Hostname: page-component-77f85d65b8-g4pgd Total loading time: 0 Render date: 2026-03-28T22:17:46.159Z Has data issue: false hasContentIssue false
  • English
  • Français

Voluntary enslavement in an Abbasid-era papyrus letter

Published online by Cambridge University Press:  20 March 2023

Jelle Bruning Open the ORCID record for Jelle Bruning [Opens in a new window]
Jelle Bruning*
Affiliation:
Leiden University Institute for Area Studies, Leiden University, Leiden, The Netherlands Email: j.bruning@hum.leidenuniv.nl
Rights & Permissions [Opens in a new window]

Abstract

Central to this article is an Arabic letter written on papyrus in an Egyptian prison in the late ninth or early tenth century ce. The author complains that he and his companions are being kept in terrible conditions and that they have received insufficient support from outside prison. Interestingly, he indicates that there is a strong inclination among the group to offer themselves as slaves in order to find relief from their crushing living conditions. By doing so, they would have transgressed Islamic law of that time, which forbade the enslavement of free inhabitants of the Realm of Islam. The letter is a unique source for the social history of slavery, especially self-enslavement, in Abbasid society. This article presents, translates, and annotates this letter and offers a detailed study of its contents.

Keywords

Abbasid EgyptimprisonmentIslamic lawslaveryself-enslavementpatronage

Information

Type
Article
Information
Journal of the Royal Asiatic Society , Volume 33 , Issue 3 , July 2023 , pp. 643 - 659
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NCCreative Common License - SA
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike licence (https://creativecommons.org/licenses/by-nc-sa/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the same Creative Commons licence is included and the original work is properly cited. The written permission of Cambridge University Press must be obtained for commercial re-use.
Copyright
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of The Royal Asiatic Society

Who is in a worse spot, in a more hopeless situation, in a more impotent condition than he who was free, in full control of himself, and made himself a slave (fa-ṣayyara nafsahu ʿabdan), the property of someone else (mamlūkan li-ghayrihi), choosing slavery without having been captured or being subjected by force?

Al-JāḥiẓFootnote 1

In his epistle on ‘Concealing Secrets and Holding the Tongue’, the source of the quote above, the well-known Abbasid-era author al-Jāḥiẓ (d. 255 ah/868 ce) uses self-enslavement as a metaphor for letting somebody in on a secret: according to him, a person wields excessive power over someone else when he knows one of his secrets. The metaphor derives its force from a combination of third/ninth-century and later Islamic law and a general notion of a slave's inferior position vis-à-vis a freeman in society. Unlike the first two Islamic centuries, when legal authorities had permitted the (often temporary) enslavement of free persons,Footnote 2 by al-Jāḥiẓ's time Muslim jurists held that free inhabitants of the Realm of Islam, whether Muslim or not, in general could not be reduced to a state of slavery. According to the few sources that say something about the legitimisation of slavery in Islamic law, slave status, and the social inferiority that came with it, was a punishment for those who had refused to become subject to Islam.Footnote 3 Despite such statements as ‘freedom is the original human condition’ (al-aṣl huwa al-ḥurriyya), occasionally found articulated in fourth/tenth-century and later legal literature but never found thoroughly analysed,Footnote 4 Muslim jurists saw no conflict between such notions of man's free origins and the institution of slavery, and embracing Islam after enslavement did not result in one's return to freedom. In fact, their discussions of legal cases in which the boundaries between slavery and freedom were blurred, concentrated on developing slavery into an ever more sophisticated institution.Footnote 5 Unsurprisingly, freedom was considered a very precious legal status that one would fiercely protect.Footnote 6 Hence al-Jāḥiẓ's statement that voluntary enslavement, that is, the wilful but illegal surrendering of one's freedom and acceptance of social inferiority, must be a sign of profound misery.

How al-Jāḥiẓ's statement relates to actual self-enslavement among Abbasid-era Muslims remains very poorly understood. Sources for the history of slavery in that period only incidentally refer to the enslavement of free persons.Footnote 7 Because the topic mostly turns up in the largely theoretical discussions of juridical texts, voluntary enslavement has hardly been studied beyond its legal parameters, while its social, economic, and other contexts are often neglected.Footnote 8 Complicating the study of self-enslavement, the texts often used for studying the history of early Islamic slavery contain little information about the possible prospects of self-enslaved persons. By and large, most historical texts and works of adab (often translated as ‘belles-lettres’) present us with images of individuals owned by a relatively small class of urban elites, especially those serving them as entertainers, concubines, and soldiers. Even when these sources can be used to study particular forms or aspects of slavery associated with elite circles of Muslim society,Footnote 9 they are almost completely silent about the reality of slavery as experienced by the vast majority of slaves in the Abbasid empire: the thousands of men, women, and children forced into serving affluent households as domestic slaves.Footnote 10 Using these sources alone makes it nearly impossible to know what may have motivated free people to give up their freedom.

In light of this lack of information about actual self-enslavement and its contexts, an until now unstudied Arabic papyrus letter deserves our full attention. Uniquely, its unnamed third/ninth- or fourth/tenth-century author writes of his strong inclination towards self-enslavement. Writing from an Egyptian prison, the author complains about his and his nine companions’ bad living conditions, asks the addressee for financial support, and warns that he and his companions will offer themselves as slaves if the addressee refuses to support them. This letter, then, offers us an unparalleled opportunity to study conditions that could lead to the (illegal) enslavement of free individuals in the Abbasid empire. Not written with the aim of meeting certain literary expectations, the letter preserves the concerns and considerations of one of those Abbasid-era Muslims whose voices are hardly ever heard in historical, legal, or other sources. Although the choice of the letter's author and his companions to give up their freedom must have been a deeply personal one, on a more general level the letter speaks to the vulnerability of those who are dependent on others and the existential insecurity that could arise when these relationships are disturbed. In addition, it raises questions about the rigidity of legal institutional boundaries and about the use of legal status for upward social and economic mobility. This article offers an analysis of this letter. But before we turn to this, we will first look at the actual document and its contents.

P.CtYBR inv. 1873 qua: text, translation, and commentary

The letter in question, currently kept in Yale University's Beinecke Rare Book and Manuscript Library under the inventory number P.CtYBR inv. 1873 qua, has been written on an almost entirely preserved light-brown papyrus sheet measuring 34.6 by 11.5 cm (see Figure 1). Its cutting lines are preserved on all sides except at the top, which has been broken off without damaging the beginning of the text. Of the four lines of a postscript written in the papyrus's upper margin, only the first letters are preserved. With the exception of this broken-off top, the upper third of the papyrus (up to line 7) exhibits hardly any damage. There are many holes in the middle third, causing damage to lines 7–11 and line 13. This damage is concentrated in the right half of these lines. There is hardly any damage in the lower third of the papyrus. A tear runs through lines 14–20, which does not affect the text's legibility. The papyrus bears traces of horizontal folding. At least five horizontal folds are still visible in the lower half of the papyrus. The distance between two of these folds ranges between 1.6 and 1.9 cm. The regular distribution of holes in lines 9 and 10 reveals that two more horizontal folds ran through these lines. Many irregularly distributed folds are visible in the upper half of the papyrus. These are likely to be secondary, that is, the result of folding the papyrus when it was discarded.

Figure 1. P.CtYBR inv. 1873 qua. Source: Yale University, Beinecke Rare Book and Manuscript Library.

The author wrote the standard opening and closing sections of his letter (lines 1–3 and 19–20) in a quite regular and neat third/ninth- or fourth/tenth-century hand, thus graphically distinguishing these parts of the letter from its core.Footnote 11 The author used this style for the first words of line 4 as well. From approximately line 7, the script becomes noticeably more cursive, with more ligatures and a less neat appearance. Compare, for example, the execution of the words in lines 6 and 15. This part of the letter frequently exhibits abusive ligatures, such as in in line 11, and in line 12, and in 16. The author wrote the poorly preserved postscript in the upper margin at a right angle to the rest of the letter. The text contains no date, but its script and layout allow it to be dated to the second half of the third/ninth or the first half of the fourth/tenth century.Footnote 12

The back of the letter is empty. Although the author may have written the letter's address in the now broken-off top of the papyrus, most addresses of third/ninth- and fourth/tenth-century letters stood approximately at the same height as the basmala on the other side of the papyrus.Footnote 13 It is most likely that our text never had an address and was despatched together with other letters or goods sent to the addressee.

