaden: Reichert, 1974).

41.  Bernard Carra de Vaux, “Les pneumatiques de Philon de Byzance,” Notices et extraits des manuscrits de 
la Bibliothèque nationale 39 (1903): 27–229, at 35; Sylvia Berryman, The Mechanical Hypothesis in Ancient Greek 
Natural Philosophy (Cambridge: Cambridge University Press, 2013), 162–63. Various chapters refer to a mosque, 
ablutions, and a minaret (respectively, chapters 63, 53, and 58), confirming that the book’s last two thirds were 
a mix of medieval Islamicate influences.

42.  Prager, Philo, 176–77 (English translation); Philo of Byzantium, Le livre des appareils pneumatiques, 
51–52 (Arabic edition), 135–37 (French translation). The chapter is illustrated in Oxford, Bodleian Library, MS 
Marsh 669, fols. 29v–30.

43.  Prager, Philo, 189–90 (English translation); Philo of Byzantium, Le livre des appareils pneumatiques, 61 
(Arabic edition), 147–48 (French translation). The chapter is illustrated in Oxford, Bodleian Library, MS Marsh 
669, fol. 37v; Istanbul, Süleymaniye Library, MS Ayasofya 3713, fol. 46v. 

44.  For an example, see Balafrej, “Automated Slaves,” 763.



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   108

Fig. 1. Clock with birds and a jāriya. Folio from an anonymous Arabic treatise, ca. 1300–
1500, ink on paper, 29 cm × 18 cm. Oxford: Bodleian Library (MS Marsh 669, fol. 13).



109  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

Fig. 2. Automated jāriya. Folio from a copy of al-Jāmiʿ bayn al-ʿilm wa-l-ʿamal al-nāfiʿ fī 
ṣināʿat al-ḥiyal of Ibn al-Razzāz al-Jazarī, probably Āmid (modern-day Diyarbakır), 1206, 

ink and opaque watercolor on paper, 33 cm × 24 cm. Istanbul: Topkapı Palace Library (MS 
Ahmet III 3472, fol. 113v).



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   110

Fig. 3. Automated jāriya. Folio from an anthology of mechanical works containing an 
Arabic translation of the Pneumatics of Philo of Byzantium, ca. 1300–1500, ink on paper, 29 

cm × 18 cm. Oxford: Bodleian Library (MS Marsh 669, fol. 29v).



111  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

The motif of the automated female slave, then, circulated rather broadly, from al-Andalus 
to Anatolia, within and beyond courtly settings—though likely always among learned, 
intellectual circles. This is testified to by the manuscripts themselves, as well their images, 
which show significant variations in terms of function, style, and level of artistry. In the case 
of the medieval Arabic Philo (Fig. 3) or the short, anonymous treatise that describes a water 
clock (Fig. 1), the copies examined do not stem from princely patronage but must have been 
made for scholars and practitioners; in addition to their lack of colophons and dedications, 
they are not illuminated and their illustrations are linear, explanatory drawings, simpler 
in appearance than the painterly images one usually finds in royal manuscripts (Fig. 2). 
Textual representations, moreover, suggest both restricted and public contexts of display, 
especially for the water clocks. While some of the devices, such as al-Jazarī’s wine-dispensing 
automaton, were designed for personal, princely consumption, others, such as the water 
clocks, were likely destined for a wider audience.45 This was probably also the case with 
al-Murādī’s contrivances. Owing to their sheer size, and given their use of waterwheels 
as primary movers, it is possible that they were meant, or at least imagined, for a public 
space, like the Damascus water clock, which was conceived for the mosque’s gate. They also 
resonated, in scale and mechanism, with the kind of public, monumental hydraulic works 
described in Andalusian sources, such as the clepsydras that were erected in eleventh-
century Toledo by the Tagus river.46 

Across these contexts, the link between the female figures these devices represented 
and domestic slavery must have been evident. It is most obviously signified through the 
sources’ insistence on the word jāriya—consistently used in all of the texts and images I 
have perused—as well as the devices’ functions, which mimic service work. 47 The machines, 
moreover, do not simply feature enslaved women, nor do they necessarily reflect historical 
realities. While echoing female slavery’s presence in medieval Islam, as representations the 
automata instill and enforce a system of relations, particularly when it comes to intersections 
of gender, class, labor, and visibility. Here, these intersections are rendered through jawārī’s 
visibility both as representation’s referents and in the story world the machines fashion. The 
very fact that jawārī, rather than free women, were chosen as a motif conveys a normative 
link between visibility, femininity, and enslaved status. This link is heightened within some 
of the machines’ narratives, notably by the female figures’ occasional movement through a 
door—as in three of al-Murādī’s specimens and the Zayyanid clock. 

45.  The circumstances of its composition are unknown; the text has survived in only one codex, completed 
in 1266, presumably at the court of Alfonso X in Toledo (Florence, Biblioteca Medicea Laurenziana, MS Orientale 
152). On this manuscript, including the long-debated question of its authorship, see Julio Samsó, Las ciencias 
de los antiguos en al-Ándalus (Madrid: MAPFRE, 1992), 249–57; Donald Hill, “An Andalusian Treatise of the 
5th/11th Century,” in Arabic Water-Clocks, 36–46; Juan Vernet, “Texto árabe de la corte de Alfonso X el Sabio 
X,” Al-Andalus: Revista de las Escuelas de estudios árabes de Madrid y Granada 43, no. 2 (1978): 405–33.

46.  Donald Hill, “The Toledo Water-Clocks of c. 1075,” History of Technology 16 (1994): 62–71.
47.  This lexical consistency is striking, since many other words would have been possible—such as fatāt 

for a young woman, imraʾa and the plural nisāʾ for adult women, or ama, ʿabda, or mamlūka for female slaves. 
But it also makes sense, given the widespread preference for euphemistic language among slaveholders, as 
exemplified by the frequent use of such denominations as jāriya and ghulām that could be understood as young 
woman or young man (see, for example, Marmon, “Intersections of Gender,” 202). 



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   112

The association of female slavery with visibility was not incidental: it appears in a 
number of literary, legal, and medical sources. Jawārī’s mobility was an integral part of 
their unfreedom in edifying anecdotes of adab or belles lettres, as Nadia Maria El-Cheikh 
has shown.48 Baber Johansen has argued that female slaves, unlike free women, were not 
necessarily subjected to gendered forms of social distance in Sunni Islamic law; their 
exemption was connected to their activities, which required that they leave the house, 
putting them “in everybody’s reach and touch.”49 More recently, Omar Anchassi has focused 
on sartorial distinctions between free and unfree women in Islamic law, showing that 
female slaves’ visibility—including the fact that generally they did not have to wear a veil—
was linked to their status as property.50 

Jawārī’s exposure to the public was described in the most brutal way in the medically 
oriented genre of slave-buying advice manuals, which provided instructions for the display 
and inspection of the enslaved at the market.51 The inspection, a process of profanation 
known as ibtidāl, worked not only to detect injury or disease but also to humiliate the 
enslaved and to mark their distinction from the free, whose bodies, unlike the slaves’, were 
protected by the principle of ḥurma or inviolability, as Hannah Barker has noted.52 Further, 
the presupposition that jawārī had greater access to a variety of social spaces may have 
supported the sexual dimension of possession, which was seen as a male privilege—the term 
jāriya, in fact, was openly associated with sexual exploitation.53 The link between jawārī and 
visibility, then, was neither benign nor peripheral, and it did not simply reflect a practical 
or a social reality; rather, it was more likely a potent cultural perception, one that facilitated 
jawārī’s commodification and exploitation. 

