INTRODUCTION1

1Some twenty-five years ago, the household was identified as an integral feature of Ottoman ruling-class society.2 «[N]ot only the sultan but also the leading officials of the empire supported households, made up of their ’inner’ and ’outer’ retinues, needed to fulfill their official fonctions.»3 As the Ottoman state grew from a small frontier principality into a far-flung empire, the sultans elevated themselves by building up an imperial household that in the fifteenth century consisted of hundreds of slaves and thousands of troops, the empire’s first standing army. Members of the military-administrative elite assembled households of their own that were miniature versions of the sultan’s. The maintenance of a household «the size of which was commensurate with rank and revenues received» was required of tīmār (military assignment) holders and higher provincial officiais.4

2Although there is some disagreement as to the timing of the change, during the seventeenth century, service in the household of a high officiai became an alternative to service in the imperial household, as a means of achieving promotion in the higher ranks of the central and provincial administrations.5 Thus, as Jane Hathaway writes, by the seventeenth century:

«The imperial household faced rivalry, or at least a diffusion of its own power, - [in Istanbul and the provinces] from lesser households. Competition came from the households of veziers and provincial governors, many of whom had begun their careers in the imperial palace, and from groups of soldiers that coalesced in the barracks of the imperial Janissaries in the capital.»6

3Just as the provincial governors’ households were smaller replicas of the imperial household, so the competition between the latter and the vezier and pasha households was replicated in the provinces, the households of the governors facing competition for influence and resources from those of the local grandees.7

4The study of these households and their role in politics is not as advanced as one might expect, given their acknowledged importance, and even less attention has been devoted to household formation among civilian elite elements such as merchants and ‘ulamā’. This is a testament to the influence of nationalist historical narratives - narratives in which the period from the late sixteenth century through the end of the eighteenth century (the heyday of the politics of households) is defined as one of decline and stagnation. Placed in juxtaposition to this dark image, the picture of progress and national awakening in the nineteenth and twentieth centuries shines all the more brightly.8

5Historians of Ottoman Egypt have paid relatively more attention to the households of local grandees than in the other former provinces of the empire, but even for Egypt the subject has not been extensively investigated. David Ayalon and P.M. Holt were pioneers in this area, though their main concern was the question of continuity between the household system of the Mamluk Sultanate and that of Ottoman Egypt, and their studies were based on a limited number of literary sources. Daniel Crecelius and Hathaway have gone some distance farther, on the basis of extensive research in Egyptian and Ottoman manuscripts and archives. Although a political history, Crecelius’ study of the regimes of ‘Alī Bey al-Kabīr and his successor, Muḥammad Bey Abū al-Dhahab, necessarily discusses the nature of the military households and household politics in that era.9 In her study of the origin and rise of the Qazdūghlī (Qazdāghlī) household, which under the leadership of ’Alī Bey, Muḥammad Bey, and their successors dominated Egypt during the mid to late eighteenth century, Hathaway pays particular attention to the household system, arguing that it should be understood as «a regional variation on a household-based elite culture that existed throughout the Ottoman Empire and that integrated the imperial center with its provinces.»10 Thus our understanding of the politics of households in Ottoman Egypt should proceed from the recognition that elite households and household politics were an empire-wide phenomenon, rather than from the perspective of Egypf s pre-Ottoman history. While there is much to recommend Hathaway’s argument, it should also be recognized that the household was not an invention of the Ottoman Empire nor unique to it. Households were also integral to the Mamluk Sultanate, and seem to have derived from elements of pre-Islamic Near Eastern and Turco-Mongol culture.11

6Another fruitful perspective on the households of Ottoman Egypt has been offered recently by scholars interested in the history of women and gender, who necessarily have focused on the ruling and civilian upper classes, due to the nature of the evidence available. Hathaway belongs to this group,12 along with Mary Ann Fay, Nelly Hanna, and Afaf Marsot,13 who have exploited chronicles, waqf documents and court records to bring to light the roles of women in building alliances between households, maintaining cohesion within them, and managing their finances. Studies such as these have suggested that due to the decentralization of politics associated with the household system and also due to their crucial role within that system, elite women in that era exercised more power and freedom than in the nineteenth century, when the household system became defunct with the consolidation of central state power.