Text

Text in top margin

Diacritical dots

3. , ; 4. ; 5. , (twice); 6. , , , ; 7. ; 8. , 9. (twice), , , ; 10. ; 11. , , (cf. comm.), , ; 12. ; 13. , ; 14. , ; 15. ; 16. ; 17. .

Translation

In the name of God, Merciful and Compassionate.

May God prolong your life; perpetuate your glory, honour, support and happiness; and may He extend His grace and benevolence to you.

You know—may God render you glorious—what we have had to endure because of (our) imprisonment and the distress we are in. We are still despoiled of our property and all our wealth has been taken. We have no one who attends to or remembers us. We needed you to remind Abū al-Ḍaḥḥāk—may God prolong his life—of our affair and to inform him of our condition. We are obedient servants, ten in number. Some of us are slaves, the others are freemen. We will make ourselves slaves so that heFootnote 14 becomes desirous of owning themFootnote 15 and you (two) hasten to deal with our affair so that we may be freed from the distress we are in. Nothing remains of the alms that we were given. We are still destitute,Footnote 16 with no one to feed or clothe us.

Be so kind—may God render you glorious—to send us money, for we have no one who attends to us. When you have heard news about …,Footnote 17 my patroness, do inform us about it so that our minds be at ease or else so that we can act in an appropriate manner, God willing.

May God prolong your life; render you glorious; make you honourable; and may He extend His grace and benevolence to you.

Translation of text in top margin

… and … and Abū al-ʿAbbās … greetings …

Commentary

  1. 4. The words are a conventional opening of an exposition. For other attestations, see Werner Diem, ‘Philologisches zu arabischen Dokumenten der Österreichischen Nationalbibliothek in Wien’, Wiener Zeitschrift für die Kunde des Morgenlandes 101 (2011), pp. 75–140 at p. 128, no. 1 (third/ninth century); P.Ryl.Arab. I § VI 10 (not dated);Footnote 18 and P.Vind.Arab. III 23 (fourth/tenth century or later). See also P.Hamb.Arab. II 43 (third/ninth century) and P.Vind.Arab. II 32 (fourth/tenth century), in which appears in restored lacunas. The early fourth/tenth-century exercise of a letter opening P.Vind.Arab. III 17.a starts with these words as well. Note that the exposition of the third/ninth-century petition P.Khalili I 16 starts with a semantically related expression: innahu lam yukhfā ʿalā al-amīr … mā ḥaṣala binā fī hādhihi al-sana, ‘the amīr … is aware of what has befallen us this year’ (lines 4–5).

  2. 5. With the exception of a slightly faded first letter, the word is quite legible (the individual letters are undotted). Considering the context of this word, and especially the author's remark that all his and his companions’ wealth has been taken from them, the word probably means ‘being despoiled of property’; cf. ḥaraba, ‘to despoil (someone) of (his) property’; aḥraba, ‘to find (someone) to be despoiled of (his) property’; and iḥtaraba, ‘to be despoiled of one's property’. The dictionaries of Kazimirski, Dozy, and Lane do not record the existence of the verb taḥarraba, but Middle Arabic texts attest to the use of reflexive verb stems for expressing the passive.Footnote 19

  3. 7. The dhāl in ends unusually low, making the letter resemble a rāʾ/zāy.

    In the current corpus of published Arabic papyri, men named Abū al-Ḍaḥḥāk also appear in the fourth/tenth-century P.Vind.Arab. I 57 and in P.Vind.Arab. III 28, dated to the fifth/eleventh century. If our author and his companions were imprisoned in a state prison during the governorate of Aḥmad b. Ṭūlūn, or that of his son and successor Khumārawayh, our author may be referring to Abū al-Ḍaḥḥāk Maḥbūb b. Rajāʾ al-Ḥaḍārī, personal scribe to these governors. This Abū al-Ḍaḥḥāk was known for his eloquence and noble character, and had direct access to the court.Footnote 20 He would have been able to bring these prisoners to the governor's attention.

  4. 8. The reconstruction of is based on the observation that the lacuna is too large for a single letter and that the final alif is connected to a preceding letter. Traces of the top and bottom of the wāw in are still visible. Very faint traces of the tāʾ are visible.

  5. 11. The beginning of the word can be read with considerable certainty. The word starts with a clearly dotted yāʾ, which is followed by another denticle and an ʿayn or ghayn. One of the latter two letters is preferred over reading a fāʾ or qāf because the author usually closes the loop of the fāʾ/qāf (cf. in line 2, in line 9, and in line 10) and leaves the loops of many ʿayns/ghayns open (for example, in line 3, in lines 6 and 15, and in line 16). What follows in that word is less clear. A hole removes the end of the verb from our sight and a small letter, such as a denticle, may have been lost in the lacuna. If intended, the dotting of the letters at the end of the word is confusing. Among the words that fit what is legible, , ‘he becomes desirous of owning them’Footnote 21 fits the context best. If the author intended a fāʾ instead of a ghayn, , ‘he throws them out’, would also fit.

    The tāʾ of is dotted. A large dot stands on top of the first wāw at the same height as the denticle that follows. Although the author did not write the dot on top of the denticle (as he usually did elsewhere in his letter), it is most likely that he intended the second person plural instead of the third person plural (, ‘and they hasten to deal with …’) because the author only refers to Abū al-Ḍaḥḥāk and the addressee as those who are able to influence his and his companions’ stay in prison. Note that the author used the plural instead of the dual classical Arabic grammar demands in this context.Footnote 22

    The palaeography of does not resemble these letters in in lines 9 and 10 and looks more like . There seems no doubt, however, that the author meant مما .

  6. 13. The author initially wrote the kāf of in an unclear way, making it possible to read the letter as a fāʾ or qāf. He subsequently added a little oblique stroke to clarify his writing. A less likely alternative reading is . In this reading, the author meant to write , ‘he gave us to drink’; in this scenario, he crossed out the misplaced qāf with the oblique stroke but did not correct his mistake.

  7. 14. The reading of the words is not entirely certain. As noted above, the author leaves the loop of most of his ʿayns/ghayns open and closes those of the fāʾ and qāf. In , the ʿayn is closed. Note that the author (unintentionally?) wrote a dot below the denticle in . The mīm in is only visible in the sharp point that connects the bottom of the alif with the preceding letter. Compare the distinctively different palaeography of an alif following a denticle in and in lines 6 and 15, for example. As in , the mīm is hardly visible before an upward stroke in (line 8) and (line 17) too. After an explicit reference to the lack of money in lines 12 and 13, a request for financial support does not seem misplaced.

  8. 16. What precedes the clearly legible is the name of the author's patroness. The dotted tāʾ in , ‘my patroness’, leaves no doubt that we are dealing with a woman, but medieval onomastic literature does not mention a female name that fits what the author wrote: a sīn/shīn, a denticle, and a final bāʾ/tāʾ/thāʾ. Perhaps the author intended to write the female name ‘Sutayt’ but wrote one instead of two denticles.Footnote 23

  9. 17. Rather than more common seems to have been the expression . See P.Heid.Arab. II 62 (fourth/tenth century), recto line 4 and P.Vind.Arab. II 12 (sixth/twelfth century), line 5.

  10. 19. The letter ends with a very common series of blessings, starting with aṭāla allāh baqāʾaka, ‘may God prolong your life’. In that blessing, a clearly legible and unconnected letter precedes . This letter cannot be part of allāh, the word that usually precedes it. The letter looks like a free-standing wāw (compare the wāws in on the same line). Two letters that closely resemble a denticle or short lām and a final hāʾ precede it. A tear removes from our sight one or two letters that originally preceded these letters. Because there is space for only one word between the preserved alif of allāh and the letter preceding , the most likely reading of the beginning of this line is , in which the writer of the letter wrote a wāw by mistake (perhaps because he wanted to write wa-aʿazzaka before he realised that the first blessing was not yet complete).