Gender and class-based hierarchies were clearly at play in the dynamic the automata built 
between themselves and their users. A few of them were explicitly described in relation to 
male, royal patrons (though this does not mean that women or lower groups were absent 
from the audience), including Zayyanid, Norman, and Artuqid ones, suggesting that the 
jawārī of such artistic and mechanical representations were symptomatic of these patrons’ 
taste and needs. Automated jawārī enhanced their users’ power position by materializing 
and reiterating some of the norms that were associated with female domestic slavery, 

48.  Nadia Maria El-Cheikh, “Women’s History: A Study of al-Tanukhi,” in Writing the Feminine: Women in 
Arab Sources, ed. Randi Deguilhem and Manuela Marín, 129–48 (London: I. B. Tauris, 2002).  

49.  Baber Johansen, “The Valorization of the Human Body in Muslim Sunni Law,” Princeton Papers in Near 
Eastern Studies 4 (1996): 71–112, at 79–80.

50.  Omar Anchassi, “Status Distinctions and Sartorial Difference: Slavery, Sexual Ethics, and the Social Logic 
of Veiling in Islamic Law,” Islamic Law and Society 28 (2021): 125–55.

51.  A well-known example is Ibn Buṭlān, “Risāla jāmiʿa li-funūn nāfiʿa fī shirā al-raqīq wa-taqlīb al-ʿabīd,” 
in Nawādir al-makhṭūṭāt, ed. ʿAbd al-Salām Muḥammad Hārūn, 1:351–89 (Cairo: Maṭbaʿat Lajnat al-Taʾlīf wa-l-
Tarjama wa-l-Nashr, 1951). On the genre and for more references, see Hagedorn, Domestic Slavery, 61–78; Barker, 
That Most Precious Merchandise, 4; idem, “Purchasing a Slave in Fourteenth-Century Cairo: Ibn al-Akfānī’s Book 
of Observation and Inspection in the Examination of Slaves,” Mamlūk Studies Review 19 (2016): 1–24. 

52.  Barker, “Purchasing a Slave,” 2–3.
53.  Marmon, “Intersections of Sex”; Antonella Ghersetti, “The Representation of Slave Girls in a Physiognomic 

Text of the Fourteenth Century,” Mamlūk Studies Review 21 (2018): 21–45.



113  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

including hypervisibility, as seen earlier. Across the foregoing treatises, in addition to 
visibility, enslaved womanhood is associated with containment and serviceability through 
the representation of the female body as vessel and vector, as in the liquid-dispensing device 
and the ball-throwing jāriya. There may also be an expectation that female housework 
should be disciplined and regimented, given the connection drawn between animated jawārī 
and the labor of keeping time. As liquid dispensers and clocks, the figurative machines, 
therefore, did not so much engage historical realities as project codes and expectations, the 
conventions that made jawārī culturally intelligible.

Power imbalance characterizes the machines’ diegetic level as well, especially when the 
jāriya accompanies a male figure of higher rank, such as the astrolabe holder in al-Murādī’s 
water clocks. The man carrying the astrolabe is endowed with the capacity to look, since 
he turns his head, the very gesture that sets the whole machinery into motion. The use of 
the astrolabe further implies that he can measure and analyze time, not simply channel it. 
By contrast, the jāriya does not embody any kind of analytical skill. These representations, 
then, may have had a disparaging, essentializing effect, framing women as physical, inferior 
entities. This view is found in a wide array of sources. Much has been written about the legal 
perspective in particular—about how certain normative texts frame wifehood as slavery and 
household chores as mindless, menial labor, primarily fit for such inferior agents as women 
and slaves.54 There is also evidence that this thinking was directly applied to automata: the 
tenth-century poet al-Mutanabbī, for example, described a mechanical jāriya as “a body 
devoid of a soul” (jāriya mā li-jismihā rūḥ).55

Yet the thesis of the reifying, patriarchal machine is not completely satisfying, especially 
if objectification is understood as the unidirectional, terminal transformation of subject into 
object (in the representational order), for example through the male gaze. In that sense, 
the idea of the automaton as objectifying amplifies, more than challenges, functionalism’s 
ideological principles, with its assumption of a dominating subject governing a body of 
servile, useful objects. Though it may seem intuitive and transhistorical, this reading 
remains quite specific to colonial modernity, which intensified the dual search for humanless 
technology and disposable labor, and it is tied to the rise of the modern subject, as noted 
again in this essay’s conclusion. Fear of the machine as repressive and belittling, moreover, 
has been shown to act as the reversed double of techno-utopianism. As Chude-Sokei has 
indicated, the rhetoric of the threatening, malevolent machine might indeed manifest 
an anxiety of reversal and replacement that underlies modern slavery and colonialism, 
repeating a white fear of the rebellious slave.56 

To say that automata represented jawārī as objects—that is, as physical, passive entities 
produced by another’s viewpoint—is to enhance a conception of technical equipment as 
subservient, one of functionalism’s underlying assumptions. We must therefore consider 

54.  Marion Holmes Katz, Wives and Work: Islamic Law and Ethics before Modernity (New York: Columbia 
University Press, 2022); Kecia Ali, Marriage and Slavery in Early Islam (Cambridge, MA: Harvard University 
Press, 2010); Leila Ahmed, Women and Gender in Islam (New Haven, CT: Yale University Press, 1992).

55.  Al-Mutanabbī, Sharḥ dīwān al-Mutanabbī, ed. ʿAbd al-Raḥmān al-Barqūqī, vol. 1 (Cairo: Maṭbaʿat 
al-Saʿāda, [1938]), 380. 

56.  Chude-Sokei, Sound of Culture, esp. chapter 1.



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   114

broadening our notions of both technology and objectification beyond the double 
conflating discourse of obedient tool and machine-like worker. This is not to suggest that 
automated jawārī were not objectifying, in the loose sense that they were stereotyping and 
dehumanizing, notably through their emphasis on visibility, acquiescence, and usability. 
Rather, the goal is to clear a space for expanding our interpretive, conceptual possibilities 
beyond the decidedly modern, polar opposition of subject and object. 