7This essay is concerned with another, related aspect of the elite households of eighteenth-century Egypt, namely the question of how they reproduced themselves over time, materially and socially. It approaches this question through a micro-historical examination of two prominent households in the Delta town of al-Manṣūra in the eighteenth and early nineteenth centuries, based on the records of the local Sharī‘a court. The al-Bayklī/Talkī and Farḥātī households (Figs. 1 and 2) were by no means as prominent and powerful as the leading households of Cairo, but even so they were part of the household system of Ottoman Egypt, and the study of lesser households such as these can contribute to our understanding of the system as a whole. Like the great households of Cairo, the al-Bayklī/Talkī and Farḥātī households were military in origin and derived most of their income from the possession of tax farms and offices (discussed below). With regard to their material reproduction, the court records show that by the early eighteenth century these two households had achieved, in effect, hereditary control of certain tax farms and endowed land and offices. One can infer from this that hereditary control of tax farms, endowed land, and offices was a characteristic of the household system in Egypt in general. Second, just as in the great households, in the lesser households endogamous marriage had a key role in social reproduction, both as a means of legitimating the succession of a new head and of incorporating non-kin members, who were often of slave origin.

8One important difference between the lesser households and the great households of Cairo is that while the latter were destroyed by Muḥammad ‘Alī Pasha, many of the former survived into the early nineteenth century. By following the history of the surviving households into the early nineteenth century we can address the question of whether and how the status of elite women changed with the demise of the household system. There are, indeed, indications that the surviving households adopted new strategies of marriage and heirship that were less favorable to women than in the eighteenth century.

9Before taking up this discussion, however, it is necessary to arrive at a definition of the kind of «household» to be discussed.

THE MEANING OF BAYT IN EIGHTEENTH-CENTURY EGYPT

10What is a family and what is a household? The answer depends to a large extent upon the social and historical context in which the question is asked. Jean-Louis Flandrin has pointed out that in France and England the meaning of family and household was conflated only during the past two hundred years. This development does not reflect the rise of the simple or «nuclear» family household, which was numerically predominant much earlier, but rather the rise of the idea of the nuclear family household as normative within these cultures. In the pre-modern era «the word ’family’ more often referred to a set of kinsfolk who did not live together, while it also designated an assemblage of co-residents who were not necessarily linked by ties of blood or marriage.»14 Similarly, in pre-modern Europe, the terni «household» was not restricted to one’s co-resident kin, but instead «the notion [was] that a household comprises all those who share their food with one another,» and who are dependent upon the same household head. The members of such a household could reside in more than one dwelling.15 These findings provide at least a starting point for understanding elite households in Ottoman Egypt. The latter also comprised large numbers of non-kin, not all of whom resided in the same dwellings, but who were dependent on a single head.

11Such a household was called a bayt (p. buyūt) in Arabie,16 and the term is usually rendered in English as «house» or «household.» Bayt carries roughly the same range of meanings and ambiguities as its English equivalents. It can mean, for example, a dwelling, a man’s wife and dependents, or a noble lineage.17 In the context of the elite households of eighteenth-century Egypt the term bayt carried an additional cluster of connotations: it could refer to a «political household» or faction comprised of numerous and dispersed individual household complexes that were attached to a leader and his personal household by relations of clientage (intiṣāb). It also could refer to the «lineage» of the political household, for one can speak of the continuity of certain buyūt across generations, of succession to the headship within them, and of their fission into successor and/or sub-households.18

12Dependency within the bayt was expressed linguistically in more than one way. To begin with, the slaves and former slaves of a household might refer to its head, their master, as «father,» and to each other «brothers.»19 They and other non-kin members of the bayt were also often referred to as atbā‘ (s. tābi‘), a term that has caused some confusion but which appears to connote «those who follow» and serve the household head, including former slaves and other followers who had their own sub-households.20 The conflating of persons of slave and non-slave status and origin under one rubric is an indicator, according to Hathaway, that «clientage overshadowed slave status in defining a person’s position» in society.21 The term tābi‘ appears to have operated in a way similar to the way that kul («slave, servant») was used to refer to persons of both slave and non-slave origin who served the imperial household.22 The tābi‘ or kul who graduated from his patron’s household and established his own household not only continued to «serve» the latter but remained dependent upon him for protection, access to lucrative offices, and even the provision of a suitable marriage partner.