Crushing living conditions in prison

The author modelled his letter after contemporary petitions, adopting their structure and using their formulas. He took care to visually separate the letter's various parts, which, indeed, neatly correspond with the parts of third/ninth-century petitions.Footnote 24 He graphically marked the conventional opening of his letter: the invocation of God and opening blessings on the addressee (lines 1–3). Beside starting the invocation above the line, he separated this part of the letter from the following part by adding an empty line after line 3.Footnote 25 Lines 4–13 contain the letter's exposition, in which the author describes his and his companions’ troubling situation. He starts this part of his letter with a verb built on the root ʿ-l-m, which is frequently used at the beginning of an exposition in contemporary documents,Footnote 26 including at the beginning of the exposition of petitions.Footnote 27 Rather than the more often found uʿlimuka, ‘I inform you’, the author of our letter has chosen to start his exposition with the equally conventional words anta taʿlimu, ‘you know’,Footnote 28 highlighting the addressee's acquaintance with the author's situation and perhaps adding to the exposition a moral appeal. Because in most lines the author used the entire width of the papyrus for his letter, the free space at the end of line 13, in addition to an empty line that follows, signals that the end of the exposition has been reached. Following the structure of the petition, lines 14–18 contain two request clauses, the first starting with the common petition formula in raʾayta an, ‘be so kind to …’.Footnote 29 The elongation of the shīn in shāʾa in line 18 and the empty space at the end of that line graphically separate this part of the letter from the closing blessings in lines 19–20.Footnote 30 The afterthought written in the top margin at a right angle to the text below it includes the request to convey greetings to others and/or greetings the author had been asked to convey on behalf of the acquaintances they have in common. The presence of this afterthought reveals that the document must be considered a letter and not an actual petition.

The format of the petition fits well with the author's primary reason for contacting the addressee: requesting immediate financial help. At the beginning of the letter's exposition, the sender reminds the addressee that he and his nine companions have been imprisoned and are being kept in very poor conditions (lines 4–5). The letter gives no information about the reason for their imprisonment or the type of prison in which they are being kept.Footnote 31 Considering the fact that the author shares in his companions’ unhappy fate and that he is aware of their situation, the letter gives the impression that the ten of them have been imprisoned together in a collective cell. At the beginning of the exposition, the author emphasises the ‘distress’ (ḍīq) he and his companions are experiencing in prison (line 5, also line 12). Descriptions of the general living conditions in Abbasid-era prisons may explain why their imprisonment caused distress.Footnote 32 Historical sources indicate that a serious lack of hygiene, a lot of vermin, and intense promiscuity were common characteristics of collective cells.Footnote 33 Graphically illustrating the unhygienic and possibly dangerous living conditions in a late third/ninth-century Egyptian prison, the History of the Patriarchs reports that a prisoner who wished to kill a severely injured inmate took ‘the prison's filth and the grim which formed on its ceilings from the foul air’ and rubbed it into the man's wounds.Footnote 34

However, what seems to have particularly aggravated the situation of the letter's author and his companions is that their possessions, including their money (lines 5–6), have been confiscated. The letter does not state why their possessions were taken from them. Without information about why the group was imprisoned, various reasons can be speculated upon. Contemporary legal texts, mostly discussing imprisonment in the context of disputes about debts, for example, write that a legal authority could confiscate someone's property in order to investigate his solvency and, if necessary, to sell (part of) it.Footnote 35 Historical sources report the seizure of property of those accused of having embezzled state funds.Footnote 36 They also relate that powerful members of society, including political authorities, at times imprisoned someone who had provoked their anger and confiscated his possessions, most likely to increase his discomfort.Footnote 37 Aḥmad b. Ṭūlūn (in office 254–270/868–884), for example, reportedly justified the long imprisonment of one of his scribes by declaring that the latter was lucky not to share in the fate of ‘those whose property I confiscated and whom I subjected to evil (al-makrūh) so that no one would hear about them and so that they would disappear without a trace’.Footnote 38 Even when authorities were intent on returning someone's seized possessions after they been released from prison,Footnote 39 confiscating a prisoner's possessions directly increased his dependence on the goodwill of others to maintain him. As is well known, prisoners had to supply their basic needs themselves,Footnote 40 despite the fact that jurists advised political and legal authorities to provide their prisoners with food and adequate clothing. Legal texts stipulate that prisoners whose belongings have been confiscated should remain able to provide for their basic needs and those of their dependants for a limited period of time.Footnote 41 Lacking the means to maintain oneself in prison was a serious cause of distress. The third/ninth-century letter P.Heid.Arab. II 8 illustrates this well when its impoverished author writes that ‘if I will be incarcerated, I will starve to death’. The author of our letter seems to have had similar fears. At the end of the exposition, he writes that the alms (ṣadaqa) he and his companions received had been insufficient and that no one had brought them food or clothing (lines 12–13).

After this complaint about neglect and poor living conditions, the author further indicates that no one ‘remembers’ them (line 6). Here, he expresses his fear about being forgotten in prison. Occupying six lines, this is a major theme in the letter—and understandably so. Lack of support from outside prison could drastically impact on an inmate's living conditions and could result in lifelong imprisonment or death. The first/seventh-century senders of the Coptic P.Mon.Epiph. 177, for example, accuse their addressees of having ‘forgotten us in the captivity wherein we are’ and urge them to send money and food or else ‘there will no life be left in us’.Footnote 42 Using similar words, an Egyptian governor of the 150s/770s allegedly said that individuals whose detention was no longer on the authorities’ minds ‘joined the dead’ while they were still alive.Footnote 43 Despite the fact that qāḍīs and jailers are likely to have kept registers listing their prisoners and stating the reasons for and the date of their incarceration,Footnote 44 historical sources give the impression that the risk of being forgotten was high for those imprisoned for criminal offences or political reasons.Footnote 45 In Egypt, Aḥmad b. Ṭūlūn was known to have frequently lost interest in his prisoners, that is, those he did not incarcerate in oubliettes especially designed for this purpose.Footnote 46 According to the fourth/tenth-century Ibn al-Dāya, he took care of (yurāʿī) a prisoner for the duration of only one year, ‘and after the year had passed he would not remember him (lam yadhkurhu)’.Footnote 47

Relief through self-enslavement

At this point in the letter, the author provides information about himself and his companions. He gives his and his companions’ legal status and number. He had hoped that the addressee would have brought their situation to the attention of one Abū al-Ḍaḥḥāk (lines 7–8), who seems to have been someone who could help them get out of prison,Footnote 48 but the addressee apparently did not do so. Because the author makes it clear that it is now too late for intercession and that his group is in direct need of sustenance, it is unlikely that this information was meant to help prison authorities identify the group. As we will see, the author gives this information in order to underscore their immediate need of help.

The sender couches his and his companions’ identity in terms that refer to dependency relationships. Anticipating the warning that the group's current situation will force them to enslave themselves so as to escape their misery, with which the author concludes the exposition (lines 10–12), he states in lines 8–10 that his group consists of ‘obedient servants’ (ghilmān mamālīk), among whom are both freemen (aḥrār) and slaves (mamālīk). The passage calls for reflection on the terminology the author uses. First, he seems to use ghilmān (sg. ghulām) not in the meaning of boys or adolescents, but in the sense of people who hold some kind of subservient position in society.Footnote 49 Immediately following ghilmān, the word mamālīk, which literally means ‘owned persons’ or ‘slaves’ but is here used metaphorically, emphasises their obedience and subordination.Footnote 50 When governed by forms of patronage, the ties that bound these ghilmān to their superiors often created (mutual) obligations, demanding loyalty, services, and protection. These social obligations formed important bases of a ghulām's inclusion in his superior's extended household.Footnote 51 Importantly, our author himself occupied exactly such a position. In lines 15–16, he asks the addressee to forward information about his patroness (mawlāt). The author indicates that he wishes to meet the social obligations that come with his client status. He states that he is willing to ‘act in an appropriate manner’ if the addressee has bad news about his patroness (lines 17–18). The author uses an idiom of patronage, too, when he states his reason for turning to the addressee for help. Referring to solicitude (ʿināya) offered in return for loyalty,Footnote 52 he writes in lines 6 and 15 that ‘we have no one who attends (yuʿnā) to us’. These words may well mean that the author believed that others neglected, or were unable to meet, their social obligations towards him and his companions.