Before and Beyond Technology

My working argument is that the aforementioned automata depicted jawārī not so much 
as objects or even tools but as “instrumental”—that is, as toollike but also as active, dynamic 
entities constituting the ground for, rather than an object of, instrumentality. I came to 
this understanding by enlarging the corpus of analysis to encompass portrayals of enslaved 
women beyond the technical artifacts that usually qualify as “Islamic technology.” Given 
the link between utilitarianism and the binaries of master and slave, user and tool, I also 
had to step away from the subject-object hierarchy, the relation whereby the object exists 
mainly in so far as it is the result of a subject’s perspective—an issue I address further in the 
conclusion. 

A major reason for opening up the discussion to other genres is that in the period under 
consideration, technology had not yet fully emerged as its own field of inquiry. Undeniably, 
though the word itself did not exist, in its triple meaning as a branch of knowledge, a 
sphere of activity that applies such knowledge for practical reasons, and the results of such 
application, such as machinery and tools, technology did have premodern antecedents, 
including ḥiyal. Still, the term “technology” can be misleading, for it mainly refers to self-
contained, external artifacts; it is also enmeshed with functionalism, given the common 
assumption of technology’s usefulness and practicality. The scope of medieval technology 
was wider. Ḥiyal, for example, which literally means ruses or stratagems, included 
mechanical contrivances but also fictions and subterfuges in legal and political domains.57 
Scholars such as Tim Ingold have argued that technology is, in fact, a distinctly modern 
concept. The emergence of “technology,” a term not used before the seventeenth century, 
marked a shift from techne—understood as skilled labor, in Arabic ṣināʿa—to an extension 
of the machinic to various forms of life and existence, including the universe, which began 
to be seen as a vast machine whose rational principles could be understood scientifically.58 

To avoid all of these connotations, I turn to the Arabic word for instrument, āla, as it 
proves more capacious and more porous than today’s notion of technology and thus allows 
us to consider a wider, cross-genre archive beyond the automaton that literally equates 
servant and machine. For one, āla commonly applied to both automata and nonmechanized 
forms of equipment. Ālāt encompassed self-moving devices as well as the tools used in a 
craft, such as the scalpel in medicine and the lute in music. Moreover, bodily parts and 
organs, too, could be defined as tools and instruments; the lung, for example, enables the 

57.  Donald Hill, “Ḥiyal,” in Encyclopaedia of Islam, 2nd ed., ed. P. Bearman et al. (Leiden: Brill Online).
58.  Ingold, “Eight Themes,” 130–31. On ṣināʿa, see Adam Mestyan, “Arabic Lexicography and European 

Aesthetics: The Origins of Fann,” Muqarnas 28 (2011): 69–100.



115  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

action of breathing by serving as a vector for air.59 Āla cuts across the distinctions of machine 
and tool, artifice and body. Instrumentality—the quality of serving as an instrument in 
processes of making, carrying, and transmitting—could be shared among man-made and 
natural entities; at least, it entailed no definitive distinction, let alone opposition, between 
intending subject and intended object. This also suggests that instrumentality could surface 
outside of the mechanical arts as a form of technological thinking, associating jawārī with 
technical tasks such as the acts of serving and carrying. 

One type of narrative in which jawārī appear as instrumental figures concerns mystical, 
Ṣūfī history. It usually consists of a providential, instructive encounter between a male 
Ṣūfī aspirant and an enslaved woman who is referred to as a jāriya. Examples turn up in 
Kitāb Muthīr al-ʿazm al-sākin ilā ashraf al-amākin (A book inciting firm resolution for 
[visiting] the noblest of places), a treatise on the ḥajj, the Muslim pilgrimage to Mecca, by 
ʿAbd al-Raḥmān Ibn al-Jawzī (ca. 1116–1200), who was an influential and prolific historian, 
judge, and preacher in Baghdad.60 In one passage, Sarī al-Saqaṭī, a ninth-century Ṣūfī saint 
of Baghdad, recounts his encounter with a jāriya ḥabashiyya (Abyssinian or Ethiopian slave) 
while on his way to Mecca via Kufa, Iraq.61 The figure’s enslaved status is later confirmed 
by her self-identification as ʿabda, an unequivocal word for a female slave. Near Kufa, Sarī 
al-Saqaṭī learns from the jāriya that she, too, is headed for the ḥajj. He warns her that the 
journey is long, to which she responds that distance is relative—that for the Lover (an 
allegorical concept for the seeker of divine wisdom, the object of the quest being the Beloved, 
who stands for God), the destination is actually near. Once at the Kaʿba, al-Saqaṭī finds the 
jāriya there already, performing the ṭawāf (the anticlockwise, ritualistic circumambulation 
of the Kaʿba); she calls him out for being surprised, adding that her vulnerability is precisely 
what allowed her to move so fast, carried by divine strength. 

The passage that immediately follows in Ibn al-Jawzī’s Muthīr al-ʿazm features another 
Ṣūfī saint, al-Shiblī (861–946). Here, too, the narrative emphasis is on the mystic’s encounter 
with a jāriya ḥabashiyya, in this case in the countryside.62 Ibn al-Jawzī’s insistence on 
the ethnic characterization of ḥabashiyya racializes the jawārī’s enslaved status while 
possibly confirming the historical predominance of domestic workers from Nubia (al-Nūba), 
East Africa (al-Zanj), and Ethiopia (al-Ḥabash) in the medieval Middle East, as attested by 
such documents of the slave trade as deeds of sale and by the literary evidence of slave-

59.  On bodily organs as tools and instruments, see, for example, these tenth-century Arabic sources: Abū 
Zakariyyā Yaḥyā Ibn ʿAdī, “Fī ithbāt ṭabīʿat al-mumkin,” ed. Carl Ehrig-Eggert, Zeitschrift für Geschichte der 
arabischen-islamischen Wissenschaften 5 (1989): 63–97 [Arabic], at 70; Ikhwān al-Ṣafā, “On the Practical Crafts,” 
On Composition and the Arts, epistles 6–8 of Epistles of the Brethren of Purity, ed. and trans. Nader El-Bizri and 
Godefroid de Callataÿ (Oxford: Oxford University Press, 2018), esp. chapter 6.

60.  For a biography of Ibn al-Jawzī and a list of his books, see Carl Brockelmann, History of the Arabic Written 
Tradition, trans. Joep Lameer, vol. 1 (Leiden: Brill, 2016), 579–85. On this particular book, also see Joseph de 
Somogyi, “Ibn al-Jauzī’s Handbook on the Makkan Pilgrimage,” Journal of the Royal Asiatic Society 70, no. 4 
(1938): 541–46.

61.  ʿAbd al-Raḥmān Ibn al-Jawzī, Kitāb Muthīr al-ʿazm al-sākin ilā ashraf al-amākin, ed. Marzūq ʿAlī Ibrāhīm, 
vol. 2 (Giza: Dār al-Rāya, 1990), 194–95.