13The term bayt also referred to a large residential complex, sometimes along with smaller, separate residences, in which the dependents of the head lived: his wives, children, and other kin along with large numbers of non-kin (slaves, servants, and military retainers). The typical household of a military grandee in eighteenth-century Egypt has been described as «an entourage of slaves, domestic servants, wives and concubines, bodyguards, and assorted clients who collected at his place of residence.»23 The main residential complex could be quite extensive, having to accommodate a large number of persons and the stores and kitchens to feed them, plus stables for the horses of the military retainers, storage for fodder, and a cistern or well to supply water. Such a residence was designed as a small fortress capable of withstanding a siege of some length. The civilian elite, at the top of which stood the wealthy merchants and high ‘ulamā’, also maintained residential complexes which, however, lacked the military component.

14The residence of a grandee in eighteenth-century Cairo was the scene of constant public activity. There were meetings with other prominent men accompanied by their retinues, and much coming and going by one’s retainers, allies, and business associates. The construction of our modem notion of separate public and private spheres was accomplished by relocating most political and economic activities outside of the household - in offices, factories, and shops - and redefining the residence as a private space.24 Yet in this period no such dichotomy existed: «no public buildings were set aside for the conduct of the state’s or the people’s affairs; instead, the household compound served as the locus of government. The highest organ of government, the imperial council, met within the walls of the imperial palace, the sultan’s home.»25

15Ottoman military-administrative cadres were obliged to maintain their own such households, with inner and outer sections, in emulation of the imperial household.26 In seventeenth- and eighteenth-century Cairo the rising influence of the grandees and the shift of the site of politics from the governor’s residence in the Citadel to their residences is signaled by the use of a new term, the «open house» (bayt maftūḥ). This referred to the consciously open or public, office-like nature of the outer sections of the residential complexes of the leading grandees.27 David Ayalon described the «open house» as «the headquarters in which assemblies and meetings were held, schemes and conspiracies hatched, and from which orders for action were sent.» One could deal a heavy blow to an opponent by plundering or even razing his bayt (main residence). And after a setback for a bayt - say, the death of its head - his successor might marry his late master’s widow, take control of his residence, and «open» it, as the phrase went. To «[re]open» the house/household was to put the bayt back in business.28

16The growing public activity in the grandees’ households was reflected in architecture. Although all great houses had inner and outer sections, Nelly Hanna has found that, parallel with the growth of the public function of such households, there was an apparently heightened concern to segregate the inner space - which, though not exclusively female, was labeled as such - with such new features as a separate «women’s gate» and «women’s kitchen.»29 This perhaps should be interpreted as a move to elevate the status of the household by making its inner space symbolically more remote.30

17A related phenomenon was the physical dispersion of the great households among more than one residence. Members of both the military and civilian elites maintained separate «small houses» and/or «harem houses,» often at some distance from the main residence that served as the headquarters of the bayt. One reason for doing so was to protect at least some valuable household goods and to cache money, as a hedge against the capture, plunder, or destruction of the main house.31 However, in addition to whatever security concerns might go into a decision to maintain «harem houses,» women from households of wealth and standing were capable of insisting - as a condition of marriage - on living in residences separate from the main house and, in the case of multiple wives, from each other.32

THE EVIDENCE FROM AL-MANṢŪRA

18Al-Manṣūra was the principal market town of the sub-province of al-Daqahliyya, which lies on the eastern shore of the Damietta branch of the Nile. It was not a grande ville equal in commercial or political importance to Aleppo, Damascus or Cairo. A regional marketing center, in 1800 it had some thirty caravanserais, compared to more than twice that number in Cairo (including Būlāq and Old Cairo). The mercantile elite (tujjār) of al-Manṣūra traded mainly in the agricultural products of its hinterland - rice, sesame, wheat, and some locally produced textiles - whereas Cairo was a center of international trade, its wealthiest merchants dealing in imported goods such as coffee, spices, and fine textiles. The highest ranking Ottoman official in al-Manṣūra was the qādī, but to judge from the eighteenth-century court cases he was not present often, and may even have resided in Cairo. Not being a provincial capital al-Manṣūra had neither a resident governor nor a citadel, and did not witness the dramatic struggles for power that were typical of Cairo. On the contrary, it was one of the places to which a grandee might be sent in exile.