Second, the author's use of the word mamālīk in the sense of ‘slaves’ is noteworthy. In contrast to the word ghilmān, which was used for free as well as enslaved individuals, and the metaphorical use of mamālīk to express obedience in line 9, the author uses the word mamālīk (sg. mamlūk) twice, at the end of line 9 and in line 10, in its literal and legal meaning, unambiguously designating those to whom it refers as someone else's property. Because of its explicit reference to slave status and inferiority, this use of the word was unpopular. Possibly following an allegedly prophetic prohibition on using this and other explicit references to a person's slave status,Footnote 53 the word mamlūk appears very infrequently in early Islamic letters. In published papyrus and early paper documents, its use is mostly linked to legal contexts.Footnote 54 It is therefore significant that our author uses the word twice in its literal meaning of ‘slaves’ in addition to its antonym aḥrār, ‘freemen’—a legal term too. Doubtlessly, in anticipation of his warning at the end of the exposition, he needs the addressee to know the legal status of the members of his group and, in particular, that some of them are freemen.

The author's description of the prospect of him and his companions having to offer themselves as slaves is an unambiguous expression of their despair; compare al-Jāḥiẓ's words cited at the beginning of this article. As noted above, third/ninth-century and later Islamic law forbade the enslavement, including self-enslavement, of free men and women within the Islamic polity.Footnote 55 Contemporary jurists made only a few exceptions. For example, they allowed the full enslavement of men and women whose legal status lay somewhere between being free and enslaved (such as a slave with an emancipation contract or a slave promised freedom after his master's death), the re-enslavement of a freedman whose former master was unable to pay off his debts, and, according to some jurists, the enslavement of a person whose legal status was unknown.Footnote 56 Nevertheless, free people who found themselves in very straitened situations are occasionally recorded as having used slavery as a means to save themselves and/or their dependants from worse. Mostly, sources report the sale of spouses and children during famines or in times of extreme financial stress,Footnote 57 but self-sales can be found as well.Footnote 58 For example, in his description of a horrific famine and outbreak of plague in Egypt in 597–598/1200–1202, ʿAbd al-Laṭīf al-Baghdādī writes,

Often, women and good-looking young boys would throw themselves at people's feet, wanting to be bought or sold by them. This was deemed lawful by a great many people, and those sold into captivity were dispersed to Iraq, deepest Khurasan, and elsewhere.Footnote 59

Without doubt, those who sold themselves under such circumstances believed that self-sale would lead to the fulfilment of their immediate need for sustenance. The cash they may have received in return for their freedom enabled them to buy food, although in theory their new masters could confiscate that property.Footnote 60 Equally if not more important, entering into slavery enabled them also to secure maintenance beyond their immediate needs and offered them a longer term perspective—an attractive prospect relative to their circumstances. According to legal theory, slave owners were obliged to provide their slaves with food, clothing, and shelter, and were not to overexploit them.Footnote 61 The really destitute may have considered slavery an institution from which they could benefit. In fact, the responsibility of slave owners to maintain their slaves did not remove from the latter their entitlement to charity when their owners failed to meet their legal obligations towards them or for other reasons. Besides the poor and needy, the Qurʾān mentions slaves as rightful beneficiaries of charity.Footnote 62 Although most later exegetes interpreted the relevant verses as only referring to slaves with an emancipation (kitāba) contract,Footnote 63 documents contemporary with our letter show that Muslims, at times at least, sympathised with (impoverished?) ‘servants’—some of whom have slave names—with whom they were acquainted and gave them charity.Footnote 64

Because of the security of maintenance and support slave status offered in theory, not all slaves wished to be set free, perceiving emancipation as an existential threat that could reduce them to beggary.Footnote 65 For example, the fourth/tenth-century historian al-Kindī reports a slave with an emancipation contract who had not found the means to pay a single instalment and wished to remain enslaved after hearing the Egyptian qāḍī Tawba b. Namir (in office 115–120/733–738) decree that he should fulfil his contract within one year.Footnote 66 Expressing a similar concern for support, a slave owned by the sender of P.Hamb.Arab. II 41 (fifth/eleventh-century) ‘asked to receive an emancipation contract and requested information about his relationship with me and my care for him (riʿāyatī lahu) and my solicitude (ʿināya)’. The sender reports that he had advised him ‘to trust the benevolence towards him and the solicitude towards his affairs that makes the decision for this emancipation contract fruitful’.

Likewise, free people in seemingly hopeless or pressing situations, such as the author of our letter, may have thought of slavery as an institution that offered them and/or their dependants relief and prospects for which it was worth transgressing contemporary Islamic law.Footnote 67 Our author's emphasis on his lack of food and clothing (line 13) suggests, indeed, that it is the theoretical security of sustenance offered by slave status that may have motivated him and his companions to consider giving up their freedom and to use an illegal change of their legal status for their social and economic advancement. Uniquely attesting to this use of slavery, our letter illustrates the possible tension between the reality of slavery (including enslavement) in Abbasid society and contemporary slavery laws in spheres of human interaction that lay beyond the purview of legal magistrates—tensions that are at present imperfectly understood.Footnote 68 But above all, the letter underscores its author's and his companions’ despair. They must have been aware of the social inequality between master and slave, the risk of being sold and transported to unknown regions, the harsh treatment slaves could receive, their frequent loss of choice and agency, and—not least important—the difficulty of regaining their freedom.Footnote 69

Concluding remarks

The letter presented and studied in this article speaks to various forms of ‘unfreedom’: physical detainment and an inmate's dependence on support from outside prison, the Islamic patronate and its social and other obligations, and self-enslavement under duress. But at the heart of its author's troubling situation lies a disturbed patron-client relationship. Under normal circumstances, such a relationship, which formed the legal basis of a client's integration into Muslim society, involved mutual loyalty and support. Importantly, however, it did not constitute an egalitarian relationship between patron and client. Clients often remained dependent on their patron for financial or other support and, as a result, are frequently found maintaining a close relationship with their patron and his family, such as working for them or living in their vicinity.Footnote 70 Their primary dependence on their patron made clients vulnerable people who had to fend for themselves when their patron-client relationship was disturbed. A telling example comes from a third/ninth-century petition addressed to an unknown amīr. Its author, a freedman, writes that his relationship with his former master had deteriorated after his emancipation. Because he was cut off from his primary source of support, the freedman turns to the amīr for financial help because he is no longer able to maintain his dependants.Footnote 71

The relationship between the sender of our letter and his patroness, too, seems to have been disturbed (although the letter does not state why). As noted above, he makes it abundantly clear that he and his nine companions did not receive the support they expected after they were imprisoned. Having lost touch with his patroness (see lines 15–18), the author turns to the addressee for support, first asking him to intercede with one Abū al-Ḍaḥḥāk (line 14) and then, by way of the letter presented above, asking him for financial help. The author's and his companions’ self-enslavement must be understood most probably in this context as well. Because in third/ninth-century and later Islamic law a bond of patronage could not be dissolved or transferred to others before the patron's death,Footnote 72 the author of our letter may have seen self-enslavement as the only means available to him to break loose from his current patroness and to establish a more advantageous relationship with another ‘patron’, in this case a slaveholder able to provide for him in his current circumstances. Surrendering their freedom in desperation, the author and his group apparently hope that this action will induce Abū al-Ḍaḥḥāk to take them under his patronage (line 11) and force him and the addressee to ‘hasten to deal with our affair so that we may be freed from the distress we are in’ (lines 11–12).