62.  Ibid., 195.



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   116

buying manuals.63 A dialogue ensues, wherein al-Shiblī asks the jāriya questions about her 
trajectory’s starting point, destination, and motivation, to which she invariably responds: 
the Beloved (al-ḥabīb). This makes her a Ṣūfī seeker and an enlightened teacher for the 
saint, whom she instructs about the necessity of silence while on the spiritual path, the 
reason being the ineffability of the divine, the impossibility of describing God before the 
quest is finished. 

A similar storyline—a Ṣūfī encounters an inspired jāriya—can be found in Tarjumān 
al-ashwāq (Interpreter of desires) by the famous Andalusi scholar and mystic Ibn ʿArabī 
(1165–1240). The Ṣūfī narrator recounts his meeting with Qurrat al-ʿAyn, a jāriya from 
al-Rūm (meaning she was either Greek or Anatolian or both).64 While at the Kaʿba, the poet 
utters a set of verses, when a jāriya appears, confronting him with a critical reading: she 
objects to his rational bent, pointing out the limits of skepticism for any true understanding 
of the divine.65 In addition to her characterization as jāriya, Qurrat al-ʿAyn’s name, too, 
may point to slave status, since it was sometimes given to unfree women.66 The reality of 
the slave trade may be further indicated by her rūmī origin.67 This gives Qurrat al-ʿAyn 
an ambivalent position, that of an inspired servant, simultaneously female slave and 
transmitter of instruction.

Like the engineer’s prototypes, these stories were authored by elite male individuals; 
they may have functioned also to underscore male superiority. Take, for example, Ibn 
al-Jawzī. This prolific scholar also wrote a book on Ṣūfī saints, Ṣifat al-ṣafwa (Characteristics 
of the elite), in which he included a number of female mystics, relying and expanding on 
earlier accounts, notably the biographical book that Abū ʿAbd al-Raḥmān al-Sulamī (d. 
1021) devoted to early female mystics, Dhikr al-niswa al-mutaʿabbidāt al-ṣūfiyyāt (Memorial 
of pious Ṣūfī women).68 Portrayals of Ṣūfī women were rather scarce, and they were less 
detailed than those of Ṣūfī men, a process of discursive marginalization that aimed to center 
normative, androcentric Sufism. Aisha Geissinger has also emphasized “the centrality of the 
gaze of the pious free elite Sunni Muslim male” in Ibn al-Jawzī’s biographies, showing that 
the male gaze both foregrounded and diminished female Ṣūfīs and ultimately downplayed 

63.  Goitein, “Slaves and Slave Girls,” 8; Craig Perry, “The Daily Life of Slaves and the Global Reach of Slavery 
in Medieval Egypt, 969–1250 CE” (PhD diss., Emory University, 2014), 39–41. On the ethnic categories of Nubians, 
East Africans, and Ethiopians, see Hagedorn, Domestic Slavery, 98–107; for deeds of sale that use them, see 
Ragib, Actes de vente d’esclaves.

64.  Ibn ʿArabī, Tarjumān al-ashwāq, trans. Reynold Nicholson (London: Royal Asiatic Society, 1911), 14. 
65.  For an analysis of the Ṣūfī tenor of the jāriya’s criticism, see Saʿdiyya Shaikh, Ṣūfī Narratives of Intimacy: 

Ibn ʿ Arabī, Gender, and Sexuality (Chapel Hill: University of North Carolina Press, 2012), chapter 2; Pablo Beneito, 
“Qurrat al-ʿAyn, the Maiden of the Kaʿba: On the Themenophany Inspiring Ibn ʿArabī’s Tarjumān,” Religions 12 
(2021): 23–40.

66.  For an example, see Ibn al-Sāʿī, Consorts of the Caliphs: Women and the Court of Baghdad, trans. Shawkat 
M. Toorawa et al. (New York: New York University Press, 2015), 57.

67.  On Black Sea slaves, see Barker, That Most Precious Merchandise; Felicia Roşu, ed., Slavery in the Black 
Sea Region, c. 900–1900: Forms of Unfreedom at the Intersection between Christianity and Islam (Leiden: Brill, 
2021).

68.  Rkia Cornell, Early Sufi Women: Dhikr an-niswa al-mutaʿabbidāt aṣ ṣūfiyyāt (Louisville: Fons Vitae, 1999).



117  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

their authority.69 Given his tendency to uphold the social, moral, and spiritual domination 
of male religious figures, Ibn al-Jawzī’s inclusion of jawārī in his book on pilgrimage, then, 
should not be taken as a celebration of female mysticism but rather as a narrative strategy, 
designed to facilitate and dramatize the male protagonist’s initiation and transformation. 

The depiction of jawārī as vectors of transmission (in the sense that they served as 
messengers, delivering spiritual insight and acting as a link between earthly and 
otherworldly domains) was neither liberatory nor redeeming; rather, it may have been 
congruent with societal views of enslaved labor. There is evidence, indeed, that enslaved 
domestic servants, both male and female, were expected to work as connectors and 
transmitters. Using the documents of the Cairo Geniza, Craig Perry has pointed out the 
mediating role of the ghulām, the jāriya’s male equivalent, likening him to a factotum or 
business agent.70 A similar function was attributed to enslaved female servants. This was 
shown by Eve Krakowski, who noted jawārī’s place in the life of freeborn Jewish women 
of medieval Egypt as their owners’ proxies, linking the domestic and public worlds.71 
Geniza documents can further speak to the jāriya’s position as a sort of factotum in Egypt 
between the eleventh and thirteenth centuries. Jawārī were often dehumanized as nameless 
commodities in dowry lists, legal dispute documents, and in the deeds recording their sale.72 
But other instances yield more detailed descriptions of their duties. One note, for example, 
requests its addressee to send five dinars with a jāriya.73 Another one directs its recipient 
to send a jāriya as a messenger.74 Meanwhile, a legal fragment indicates that a jāriya may 
have had to make her female owner’s funeral arrangements after the latter’s death.75 None 
of these sources give us access to jawārī’s actual quotidian experiences, but they convey 
a variegated understanding of jawārī’s authorized responsibilities, with an emphasis on 
mediatory functions. 

Like documentary sources, literary accounts can shed light on representations of daily 
life, though stories that feature mundane domestic activity outside of elite circles appear 
to be rare. One exception turns up in Kitāb al-Mukhtār fī kashf al-asrār (The book of the 
selected disclosure of secrets / The book of charlatans) by the thirteenth-century author 
ʿAbd al-Raḥīm al-Jawbarī. The story tells of the narrator’s encounter with a member of 
the Banū Sāsān—a Romani tribal confederation—in Konya, Anatolia, in 1219–20 and of his 

69.  Aisha Geissinger, “Female Figures, Marginality, and Qur’anic Exegesis in Ibn al-Jawzī’s Ṣifat al-Ṣafwa,” in 
Islamic Interpretive Tradition and Gender Justice: Processes of Canonization, Subversion, and Change, ed. Nevin 
Reda and Yasmin Amin, 151–78 (Montreal: McGill-Queen’s University Press, 2020).