19Thus the elite households of eighteenth-century al-Manṣūra differed from those of Cairo in one important respect, namely their distance from power, and, perhaps frequently, their distance from the political struggles that took place in Cairo. Most likely they were among those al-Jabartī called «the old houses» (al-buyūt al-qadīma), i.e. the remains or local branches of households that had been politically vanquished or sidelined during the rise of the Qazdūghlī bayt.33 A second and related difference is that these households would have been less wealthy and hence smaller than the ruling households in Cairo. However, their male members probably lived longer and their economic interests may have been more secure. At least some of these «old houses» survived into the mid nineteenth century, largely due to their non-involvement in the political struggles in the capital.

20In sum, then, the two lineages of «old houses» discussed below would have differed from the great households of Cairo in the late eighteenth century in being smaller and less wealthy, in having a greater degree of localization, and greater longevity. In spite of their differences, the great and lesser households are comparable in so far as they sought to acquire and retain status and resources, and resorted to similar strategies of marriage and heirship in pursuit of that goal.

21The Sharī‘a court registers of Ottoman Egypt contain a variety of types of cases that permit the reconstruction of family and household histories, such as probate inventories and inheritance disputes, divorces, contracts, and various kinds of litigation. Whereas women are often absent from other kinds of sources, the court registers have the great merit of containing many cases involving women, who are identified by name and father’s name, making it possible to trace the female as well as the male linkages within families and households. On the other hand, the court registers do not by any means offer a complete picture of any family’s or household’s history, much less a picture of the whole society. The political and economic elite - the military-administrative cadre, the leading merchants, and the high ‘ulamā’ - are over-represented in this source because it was they who regularly participated in public affairs, who had sufficient wealth to use the court frequently for notary purposes, and whose investments in commerce and property were on a scale sufficient to resuit in frequent litigation. The fees charged at the court must also have discouraged more humble persons from making use of it. For these reasons, the lower the social stratum, the less represented they are in the court register.

22Even for the elite, there seems to have been no requirement that their affairs be recorded at the court. For example, no marriage contracts are recorded in this register series at all, and only the occasional divorce suit - and they are suits, i.e. litigation resulting from a dispute, usually alleged nonpayment of the dower. In other words, few persons seem to have felt it necessary to go to the court to notarize marriage contracts or divorce agreements, so long as there was no dispute involved. Similarly, one has the impression that partnerships and transactions in commerce and property were notarized usually after a disputed or an especially complicated affair had been settled out of court -this, from the absence of a large number of relatively simple business dealings in the court record. We have, therefore, at best an incomplete record of the activities of even the elite. This evidence will bear very little quantitative analysis, though as qualitative evidence it is sufficient for at least a preliminary discussion of patterns of marriage and heirship among the elite.

TWO HOUSEHOLD LINEAGES IN AL-MANṢŪRA: HEREDITARY POSSESSION OF TAX FARMS AND OFFICES

23While the material assets of an elite household included slaves, urban properties, household goods, horses, and weapons, its wealth derived mainly from the control of endowments of property and land (waqfs and rizqas) and/or of the offices (waẓā’if) supported by these endowments, as well as from tax farms (iltizāms). Individuals and households controlled, possessed, or held endowments, offices, and tax farms, but strictly speaking they did not own them. Endowments in support of charitable and religious works could be controlled by occupying the position of administrator (nāẓir) - and typically the founder of a private endowment often specified him - or herself and a line of descendents or freed slaves as the administrators. In this way a family or household lineage could control an endowment, enjoying much of its income, over generations. The various other offices funded by endowments, such as the imāma and the khiṭāba of a mosque (the posts of prayer leader and preacher), were also kept in a family and passed on from one generation to the next. Such offices could also be, in effect, purchased. One can therefore speak of a market in endowment-funded offices, some of which were sufficiently lucrative to be sought by the elite.34