Acknowledgements

I would like to thank Matthew Gordon and Robert Hoyland for commenting upon a draft of this article; and I am grateful to Khaled Younes for critically reviewing the edition. Needless to say, any remaining mistakes are my own.

Conflicts of interest

None.

References

1 Al-Jāḥiẓ, , Majmūʿ rasāʾil al-Jāḥiẓ, (eds) Kraus, Paul and al-Ḥājirī, Muḥammad Ṭ. (Cairo, 1943), p. 44Google Scholar; translated in Rosenthal, Franz, The Muslim Concept of Freedom Prior to the Nineteenth Century (Leiden, 1960), p. 92CrossRefGoogle Scholar. I have added the transliterated Arabic.

2 Schneider, Irene, Kinderverkauf und Schuldknechtschaft: Untersuchungen zur frühen Phase des islamischen Rechts (Stuttgart, 1999)Google Scholar; Schneider, Irene, ‘Freedom and slavery in early Islamic time (1st/7th and 2nd/8th centuries)’, Al-Qanṭara 28.2 (2007), pp. 353–382CrossRefGoogle Scholar; Motzki, Harald, ‘Der Prophet und die Schuldner: Eine Ḥadīth-Untersuchung auf dem Prüfstand’, Der Islam 77 (2000), pp. 1–83CrossRefGoogle Scholar.

3 Schneider, Kinderverkauf und Schuldknechtschaft, pp. 27 and 30; Oßwald, Rainer, Das islamische Sklavenrecht (Würzburg, 2017), pp. 24–26Google Scholar. For the inferiority of slaves, see Rosenthal, Concept of Freedom, pp. 81–98.

4 Schneider, Kinderverkauf und Schuldknechtschaft, pp. 23–28; Oßwald, Islamische Sklavenrecht, pp. 22–24; Franz, Kurt, ‘Slavery in Islam: legal norms and social practice’, in Slavery and the Slave Trade in the Eastern Mediterranean (c. 1000–1500 CE), (eds) Reuven Amitai and Christoph Cluse (Turnhout, 2017), pp. 51–141Google Scholar at pp. 81–82 with note 99.

5 Franz, ‘Slavery in Islam’, pp. 82–83; Schneider, ‘Freedom and slavery’, pp. 356–357.

6 Rosenthal, Concept of Freedom, p. 29.

7 See Bruning, Jelle, ‘Slave trade dynamics in Abbasid Egypt: the papyrological evidence’, Journal of the Economic and Social History of the Orient 63.5–6 (2020), pp. 682–742CrossRefGoogle Scholar at pp. 689–690 for references to illegal slave-making during civil wars and for the activity of slavers within the Realm of Islam. Yūsuf Rāġib, Actes de vente d'esclaves et d'animaux d’Égypte médiéval (Cairo, 2002–2006), vol. ii, pp. 83–84, para. 220 discusses a passage in al-Ḥarīrī's Maqāmāt about slave dealers misleading their customers by offering free persons for sale.

8 Most relevant publications focus on the (il)legality of enslaving free people. Schneider, Kinderverkauf und Schuldknechtschaft; Schneider, ‘Freedom and slavery’; and Motzki, ‘Der Prophet und die Schuldner’ study the enslavement of free men and women in pre-classical Islamic legal traditions. Oßwald, Islamische Sklavenrecht, pp. 31–32 and 41–42 studies this topic from the viewpoint of classical Mālikī law. Rāġib, Actes de vente, vol. ii, pp. 27–28 and Hannah Barker, That Most Precious Merchandise: The Mediterranean Trade in Black Sea Slaves, 1260–1500 (Philadelphia, 2019), p. 118 discuss clauses in deeds recording slave sales which state that the sold person is a legal slave. Historians of slavery under Ottoman rule have included other perspectives, including social context, in their discussions of voluntary enslavement or the forced enslavement of free people. See, for example, Clarence-Smith, William G., Islam and the Abolition of Slavery (London, 2006), pp. 74–78Google Scholar and Toledano, Ehud R., The Ottoman Slave Trade and Its Suppression: 1840–1890 (Princeton, 2014), pp. 18–19Google Scholar.

9 As argued by Gordon, Matthew S., ‘Yearning and disquiet: al-Jāḥiẓ and the Risālat al-Qiyān’, in Al-Jāḥiẓ: A Muslim Humanist for Our Time, (eds) Arnim Heinemann et al. (Würzburg, 2009), pp. 253–269Google Scholar.

10 Gordon, Matthew S., ‘Preliminary remarks on slaves and slave labor in the third/ninth century ʿAbbāsid empire’, in Slaves and Households in the Near East, (ed.) Laura Culbertson (Chicago, 2011), pp. 71–84Google Scholar at p. 71; Perry, Craig, ‘Historicizing slavery in the medieval Islamic world’, International Journal of Middle East Studies 49 (2017), pp. 133–138CrossRefGoogle Scholar at p. 133. See also Bray, Julia, ‘Men, women and slaves in Abbasid society’, in Gender in the Early Medieval World: East and West, 300–900, (eds) Leslie Brubaker and Julia M. H. Smith (Cambridge, 2004), pp. 121–146Google Scholar at pp. 136–139.

11 For this phenomenon, see Grob, Eva Mira, Documentary Arabic Private and Business Letters on Papyrus: Form and Function, Content and Context (Berlin and New York, 2010), pp. 193–197CrossRefGoogle Scholar.

12 Relevant characteristics: the obliquely written word bi-sm in the basmala starts above the line at about twice the height of the script; the sīn and shīn are written with and without teeth; with the exception of an elongated jīm in in line 6 and an elongated shīn in شا in line 18, the document exhibits no use of linea dilatans; a large initial ʿayn in in lines 3 and 20, in in line 13, in in line 16, and in in line 17; in general, the orientation of the overall writing line vis-à-vis the margins is straight; and the papyrus has no left margin. See Eva Mira Grob, ‘A catalogue of dating criteria for undated Arabic papyri with “cursive” features’, in Documents et histoire: Islam, VIIe–XVIe s.: Actes des journées d’études Musée du Louvre/EPHE, mai 2008, (ed.) Anne Regourd (Geneva, 2013), pp. 115–135 at pp. 130–134 and 137–138. Theodore C. Petersen, A Collection of Papyri: Egyptian, Greek, Coptic, Arabic, Showing the Development of Handwriting, Dating Mainly from the Second Century B.C. to the Eighth Century A.D. (New York, 1964), p. 83, no. 135 dates the document to 135/849–50 (sic). The document does not support this date.

13 Grob, Documentary Arabic, pp. 95 and 181.

14 That is, Abū al-Ḍaḥḥāk.

15 That is, the self-enslaved persons. The suffix -hā in refers to , ‘ourselves’, on the preceding line.

16 Instead of ‘destitute’, one could also read ‘naked’; cf. ‘with no one … to clothe us’ later on the same line.

17 The author wrote here the name of his patroness, perhaps Sutayt. See the commentary to line 16 below.

18 Papyrological abbreviations used in this article are those of ‘The Checklist of Arabic Documents’ (available at http://www.naher-osten.lmu.de/isapchecklist [accessed 22 December 2022]) and ‘Checklist of Editions of Greek, Latin, Demotic, and Coptic Papyri, Ostraca, and Tablets’ (available at https://papyri.info/docs/checklist [accessed 22 December 2022]).

19 Joshua Blau, A Grammar of Christian Arabic Based Mainly on South-Palestinian Texts from the First Millennium (Louvain, 1966–1967), vol. i, p. 164, para. 59; Joshua Blau, A Handbook of Early Middle Arabic (Jerusalem, 2002), p. 37, para. 36.