70.  Perry, “Daily Life of Slaves,” 12; also see Balafrej, “Domestic Slavery,” 1020–21.
71.  Eve Krakowski, Coming of Age in Medieval Egypt: Female Adolescence, Jewish Law, and Ordinary Culture 

(Princeton, NJ: Princeton University Press, 2017), 199–200.
72.  On women-owned domestic slaves, see Perry, “Daily Life of Slaves,” 95–98. Documents available online 

through the Princeton Geniza Project (PGP) include the following, held at Cambridge University Library: T-S NS 
J319 + T-S NS 190.108 + T-S NS 190.114, S. D. Goitein’s unpublished edition; T-S NS 320.50c; T-S NS J449.

73.  University of Manchester Library, JRL SERIES B 6116, available online through the PGP.
74.  Cambridge University Library, Moss. II,127.2, available online through the PGP. 
75.  Cambridge University Library, Moss. VII,129.2, available online through the PGP.

https://geniza.princeton.edu/en/documents/4316/
https://geniza.princeton.edu/en/documents/4316/
https://geniza.princeton.edu/en/documents/25428/
https://geniza.princeton.edu/en/documents/26488/
https://geniza.princeton.edu/en/documents/28829/
https://geniza.princeton.edu/en/documents/26696/
https://geniza.princeton.edu/en/documents/27013


Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   118

interest in understanding the Banū Sāsān’s success and talent in the arts of trickery.76 At 
the Sāsānī’s home, a jāriya appears a few times, carrying out a range of tasks; at dinner, she 
brings a jug of hot water with a washbasin, then food and drinks, before calling in another 
jāriya, more an entertainer than a servant, who plays an oud and then a harp. The next 
morning, the servant acts as the Sāsānī’s helper, bringing the accessories he needs for his 
subterfuge at the mosque—rags, a garbage sack, and a headband—and then scattering dust 
all over him to make him look a pauper. While her patron gets ready, she helps the guest 
again, taking him to the bathhouse and offering him food, as well as perfumes, before 
he goes to attend the host’s scam. Both this early thirteenth-century account and the 
scattered Geniza examples define the servant figure as a personal assistant, a manager, 
and an emissary while also confirming the common association of visibility and jawārī’s 
enslaved status. 

The media operations that jawārī carried out in the foregoing representations involved 
the culling and sharing of information, which itself entailed spatial and technical literacy. 
These duties, which were essential to premodern service and communication, were not 
dissimilar to those of the ghulām-factotum who, as a messenger, functioned to connect 
people and places, sometimes across vast swaths of the world, as recorded in business 
letters.77 This may explain jawārī’s inclusion in both technical and mystical writings as 
mediating, vectorial entities. This study’s jump from the mechanical arts to spiritual 
literature, then, may illuminate a continuum across genres, revealing another, transversal 
convention for jawārī’s representability: not only as mobile and hypervisible but as signs 
of instrumentality, charged with mediatory functions—as transmitters and vehicles—that 
were both vital and coerced. Technical and mystical sources find another commonality 
in the way they hybridize high endeavors, whether technological experimentation or the 
pursuit of spiritual instruction, with lowly labor. The contrast may seem surprising, but 
it likely served to highlight the power of patron and saint while revealing the (potentially 
constitutive) enmeshment of slavery with certain cultural and scientific domains.78

Instrumental Jawārī

Enslaved women or jawārī crisscrossed a variety of imaginaries, genres, and compositional 
forms, both verbal and visual, from al-Andalus to Anatolia and beyond. The presence of 
jawārī across such a range of discourses and locales reflects the importance of female 
domestic labor in medieval Islam. These representations also raise the question of their 
semantic, allegorical function, likely as a patriarchal maneuver, equating the female body, 
abstract or even forced action (either through mechanical or divinely inspired power), and 

76.  Jamāl al-Dīn ʿAbd al-Raḥīm al-Jawbarī, Kitāb al-Mukhtār fī kashf al-asrār, ed. Manuela Dengler, trans. 
Humphrey Davies (New York: New York University Press, 2020), 104–13. 

77.  Balafrej, “Domestic Slavery,” 1020–21.
78.  As such, depictions of jawārī as instrumental may partake in the subaltern genealogy of modern media 

and technology. See Markus Krajewski, The Server: A Media History from the Present to the Baroque, trans. 
Ilinca Iurascu (New Haven, CT: Yale University Press, 2018), though Krajewski does not include non-Western 
history or Black studies. I thank Tung-Hui Hu for this reference.



119  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

timely, reliable service. Yet approaching these portrayals of unfree domestic labor uniquely 
through the trope of the objectifying machine is problematic, for it centers a distinction of 
subject and object that may not have been so relevant to the medieval context. 

In order to circumvent the modern strictures of both technology and objectification, 
I have characterized the figure of the jāriya, as represented in the above corpus, as 
“instrumental,” in reference to the term āla (“instrument”) that has served to designate 
machines and instruments, including bodily organs. On the one hand, “instrumental” 
applies to something (like music) insofar as it is performed through an instrument or a 
tool, which implies the participation of a source of agency and volition that is usually 
perceived as superior. In this regard, the expression “instrumental jāriya” acknowledges 
the realities of enslavement and patriarchy, the fact that jawārī’s labor was produced under 
the patronage, often exploitative, of an owner. 

On the other hand, “instrumental” can also be employed to say of an element that 
it is significant and pivotal, that it plays an essential role. Indeed, there is a difference 
between being instrumental and being useful; this is similar to the familiar distinction 
between instrument and tool, in Arabic between āla and adāt. In common parlance a tool 
is often thought of as a basic, disposable object. By contrast, an instrument usually stands 
for an elaborate, amplifying medium, as exemplified by musical or surgical instruments. 
Consequently, “instrumental” indicates a complex form of objectification, whether in 
ālāt mutaḥarrika that bridged forced labor and technical sophistication or in the spiritual 
encounters that posited jawārī as mediators of truth, not in command of their power yet 
acting as potent, inspired intercessors. 

The notion of the “instrumental” troubles the prevalent understanding of objectification 
as an extreme, polarizing operation that transforms the world into a field of passive, 
inanimate things, as it implies a measure of activity, if not agency. As such, it also complicates 
the traditional view of domestic work as inconsequential—that is, as a form of repetitive, 
unremarkable labor, centered on the quotidian, unimportant tasks of housekeeping and 
mainly restricted to the alleged privacy of the home. Automata and other representations 
indeed heightened jawārī’s functions as narrative, spatial, and social connectors and as 
vectors for a range of substances, both material and abstract, including liquids, time, 
and knowledge. Unfree service appeared as dynamic and productive, with an intellectual 
component, and as a media practice that involved the gathering, processing, and transfer 
of information. This did not attenuate jawārī’s difficult conditions. Their visible, mediatory 
participation was likely a function of enslavement and an authorized discourse, partaking in 
gendered, class-based hierarchies. 