24By the eighteenth century there was an analogous situation in iltizām rights. An iltizām consisted of the right to collect the tax due from a particular revenue source, and to keep a portion of it, which was purchased from the state subject to certain conditions. Virtually all open field agricultural land was mīrī, or state revenue land, in which the peasant cultivators held usufructuary rights. The tax farmers of the villages or multazims were not in the position of the feudal landlords of Europe: they were not landowners, but enjoyed their revenue collection rights, legally speaking, at the pleasure of the ruler. Nor were the peasants tied to the soil as in European serfdom. However, by the end of the seventeenth century the multazims were being granted life tenure, and their heirs were favored in the reassignment of iltizāms vacated by their death, making the iltizāms in effect inheritable. By the early eighteenth century, also, a market in iltizām rights had developed: they were bought and sold by individuals, and could be mortgaged.35

25In strictly legal terms the possession of an iltizām meant that one was a servant of the state, occupying one’s office with the permission of the ruler or his deputy. So long as the iltizāms were regarded as state offices, and the multazims as state servants, the latter were recruited from the political class, which was overwhelmingly military, and normatively adult, male, and Muslim. Thus in the early seventeenth century most multazims were from the military elite, the only exceptions being tribal chiefs, in the districts they controlled, and a handful of prominent ‘ulamā’. By the late eighteenth century the composition of the multazims had changed: tax-farming rights had been acquired by a greater number of ‘ulamā’, as well as by merchants and women, although «men of the sword» still predominated.36 On one hand, the participation of a significant number of civilians and especially of women. in tax farming was a consequence of the tendency to treat iltizām rights as assets in a financial portfolio. And since elite women participated extensively in commerce and property ownership, it is logical that they too would acquire iltizām rights. However, their entry into the ranks of the multazims also is an indication of the extent to which, conceptually, the holding of iltizām rights had become separated from the idea of state service.

26The histories of our two Manṣūrī households illustrate the hereditary control of iltizāms and endowment-funded offices that is implicit in these developments. In representing these household lineages (Figs. 1 and 2) I have indicated the members of slave or unknown origin in normal type, and kin relations in boldface type.37

27Figure I shows part of the al-Bayklī/Talkī household lineage of al-Manṣūra during the eighteenth and early nineteenth centuries.38 The earliest head of the lineage that can be identified was Amīr Muḥammad al-Bayklī (d. 1694), katkhudā of the ‘Azabān regiment in Cairo. As katkhudā (kāhya)39 of the second most powerful regiment in the capital, Amīr Muḥammad was of sufficient stature to receive a biographical notice in al-Jabartī’s chronicle. We are told that he kept a residence in Suq al-Silāh which his son Yūsuf, who also became a katkhudā of the ’Azabān, converted into a wikāla in 1698-1699.40 Amīr Muḥammad also maintained a residence in al-Manṣūra and endowed a mosque there, known as Jāmi‘ Muḥammad Katkhudā; at least part of the administration (al-niẓāra) and of the rizqa land supporting this mosque remained in the hands of his household lineage a century later (see below).41 This suggests that his interests in the districts around al-Manṣūra were much more extensive than the four qīrāṭs (twenty-fourths) or one-sixth share of the iltizām rights of the village of Mīt Khamīs that he is known to have held.42 One of his slaves, Amīr Ibrāhīm Abāẓa, also held four qīrāṭs of the iltizām of this village,43 which may have been vacated by the death of his master. One of the slaves of Ibrāhīm Abāẓa, who held two qīrāṭs, represents the third consecutive generation in this household lineage to possess iltizām rights in Mīt Khamīs.44

28After the abolition ofthe iltizām system in 1813-14 by Muḥammad ‘Alī Pasha, many of the multazims of Lower Egypt were permitted to retain their ūsya (a portion of the land in their villages that they held in usufruct) and were paid a stipend in partial compensation for their lost revenues.45 In the court records they and their heirs continued to be called multazims, and cases involving ūsya land and stipends were recorded up to the mid nineteenth century, making it possible to continue to follow the old multazim families. It is evident that the descendants of Amīr Muḥammad al-Bayklī still held iltizām rights in Mīt Khamīs in 1813-14, more than a century after his death, since Ismā’īl Talkī (fl. 1821-42) was still described many years later as the multazim of Mīt Khamīs by right of two qīrāṭs.46