20 For descriptions of Maḥbūb b. Rajāʾ and his position at the Tulunid court, see, for example, al-Balawī, Sīrat Aḥmad b. Ṭūlūn, (ed.) Muḥammad Kurd ʿAlī (Cairo, 1939), p. 145 and Ibn ʿAsākir, Taʾrīkh madīnat Dimashq, (ed.) ʿUmar b. Gharāma al-ʿAmrawī (Beirut, 1415–1421/1995–2000), vol. lvii, pp. 71–72.

21 Albert Biberstein de Kazimirski, Dictionnaire arabe-français (Paris, 1860), vol. i, pp. 146–147; Edward W. Lane, An Arabic-English Lexicon Derived from the Best and Most Copious Eastern Sources (London, 1863–1893), vol. i, pp. 231–232.

22 For the plural superseding the dual in Middle Arabic texts, see Blau, Grammar, vol. i, pp. 209–213, para. 106.

23 Recorded in Ibn Mākūlā, al-Ikmāl fī rafʿ irtiyāb fī al-muʾtalif wa'l-mukhtalif fī al-asmāʾ wa'l-kunā wa'l-ansāb, (ed.) ʿAbd al-Raḥmān b. Yaḥyā al-Muʿallimī (Cairo, 1993–1994), vol. v, p. 32; and Ibn ʿAsākir, Taʾrīkh, vol. lxix, p. 191.

24 For the structure of third/ninth-century petitions, see Geoffrey Khan, ‘The historical development of the structure of medieval Arabic petitions’, Bulletin of the School of Oriental and African Studies 53.1 (1990), pp. 8–30 at pp. 11 and 13.

25 On the marking of the basmala, see Grob, Documentary Arabic, pp. 188 and 191–192. On spacing in Arabic letters, see ibid., p. 190.

26 See CPR XXXII 4, commentary to line 5, and P.Heid.Arab. II 10, commentary to line 2.

27 For example, the third/ninth-century documents Khan, ‘Historical development’, no. 1, line 4; no. 2, line 3; no. 3, line 2; P.Marchands II 29; and the document edited in Naïm Vanthieghem, ‘Violences et extorsions contre des moines dans la region d'Assiout: Reédition de P.Ryl.Arab. II 11’, Journal of Coptic Studies 18 (2016), pp. 185–196 at p. 189, line 5. See also CPR XVI 3; P.Heid.Arab. II 10, 27, 50 and 52; and P.Khalili I 23. For a fourth/tenth-century example, see P.Heid.Arab. III 28, line 1. Other petitions or petition-like letters use semantically cognitive expressions, such as the second/eighth-century documents published as Jean David-Weill et al., ‘Papyrus arabes du Louvre, III’, Journal of the Economic and Social History of the Orient 21.2 (1978), no. 23 recto (line 3) and CPR XVI 9 (line 7), which expositions start with ukhbiruka, ‘I inform you’.

28 See the commentary to line 4.

29 Khan, ‘Historical development’.

30 For such use of linea dilatans, see Grob, Documentary Arabic, p. 188.

31 For different types of Abbasid-era prisons and those who held authority over them, see Irene Schneider, ‘Imprisonment in pre-classical and classical Islamic law’, Islamic Law and Society 2 (1995), pp. 157–173 at p. 169; Mathieu Tillier, ‘Prisons et autorités urbaines sous les Abbassides’, Arabica 55 (2008), pp. 387–408, especially at pp. 392–402. For Egypt, see also Petra M. Sijpesteijn, ‘Policing, punishing and prisons in the early Islamic Egyptian countryside (640–850 CE)’, in Authority and Control in the Countryside: From Antiquity to Islam in the Mediterranean and Near East (Sixth–Tenth Century), (eds) Alain Delattre, Marie Legendre and Petra M. Sijpesteijn (Leiden and Boston, 2018), pp. 547–588 at pp. 567–571 and Mathieu Tillier and Naïm Vanthieghem, ‘Un registre carcéral de la Fusṭāṭ Abbasside’, Islamic Law and Society 25 (2018), pp. 319–358 at pp. 347–357.

32 Distress is frequently associated with imprisonment; see, for example, History of the Patriarchs of the Coptic Church of Alexandria, (ed. and trans.) Basil T. A. Evetts, vol. iii: Agathon to Michael I (766), in Patrologia Orientalis 5 (1910), pp. 1–215 at pp. 138, 179 and 182. The third/ninth-century author of P.Khalili I 16, a petition, also writes about the ‘disease and distress’ caused by imprisonment. See also Rosenthal, Concept of Freedom, p. 64.

33 Mathieu Tillier, ‘Vivre en prison à l’époque Abbasside’, Journal of the Economic and Social History of the Orient 52 (2009), pp. 635–659; Mathieu Tillier, ‘Dans les prisons d'Ibn Ṭūlūn’, in Savants, amants, poètes et fous: Séances offertes à Katia Zakharia, (ed.) Catherine Pinon (Beirut and Damascus, 2019), pp. 233–251 at pp. 242–243.

34 History of the Patriarchs of the Egyptian Church, Known as the History of the Holy Church, (ed. and trans.) Yassā ʿAbd al-Masīḥ and Oswald H. E. Burmester (Cairo, 1943), vol. ii, p. 55.

35 For expropriation (ḥajr), see Émile Tyan, ‘Iflās et procedure d’éxécution sur les biens en droit musulman (maḏhab ḥanafite)’, Studia Islamica 21 (1964), pp. 145–166.

36 Al-Kindī, Kitāb al-Wulāt wa-kitāb al-Quḍāt, in The Governors and Judges of Egypt, (ed.) Rhuvon Guest (Leiden, 1912), pp. 463–466; see also History of the Patriarchs of the Coptic Church of Alexandria, (ed. and trans.) Basil T. A. Evetts, vol. iv: Mennas I to Joseph (849), in Patrologia Orientalis 10 (1915), pp. 357–551 at p. 546. For a general discussion, see EI 2, s.v. ‘Muṣādara’.

37 For example, al-Balawī, Sīra, pp. 317, 318.

38 Ibid., p. 166.

39 For examples of the restitution of confiscated possessions, see ibid., pp. 329–330; al-Kindī, al-Wulāt wa'l-Quḍāt, p. 465; and History of the Patriarchs, (eds) ʿAbd al-Masīḥ and Burmester, vol. ii, p. 51 (tr.: vol. i, p. 75).

40 Tillier, ‘Vivre en prison’, p. 648; Sijpesteijn, ‘Policing, punishing and prisons’, p. 569.

41 Shāfiʿī scholars allowed a debtor to keep enough money for one day: al-Shāfiʿī, Kitāb al-Umm, (ed.) Rifʿat F. ʿAbd al-Muṭṭalib (al-Manṣūra, 1422/2001), vol. iv, p. 437; al-Buwayṭī, Mukhtaṣar al-Buwayṭī, (ed.) Ayman b. Nāṣir al-Salāyimī (unpublished MA thesis, al-Jāmiʿa al-Islāmiyya bi'l-Madīna, 2009–2010), p. 704, no. 2307; al-Muzanī, Mukhtaṣar al-Muzanī fī furūʿ al-shāfiʿiyya, (ed.) Muḥammad ʿA. Shāhīn (Beirut, 1419/1994), p. 144. Most Mālikī scholars held that a debtor should be allowed to keep as much money as he would need to sustain himself for a month: Saḥnūn, al-Mudawwana al-kubrā (Beirut, 1415/1994), vol. iv, p. 73, and the sources cited in Ibn Abī Zayd al-Qayrawānī, al-Nawādir wa'l-ziyādāt ʿalā mā fī al-Mudawwana min ghayrihā min al-ummahāt, (ed.) Muḥammad Amīn Būkhubza (Beirut, 1999), vol. x, p. 8. See also Ibn al-Qāṣṣ, Adab al-qāḍī, (ed.) Ḥusayn Kh. al-Jubbūrī (Taif, 1409/1989), vol. ii, pp. 418–419, no. 461.

42 The translation is by Walter Crum, who edited the text. P.Mon.Epiph. 176 contains a similar accusation of being forgotten by the addressee.