In addition to the myth of technological neutrality, then, instrumental jawārī challenge 
the object-subject chasm by appearing less as the object of a subject’s perspective—though 
many were meant to benefit a powerful patron—than as a sign of instrumentality, as a 
vehicle for the capacity to carry and to channel. Again, this is not to say that these depictions 
were less dehumanizing. Rather, it is to try and envision the representational terms in 
which objectification operated before the emergence of the modern subject. In Western 
philosophy, the splitting of object and subject bespeaks a relationship of domination that is 
fixed and unilateral; the objectification of an entity indeed implies that the entity is placed 



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   120

before a subject, as the framed object of its viewpoint—that is, as a graspable element of 
perception, inquiry, and use. The “I” or ego that experiences itself as subject of thought 
did not develop before the eighteenth century, when being was identified with the being of 
the individual, thinking subject, who experiences the world as the object’s other.79 This also 
explains the link between the reflective subject and the quest for technological mastery. 
As Paul Ricoeur summarized, “the claim to master the existent as a whole, in technology, is 
only a consequence [. . .] of this emergence of man on the stage of his own representation.”80 

In earlier periods, the subject had not yet become the source of all representation. In 
medieval philosophical parlance, the subject, in fact, was not the “I” but a substrate, as 
indicated by both the Latin subjectum and the Arabic mawḍūʿ, which signify “basis” or, 
more literally, “that which is posited”—a sort of ground, then, through which something 
comes to be, rather than ego or self.81 The premodern subject, at least as conveyed by 
the words subjectum and mawḍūʿ, operated as a support in both a physical and a logical 
sense, respectively as a substrate for change in the world and as a support for predicates 
in a proposition.82 This may indicate yet another possible interpretation of jawārī’s 
representations: not only as instrumental figures but as medieval subjects, one of the 
possible grounds from which mechanisms of transmission were materialized, replicated, 
and conceptualized.

Bibliography

Adas, Michael. Machines as the Measure of Men. Ithaca, NY: Cornell University Press, 1989.

Ahmed, Leila. Women and Gender in Islam. New Haven, CT: Yale University Press, 1992.

Ali, Kecia. Marriage and Slavery in Early Islam. Cambridge, MA: Harvard University Press, 
2010.

Amari, Michele. Le epigrafe arabiche di Sicilia, trascritte, tradotte e illustrate. Palermo: L. 
Pedone-Lauriel, 1971.

Anchassi, Omar. “Status Distinctions and Sartorial Difference: Slavery, Sexual Ethics, and the 
Social Logic of Veiling in Islamic Law.” Islamic Law and Society 28 (2021): 125–55.

79.  Though the emergence of the modern subject in Western philosophy is commonly attributed to Descartes, 
a number of philosophers have corrected this belief, showing that the modern subject was developed by later 
philosophers such as Kant; see, for example, Jean-Luc Marion, “Descartes hors sujet,” Les études philosophiques 
88, no. 1 (2009): 51–62. 

80.  Paul Ricoeur, The Conflict of Interpretations: Essays in Hermeneutics (Evanston, IL: Northwestern 
University Press, 1974), 229. 

81.  Ibid., 228–29. 
82.  Etienne Balibar, Barbara Cassin, and Alain de Libera, “Subject,” trans. David Macey, Radical Philosophy 

138 (2006): 15–41.



121  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

Antrim, Zayde. “Qamarayn: The Erotics of Sameness in the 1001 Nights.” Al-ʿUṣūr al-Wusṭā 
28 (2020): 1–44.

Atanasoski, Neda, and Kalindi Vora. Surrogate Humanity: Race, Robots, and the Politics of 
Technological Futures. Durham, NC: Duke University Press, 2019.

ʿAthamina, Khalil. “How Did Islam Contribute to Change the Legal Status of Women: The Case 
of the Jawārī, or the Female Slaves.” Al-Qanṭara 28 (2007): 383–408.

Balafrej, Lamia. “Automated Slaves, Ambivalent Images, and Noneffective Machines in 
al-Jazarī’s Compendium of the Mechanical Arts, 1206.” 21: Inquiries into Art, History, 
and the Visual 3, no. 4 (2022): 737–74.

– – –. “Domestic Slavery, Skin Color, and Image Dialectic in Thirteenth-Century Arabic 
Manuscripts.” Art History 44, no. 5 (2021): 1012–36.

Balibar, Etienne, Barbara Cassin, and Alain de Libera. “Subject.” Translated by David Macey. 
Radical Philosophy 138 (2006): 15–41.

Barker, Hannah. “Purchasing a Slave in Fourteenth-Century Cairo: Ibn al-Akfānī’s Book of 
Observation and Inspection in the Examination of Slaves.” Mamlūk Studies Review 19 
(2016): 1–24.

– – –. That Most Precious Merchandise: The Mediterranean Trade in Black Sea Slaves, 1260–
1500. Philadelphia: University of Pennsylvania Press, 2019.

Beneito, Pablo. “Qurrat al-ʿAyn, the Maiden of the Kaʿba: On the Themenophany Inspiring Ibn 
ʿArabī’s Tarjumān.” Religions 12 (2021): 23–40.

Benjamin, Ruha. Race after Technology: Abolitionist Tools for the New Jim Code. Medford, 
MA: Polity Press, 2019.

Berryman, Sylvia. The Mechanical Hypothesis in Ancient Greek Natural Philosophy. 
Cambridge: Cambridge University Press, 2013.

Binder, O. O. “You’ll Own ‘Slaves’ by 1965.” Mechanix Illustrated, January 1957, 62–65.

Bray, Francesca. “Gender and Technology.” Annual Review of Anthropology 36 (2007): 37–53.

Brockelmann, Carl. History of the Arabic Written Tradition. Translated by Joep Lameer. Vol. 
1. Leiden: Brill, 2016.

Caiozzo, Anna. “Entre prouesse technique, cosmologie et magie: L’automate dans l’imaginaire 
de l’Orient médieval.” In La fabrique du corps humain: La machine modèle du vivant, 
edited by Véronique Adam and Anna Caiozzo, 43–79. Grenoble: MSH-Alpes, 2010.

Canard, Marius. “Quelques aspects de la vie sociale en Syrie et Jazīra au dixième siècle d’après 
les poètes de la cour Ḥamdanide.” In Arabic and Islamic Studies in Honour of Hamilton 
A. R. Gibb, edited by G. Makdisi, 168–90. Leiden: Brill, 1965.

https://journals.library.columbia.edu/index.php/alusur/article/view/8287
https://journals.ub.uni-heidelberg.de/index.php/xxi/article/view/91685
https://journals.ub.uni-heidelberg.de/index.php/xxi/article/view/91685


Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   122

Carra de Vaux, Bernard. “Les pneumatiques de Philon de Byzance.” Notices et extraits des 
manuscrits de la Bibliothèque nationale 39 (1903): 27–229.

Chism, Christine. “Lost Worlds: Encyclopaedism and Riddles in the Tale of Tawaddud/
Theodor.” In Bestsellers and Masterpieces: The Changing Medieval Canon, edited by 
H. Blurton and D. F. Reynolds, 234–61. Manchester: Manchester University Press, 2022.

Chude-Sokei, Louis. The Sound of Culture: Diaspora and Black Technopoetics. Middletown, 
CT: Wesleyan University Press, 2016.

Cornell, Rkia. Early Sufi Women: Dhikr an-niswa al-mutaʿabbidāt aṣ ṣūfiyyāt. Louisville: Fons 
Vitae, 1999.

Duggan, T. M. P. “Diplomatic Shock and Awe: Moving, Sometimes Speaking, Islamic 
Sculptures.” Al-Masāq 21 (2009): 229–67.

El-Cheikh, Nadia Maria. “Women’s History: A Study of al-Tanukhi.” In Writing the Feminine: 
Women in Arab Sources, edited by Randi Deguilhem and Manuela Marín, 129–48. 
London: I. B. Tauris, 2002.

Fierro, Maribel. “The ‘Bestsellers’ of al-Andalus.” In Artistic and Cultural Dialogues in the Late 
Medieval Mediterranean, edited by María Marcos Cobaleda, 31–56. Cham: Palgrave-
MacMillan, 2021.

Flood, Finbarr Barry. The Great Mosque of Damascus: Studies on the Makings of an Umayyad 
Visual Culture. Leiden: Brill, 2001.

Geissinger, Aisha. “Female Figures, Marginality, and Qur’anic Exegesis in Ibn al-Jawzī’s 
Ṣifat al-Ṣafwa.” In Islamic Interpretive Tradition and Gender Justice: Processes of 
Canonization, Subversion, and Change, edited by Nevin Reda and Yasmin Amin, 151–78. 
Montreal: McGill-Queen’s University Press, 2020.

Ghersetti, Antonella. “The Representation of Slave Girls in a Physiognomic Text of the 
Fourteenth Century.” Mamlūk Studies Review 21 (2018): 21–45.

Goitein, S. D. “Slaves and Slave Girls in the Cairo Geniza Records.” Arabica 9 (1962): 1–20.

Gordon, Matthew, and Kathryn Hain, eds. Concubines and Courtesans: Women and Slavery 
in Islamic History. Oxford: Oxford University Press, 2017.

Hagedorn, Jan Hinrich. Domestic Slavery in Syria and Egypt, 1200–1500. Bonn: Bonn University 
Press, 2020.

Halberstam, Jack. “Automating Gender: Postmodern Feminism in the Age of the Intelligent 
Machine.” Feminist Studies 17, no. 3 (1991): 439–60.

al-Ḥasan, Aḥmad Yūsuf, and Donald Hill. Islamic Technology: An Illustrated History. New 
York: Cambridge University Press, 1986.



123  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

Hill, Donald. “An Andalusian Treatise of the 5th/11th Century.” In Arabic Water-Clocks, 
36–46. Aleppo: Institute for the History of Arabic Science, 1981.

– – –. “Arabic Mechanical Engineering: Survey of the Historical Sources.” Arabic Sciences and 
Philosophy 1 (1991): 167–86.

– – –. “Ḥiyal.” In Encyclopaedia of Islam, 2nd ed., ed. P. Bearman et al. Leiden: Brill Online.

– – –. Kitāb Arshimīdas fī ʿamal al-binkamāt. London: Turner & Devereux, 1976.

– – –. “The Pseudo-Archimedes Treatise.” In Arabic Water-Clocks, 15–35. Aleppo: Institute for 
the History of Arabic Science, 1981.

– – –. “The Toledo Water-Clocks of c. 1075.” History of Technology 16 (1994): 62–71.

Ibn ʿAdī, Abū Zakariyyā Yaḥyā. “Fī ithbāt ṭabīʿat al-mumkin.” Edited by Carl Ehrig-Eggert. 
Zeitschrift für Geschichte der arabischen-islamischen Wissenschaften 5 (1989): 63–97.

Ibn ʿArabī. Tarjumān al-ashwāq. Edited and translated by Reynold Nicholson. London: Royal 
Asiatic Society, 1911.

Ibn Buṭlān. “Risāla jāmiʿa li-funūn nāfiʿa fī shirā al-raqīq wa-taqlīb al-ʿabīd.” In Nawādir 
al-makhṭūṭāt, edited by ʿAbd al-Salām Muḥammad Hārūn, 1:351–89. Cairo: Maṭbaʿat 
Lajnat al-Taʾlīf wa-l-Tarjama wa-l-Nashr, 1951.

Ibn al-Jawzī, ʿAbd al-Raḥmān. Kitāb Muthīr al-ʿazm al-sākin ilā ashraf al-amākin. Edited by 
Marzūq ʿAlī Ibrāhīm. Vol. 2. Giza: Dār al-Rāya, 1990.

Ibn al-Sāʿī. Consorts of the Caliphs: Women and the Court of Baghdad. Translated by Shawkat 
M. Toorawa et al. New York: New York University Press, 2015.

Ikhwān al-Ṣafā.  “On the Practical Crafts.” In On Composition and the Arts, epistles 6–8 of 
Epistles of the Brethren of Purity. Edited and translated by Nader El-Bizri and Godefroid 
de Callataÿ. Oxford: Oxford University Press, 2018.

Ingold, Tim. “Eight Themes in the Anthropology of Technology.” Social Analysis 41, no. 1 
(1997): 106–38.

al-Jawbarī, Jamāl al-Dīn ʿAbd al-Raḥīm. Kitāb al-Mukhtār fī kashf al-asrār. Edited by Manuela 
Dengler. Translated by Humphrey Davies. New York: New York University Press, 2020.

al-Jazarī, Ibn al-Razzāz. The Book of Knowledge of Ingenious Mechanical Devices. Translated 
by Donald Hill. Dordrecht: Reidel, 1974.

– – –. Al-Jāmiʿ bayn al-ʿilm wa-l-ʿamal al-nāfiʿ fī ṣināʿat al-ḥiyal. Edited by Aḥmad Yūsuf 
al-Ḥasan. Aleppo: Institute for the History of Arabic Science, 1979.

Johansen, Baber. “The Valorization of the Human Body in Muslim Sunni Law.” Princeton 
Papers in Near Eastern Studies 4 (1996): 71–112.



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   124

Katz, Marion Holmes. Wives and Work: Islamic Law and Ethics before Modernity. New York: 
Columbia University Press, 2022.

Khan, Geoffrey. “A Petition to the Fāṭimid Caliph al-’Āmir.” Journal of the Royal Asiatic Society 
of Great Britain and Ireland 122, no. 1 (1990): 44–54.

King, David A. Islamic Astronomical Instruments. London: Variorum Reprints, 1987.

Krajewski, Markus. The Server: A Media History from the Present to the Baroque. Translated 
by Ilinca Iurascu. New Haven, CT: Yale University Press, 2018.