43 History of the Patriarchs (ed. and trans. Evetts), vol. iv, p. 377.

44 For examples of lists of prisoners, see CPR XXII 35, P.Horak 64–66, and SPP X 252, all dated to the second/eighth century. See also Tillier, ‘Prisons et autorités urbaines’, pp. 391–392; Mathieu Tillier, ‘Les prisonniers dans la société musulmane (IIe/VIIIe–IVe/Xe siècle)’, in Dynamiques sociales au Moyen Âge en Occident et en Orient, (ed.) Élisabeth Malamut (Aix-en-Provence, 2010), p. 191–212 at p. 195; and Tillier and Vanthieghem, ‘Un registre carcéral’, pp. 349–350.

45 Rosenthal, Concept of Freedom, pp. 44–45; Schneider, ‘Imprisonment’, p. 169; Tillier, ‘Les prisonniers’, p. 197.

46 Like its prototype in Baghdad, Ibn Ṭūlūn's Muṭbaq had oubliettes. See Tillier, ‘Prisons d'Ibn Ṭūlūn’, p. 234.

47 Ibn al-Dāya, Kitāb al-Mukāfaʾa wa-ḥusn al-ʿuqba, (ed.) Maḥmūd M. Shākir (Beirut, n.d.), p. 9; cf. al-Balawī, Sīra, p. 234. See also ibid., p. 220, where Ibn Ṭūlūn is said to have remembered (dhakara) a prisoner who had been imprisoned in the Muṭbaq for over a year.

48 See also the commentary to line 7 above.

49 See EI 2, s.v. ‘Ghulām’ for the wide range of meanings of this word.

50 Koby Yosef, ‘The term mamlūk and slave status during the Mamluk sultanate’, Al-Qanṭara 34.1 (2013), pp. 7–34, discusses on pp. 10–12 the use of the word mamlūk to express obedience and subordination in the Mamluk period. The word was used in this sense in the Abbasid period as well, albeit not very often. For instance, the fourth/tenth-century author of P.Vind.Arab. III 38, a petition addressed to a ‘lord’ or ‘patron’ (mawlā), formulates his request as follows, using terminology that denotes servile status to express his subordination to the addressee: ‘his slave (ʿabd) asks him to look into his, that is his slave's (mamlūk), situation’. The third/ninth- or fourth/tenth-century P.Ryl.Arab. I § VI 10, a petition too, uses slave terminology in a similar way. Its author calls the addressee his ‘lord’ or ‘patron’ (mawlā) and identifies himself as the addressee's ‘slave (ʿabd) and his servant (khādim)’. For a possible early second/eighth-century example of the use of ʿabd, ‘slave’, to express subordination, see the document transcribed in Esther Garel and Naïm Vanthieghem, ‘Nouveaux textes sur les pagarques du Fayoum au VIIIe siècle’, in Christians and Muslims in Early Islamic Egypt, (ed.) Lajos Berkes (Ann Arbor, 2022), pp. 87–125, at p. 88, n.12.

51 Roy Mottahedeh, Loyalty and Leadership in an Early Islamic Society (Princeton, 1980), pp. 82–95; Marina Rustow, ‘Formal and informal patronage among Jews in the Islamic East: evidence from the Cairo Geniza’, Al-Qanṭara 29.2 (2008), pp. 341–382.

52 Rustow, ‘Formal and informal patronage’, pp. 360–362, argues that ʿināya belonged to the idiom of patronage in fifth/eleventh through to seventh/thirteenth-century Egypt. By that time, solicitude had already been a way to express patronage for two centuries. The author of the second/eighth-century private letter CPR XVI 26, for example, points at his ʿināya towards the addressee as a reason why they should treat him lovingly. The author of another private letter, P.Hamb.Arab. II 45 (dated to before 310/922–923), identifies himself as the addressee's protégé (ṣanīʿa; see Mottahedeh, Loyalty and Leadership, p. 83, pace Werner Diem, ‘Philologisches zu den arabischen Papyri der Hamburger Staats- und Universitäts-Bibliothek’, Zeitschrift für arabische Linguistik 45 [2006], pp. 7–54 at p. 44) and writes that one Abū al-Ḥusayn had helped him in a dispute while ‘he was not aware of your solicitude (ʿināya) towards me’ (with Diem, ‘Papyri der Hamburger Staats- und Universitäts-Bibliothek’, pp. 44–45). The formulaic request to convey greetings to ‘all those whom your solicitude (ʿināya) encompasses’ (cf. Rustow, ‘Formal and informal patronage’, p. 361) is found in a number of fourth/tenth-century letters: CPR XXXII 4, P.Cair.Arab. V 327 (with corrections in Werner Diem, ‘Philologisches zu arabischen Dokumenten: II. Dokumente aus der Sammlung der Egyptian Library in Kairo’, Zeitschrift für arabische Linguistik 56 [2012], pp. 27–78 at pp. 60–63), P.Vind.Arab. I 11, and P.Vind.Arab. II 17; see also the late third/ninth-century P.Hamb.Arab. II 24, which has in line 14 ‘and peace upon Abū Bakr and all those to whose affairs you attend’ (tuʿnā bi-amrihi).

53 For example, al-Ṭaḥāwī, Sharḥ mushkil al-āthār, (ed.) Shuʿayb al-Arnāʾūṭ (Beirut, 1415/1994), vol. iv, pp. 229–231, with references in the footnotes. See also Rosenthal, Concept of Freedom, p. 30; Rāġib, Actes de vente, vol. ii, p. 24, para. 58 and Yūsuf Rāġib, ‘Esclaves et affranchis trahis par leur nom dans les arts de l'Islam médiéval’, in Les non-dits du nom: Onomastique et documents en terre d'Islam. Mélanges offerts à Jacqueline Sublet, (eds) Christian Müller and Muriel Roiland-Rouabah (Beirut, 2013), pp. 247–301 at p. 252.

54 In the current corpus of published papyri and paper documents dating from the Abbasid period, the word mamlūk appears in a non-legal context only in the second/eighth- or third/ninth-century P.Prag.Arab. 19, a fragment of a list of day-workers including mamālīk (line 12), and in the fourth/tenth-century P.Vind.Arab. III 38, referred to in note 50 above. In other Abbasid-era documents, the word appears in a deed of slave acquisition from 282/895 (P.Vente 4), an emancipation contract from 304/916–917 (Adolf Grohmann, ‘Arabische Papyri aus den Staatlichen Museen zu Berlin’, Der Islam 22 [1935], pp. 1–68, no. 7) and a fourth/tenth-century petition asking a qāḍī al-quḍāt to give a slave the inheritance he is entitled to (P.Vind.Arab. III 64).

55 Cf. Schneider, Kinderverkauf und Schuldknechtschaft, pp. 304–307 and Motzki, ‘Der Prophet und die Schuldner’ for the existence of debt bondage and self-dedition among pre-third/ninth-century Muslim communities.

56 Oßwald, Islamische Sklavenrecht, pp. 31–32, 41, 194. For the enslavement of a person with unknown legal status, see Ibn Abī Zayd al-Qayrawānī, al-Nawādir wa'l-ziyādāt, vol. ix, pp. 394 and 396, ascribing this opinion to the Mālikī jurists Ibn al-Mawwāz, Ibn ʿAbd al-Ḥakam, and Saḥnūn. See also Rosenthal, Concept of Freedom, p. 32 and Cristina de la Puente, ‘Islamic law, slavery, and feelings: a fourth/tenth-century Andalusi notarial model on the manumission of an unruly and bad-tempered female slave’, Hawwa 19 (2021), pp. 294–313.