Krakowski, Eve. Coming of Age in Medieval Egypt: Female Adolescence, Jewish Law, and 
Ordinary Culture. Princeton, NJ: Princeton University Press, 2017.

Latour, Bruno. Pandora’s Hope: Essays on the Reality of Science Studies. Cambridge, MA: 
Harvard University Press, 1999.

Lugones, María. “Heterosexualism and the Colonial/Modern Gender System.” Hypatia 22, no. 
1 (2007): 186–209.

Maddison, Francis, and Emilie Savage-Smith. Science, Tools and Magic. Part 1: Body and 
Spirit, Mapping the Universe. London: Nour Foundation with Azimuth Editions and 
Oxford University Press, 1997.

Marion, Jean-Luc. “Descartes hors sujet.” Les études philosophiques 88, no. 1 (2009): 51–62.

Marmon, Shaun E. “Domestic Slavery in the Mamluk Empire: A Preliminary Sketch.” In 
Slavery in the Islamic Middle East, edited by Shaun E. Marmon, 1–23. Princeton, NJ: M. 
Wiener, 1999.

– – –. “Intersections of Gender, Sex, and Slavery: Female Sexual Slavery.” In The Cambridge 
World History of Slavery, AD 500–AD 1420, edited by Craig Perry, David Eltis, Stanley 
Engerman, and David Richardson, 185–213. Cambridge: Cambridge University Press, 
2021.

Mestyan, Adam. “Arabic Lexicography and European Aesthetics: The Origins of Fann.” 
Muqarnas 28 (2011): 69–100.

Munn, Luke. Automation Is a Myth. Stanford, CA: Stanford University Press, 2022.

al-Murādī, Ibn Khalaf. Kitāb al-Asrār fī natāʾij al-afkār. Edited and translated by Ahmed Ragad. 
Milan: Leonardo3, 2018.

al-Mutanabbī. Sharḥ dīwān al-Mutanabbī. Edited by ʿAbd al-Raḥmān al-Barqūqī. Vol. 1. Cairo: 
Maṭbaʿat al-Saʿāda, [1938].

Norozi, Nahid. “The ‘Metal Army’ of Alexander in the War against Indian King Porus in Three 
Persian Alexander Books (Tenth-Fourteenth Centuries).” Iranian Studies 52 (2019): 
903–22.



125  •  Lamia BaLafrej

Al-ʿUṣūr al-Wusṭā 31 (2023)

Parisi, Luciana. Abstract Sex: Philosophy, Bio-Technology and the Mutations of Desire. 
London: Continuum, 2004.

Pattison, Joel. “A Golden Tree in the ‘Garden of Pages’: The Genoese Embassy to Morocco of 
1292.” Journal of Medieval Worlds 1, no. 4 (2019): 1–9.

Perry, Craig. “The Daily Life of Slaves and the Global Reach of Slavery in Medieval Egypt, 
969–1250 CE.” PhD dissertation, Emory University, 2014.

– – –. “Historicizing Slavery in the Medieval Islamic World.” International Journal of Middle 
East Studies 49, no. 1 (2017): 133–38.

– – –. “Slavery and Agency in the Middle Ages.” In The Cambridge World History of Slavery, 
AD 500–AD 1420, edited by Craig Perry, David Eltis, Stanley Engerman, and David 
Richardson, 240–67. Cambridge: Cambridge University Press, 2021.

Phan, Thao. “The Materiality of the Digital and the Gendered Voice of Siri.” Transformations 
49 (2017): 23–33.

Philo of Byzantium. Le livre des appareils pneumatiques et machines hydrauliques par Philon 
de Byzance. Edited and translated by Bernard Carra de Vaux. Paris: Klincksieck, 1902.

Prager, Frank D. Philo of Byzantium, Pneumatica: The First Treatise on Experimental Physics, 
Western Version and Eastern Version. Wiesbaden: Reichert, 1974.

al-Qazwīnī, Zakariyyāʾ b. Muḥammad. Athār al-bilād wa-akhbār al-ʿibād. Beirut: Dār Ṣādir, 
1960.

Ragab, Ahmed. “Making History: Identity, Progress and the Modern-Science Archive.” Journal 
of Early Modern History 21 (2017): 433–44.

Ragib, Yusuf. Actes de vente d’esclaves et d’animaux d’Egypte médiévale. Vol. 1. Cairo: Institut 
français d’archéologie orientale, 2006.

Resnikoff, Jason. “The Myth of Black Obsolescence.” International Labor and Working-Class 
History 102 (2022): 124–45.

Rice, D. S. “A Drawing of the Fatimid Period.” Bulletin of the School of Oriental and African 
Studies 21 (1958): 31–39.

Ricoeur, Paul. The Conflict of Interpretations: Essays in Hermeneutics. Evanston, IL: 
Northwestern University Press, 1974.

Roşu, Felicia, ed. Slavery in the Black Sea Region, c. 900–1900: Forms of Unfreedom at the 
Intersection between Christianity and Islam. Leiden: Brill, 2021.

Saliba, George. Islamic Science and the Making of the European Renaissance. Cambridge, MA: 
MIT Press, 2007.

Samsó, Julio. Las ciencias de los antiguos en al-Ándalus. Madrid: MAPFRE, 1992.



Al-ʿUṣūr al-Wusṭā 31 (2023)

Instrumental Jawārī  •   126

Schine, Rachel. “Translating Race in the Islamic Studies Classroom.” Al-ʿUṣūr al-Wusṭā 30 
(2022): 320–83.

Schmidt, Laura Lee. “Islamic Automata in the Absence of Wonder.” SM thesis, Massachusetts 
Institute of Technology, 2010.

Shaikh, Saʿdiyya. Ṣūfī Narratives of Intimacy: Ibn ʿArabī, Gender, and Sexuality. Chapel Hill: 
University of North Carolina Press, 2012.

Somogyi, Joseph de. “Ibn al-Jauzī’s Handbook on the Makkan Pilgrimage.” Journal of the 
Royal Asiatic Society 70, no. 4 (1938): 541–46.

Staudenmaier, John M. Technology’s Storytellers: Reweaving the Human Fabric. Cambridge, 
MA: MIT Press, 1985.

Urban, Elizabeth. Conquered Populations in Early Islam. Edinburgh: Edinburgh University 
Press, 2020.

Vernet, Juan. “Texto árabe de la corte de Alfonso X el Sabio X.” Al-Andalus: Revista de las 
Escuelas de estudios árabes de Madrid y Granada 43, no. 2 (1978): 405–33.

Zielinski, Siegfried, and Peter Weibel, eds. Allah’s Automata: Artifacts of the Arab-Islamic 
Renaissance (800–1200). Ostfildern: Hatje Cantz, 2015.

https://journals.library.columbia.edu/index.php/alusur/article/view/uw30schine
https://dspace.mit.edu/handle/1721.1/59207