57 In 151/768–769 wealthy persons are recorded as trying to take advantage of the financial pressure on the poor as a result of the poll-tax to buy their children; see The Chronicle of Zuqnīn: Parts III and IV, A.D. 488–775, (trans.) Amir Harrak (Toronto, 1999), p. 244. For other examples of the illegal sale of wives and children under duress, see Ibn Faḍl Allāh al-ʿUmarī, Masālik al-abṣār fī mamālik al-amṣār (Abu Dhabi, 1424/2003), vol. xxvii, p. 518; Ibn Ḥajar al-ʿAsqalānī, Kitāb al-Durar al-kāmina fī aʿyān al-miʾa al-thāmina (Hyderabad, 1348–1350/1929–1931), vol. iv, p. 319; al-Maqrīzī, Kitāb Ighāthat al-umma bi-kashf al-ghumma, (eds) Muḥammad M. Ziyāda and Jamāl al-Dīn M. al-Shayyāl (Cairo, 1359/1940), p. 34. See Ibn Abī Zayd al-Qayrawānī, al-Nawādir wa'l-ziyādāt, vol. v, p. 158, citing the Cordoban jurist al-ʿUtbī (d. 255/869), for the Egyptian ʿAbd al-Raḥmān b. al-Qāsim's (d. 191/806) discussion of a legal case involving a man who sold his free wife during a famine. In al-Ḥarīrī's (d. 516/1122) thirty-fourth maqāma, Abū Zayd sells his son because, he explains after the sale has been cancelled, ‘at the time I had no choice, I didn't have a dime’; see al-Ḥarīrī, Maqāmāt Abī Zayd al-Sarūjī, (trans.) Michael Cooperson, Impostures (New York, 2020), p. 322.

58 In addition to the example that follows, in the early second/eighth century, monks in northern Syria were almost sold in order to compensate for lost gold entrusted to their monastery (Chronicle of Zuqnīn, pp. 153–154). According to Ibn ʿIdhārī, al-Bayān al-mughrib fī akhbār al-Andalus wa'l-Maghrib, (eds) G. S. Colin and É. Lévi-Provençal (Beirut, 1983), vol. i, p. 140, one Ibn Jimāl, a late third/ninth-century qāḍī of Raqqāda in North Africa sold himself in his youth during ‘days of hardship’. Self-sale also appears in a fictional story in a fifth/eleventh-century faraj baʿd al-shidda book by the North African scholar and rabbi Ibn Shāhīn (d. 453/1062); see Mark R. Cohen, Poverty and Charity in the Jewish Community of Medieval Egypt (Princeton and Oxford, 2005), pp. 167–168.

59 ʿAbd al-Laṭīf al-Baghdādī, Kitāb al-Ifāda wa'l-iʿtibār, (ed. and trans.) Tim Mackintosh-Smith, A Physician on the Nile: A Description of Egypt and Journal of the Famine Years (New York, 2021), p. 155.

60 Oßwald, Islamische Sklavenrecht, pp. 69–71.

61 Ingrid Mattson, ‘Status-based definitions of need in early Islamic zakat and maintenance laws’, in Poverty and Charity in Middle Eastern Contexts, (eds) Michael Bonner, Mine Ener and Amy Singer (Albany, 2003), pp. 31–51 at pp. 44–47; Oßwald, Islamische Sklavenrecht, pp. 182–185.

62 Qurʾān 2:177 and 9:60.

63 Al-Ṭabarī, Jāmiʿ al-bayān ʿan taʾwīl āy al-Qurʾān, (ed.) ʿAbd Allāh b. ʿAbd al-Muḥsin al-Turkī (Cairo, 1422/2001), vol. iii, p. 84 and xi, pp. 523–525.

64 The author of P.World, 149–150, a payment order dated 200/816, transfers one-sixth of a dinar to a khādim (servant, eunuch) named Nuṣayr ‘for balancing his legal claim’. Another third/ninth-century document is an order for the delivery of Palestinian oil to an unnamed ghulām at a reduced price (Petra M. Sijpesteijn, ‘Making the private public: a delivery of Palestinian oil in third/ninth-century Egypt’, Studia Orientalia Electronica 2 [2014], pp. 74–91 at p. 76). The fifth/eleventh-century P.Heid.Arab. III 19 is an order to give a ghulām named Mufliḥ (‘Lucky’) a jar of wine at the author's expense. See also the discussion in Sijpesteijn, ‘Making the private public', pp. 80–83.

65 Hain, Kathryn A., ‘Epilogue: avenues to social mobility available to courtesans and concubines’, in Concubines and Courtesans: Women and Slavery in Islamic History, (eds) Matthew S. Gordon and Kathryn A. Hain (New York, 2017), pp. 324–339Google Scholar at pp. 328–329; Oßwald, Islamische Sklavenrecht, pp. 204–205.

66 Al-Kindī, al-Wulāt wa'l-Quḍāt, pp. 343–344. See also al-Khaṣṣāf, Kitāb Adab al-qāḍī, (ed.) Farḥāt Ziyāda (Cairo, 1978), pp. 423–425, who discusses two cases in which a woman first acknowledged her slave status and later claimed to have been born free.

67 Cf. Ehud R. Toledano, As if Silent and Absent: Bonds of Enslavement in the Islamic Middle East (New Haven and London, 2007), pp. 156–203, who argues that the vulnerability of slaves and freedmen in the late Ottoman empire caused by a disturbed master-slave/patron-client relationship made them more inclined than others to criminal behaviour.

68 A study of possible tensions between slavery laws and legal practice in Abbasid society remains a desideratum, especially because Arabic papyrus and early paper documents indicate that such tensions existed. See, for example, a slave dealer's ledger recording the illegal sale of a minor (ṣabī) separate from one of his parents (Donald S. Richards, ‘Fragments of a slave dealer's day-book from Fusṭāṭ’, in Documents de l'Islam médiéval: Nouvelles perspectives de recherche, [ed.] Yūsuf Rāġib [Cairo, 1991], 89–96); a private letter referring to the unlawful sale of a mudabbar slave (Bruning, ‘Slave trade dynamics’, no. 4, with the discussion on pp. 696–699); and a letter stating that a runaway slave woman got married without her owner's permission (P.Philad.Arab. 74). See also the references in note 7 above.

69 Islamic jurists stipulated that illegally enslaved persons could only regain their freedom if they could provide proof of their free origins. See Rosenthal, Concept of Freedom, pp. 32–33; Schneider, Kinderverkauf und Schuldknechtschaft, pp. 24–26; and Rāġib, Actes de vente, vol. ii, pp. 27–28, paras. 66 and 67.

70 Onimus, Clément, ‘Les mawālī en Égypte dans la documentation papyrologique, Ier-Ve s. H.’, Annales islamologiques 39 (2005), pp. 81–107Google Scholar at pp. 95–100. For clients living in their former master's house or in his vicinity, not discussed by Onimus, see, for example, P.JoySorrow 17 (second/eighth century) and Tillier and Vanthieghem, ‘Un registre carcéral’, no. 1 (possibly 190/806), lines 20–22.

71 Khan, ‘Historical development’, no. 1. The second/eighth-century letter P.JoySorrow 17 offers another example. Its author helps a freedman named Muqassam establish new social ties (or improve existing ones, the letter is unclear) after his old master died.

72 Crone, Patricia, Roman, Provincial and Islamic Law: The Origins of the Islamic Patronate (Cambridge, 1987), pp. 36–38CrossRefGoogle Scholar; Gleave, Robert, ‘Patronate in Shīʿite law’, in Patronate and Patronage in Early and Classical Islam, (eds) Monique Bernards and John Nawas (Leiden and Boston, 2005), pp. 134–166CrossRefGoogle Scholar, especially pp. 155–163; Mitter, Ulrike, Das frühislamische Patronat: Eine Studie zu den Anfängen des islamischen Rechts (Würzburg, 2006), pp. 152–154Google Scholar. For the transference of the patronate to others in pre-third/ninth-century Islamic law, see Mitter, Ulrike, ‘Wenn Frauen Sklaven freilassen: Zur Stellung der Patronin im frühislamischen Recht’, Asiatische Studien 56 (2002), pp. 407–437Google Scholar at pp. 427–430, and Mitter, Das frühislamische Patronat, pp. 339–525.

Figure 0

Figure 1. P.CtYBR inv. 1873 qua. Source: Yale University, Beinecke Rare Book and Manuscript Library.