29The administration of the Mosque of Muḥammad Katkhudā in al-Manṣūra also remained at least partly in the hands of his lineage. In 1805 Muḥammad Ḥasan Talkī (fl. 1805-12) ceded half the niẓāra of the mosque to his great uncles ‘Alī and Sālih. The other half may have remained in his possession, for his son Ismā‘īl Talkī had an unspecified share in the administration of the mosque and its endowment, which he probably inherited.47

30Consistent with the long-term, hereditary possession of iltizām shares in the same villages, when it was necessary to exchange shares there was a preference for keeping them within the lineage. Additionally, there was a tendency to cluster the shares held by the members of the lineage in the same and nearby villages, suggesting that they sought to maintain local influence or a local power base, a strategy also followed by the more influential multazims based in Cairo.48 For example, Amīr Ibrāhīm Abāẓa held an undetermined share of the iltizām of Jadīdat al-Hāla during his lifetime. In 1744 his slave, al-Zaynī Khalīl al-Asmar («the Brown,» probably an African), sold five of the six qīrāṭs he held in that village to his co-slave Amīr Aḥmad.49 The remaining qīrāṭ was purchased by a member of another branch of the lineage: Aḥmad, the slave of Muṣṭafā Afandī, who was the slave of Amīr Yusūf al-Bayklī, the son of Amīr Muḥammad al-Bayklī.50 Al-Sayyid Aḥmad, grandson of Amīr Aḥmad, the slave of Ibrāhīm Abāẓa, also held a qīrāṭ of Jadīdat al-Hāla, which he mortgaged in 1760,51 indicating a long-term interest in iltizām shares in this village as well.

31The household lineage of Amīr Khalīl Farḥātī (fl. 1759-1774) (Fig. 2), a jurbajī (çorbaci) of the Tufenkjian corps in al-Manṣūra, was linked to the al-Bayklī/Talkī lineage by the second marriage of Zahra, the granddaughter of Muṣṭafā Rajab al-Bayklī, to ‘Abd al-Raḥmān Ḥasan Farḥātī (fl. 1812). In addition to that the interests of the two bayts coincided in the village of Mīt Khamīs. Amīr ‘Alī, the slave of Amīr Muḥammad al-Farḥātī, purchased one qīrāṭ of the iltizām of Mīt Khamīs in 1744, and the following year ‘ā’isha al-Baydā’, the freed slave of Amīr Fakhr al-Farḥātī, purchased another qīrāṭ there.52 Since the relationship of these two individuals to the line of Amīr Khalīl Farḥātī is unclear, they have not been included in Fig. 2. As for Amīr Khalīl, he and several of his descendants held rights in the rizqa land located in Mīt Khamīs, and his great-great granddaughter Gülsun Khātūn (fl. 1802-36) held six qīrāṭs or a fourth of the iltizām of this village.

32The Farḥātī lineage also maintained shares in the rizqa land and iltizām rights in the villages of Shawā and Shawā Sallant for three consecutive generations during the late eighteenth and early nineteenth centuries. Amīr Muḥammad Farḥātī al-Kabīr (fl. 1781-1802), the son of Khalīl Farḥātī, held a half-share of all the offices in the mosque in Shawā, and a corresponding proportion of the rizqa land designated for its support.53 ‘Alī Farḥātī (fl. 1759-66), the slave of Khalīl Farḥātī, purchased two qīrāṭs of the iltizām of Shawā in 1760. His granddaughters Ṣāliḥa and Khadīja held two additional qīrāṭs of Shawā in 1760, which they sold while still minors.54

33One of the wives of Muḥammad al-Kabīr was Laṭīfa Khātūn (fl. 1800-13), the daughter of Amīr ‘Umar, Aghā of the Jawisān in al-Manṣūra, and multazima of Shawā Sallant by right of five qīrāṭs. Their daughter Sitaita purchased two qīrāṭs of Shawā Sallant from her mother in 1800. Khadūja (fl. 1802-22), another daughter of Muḥammad al-Kabīr by a wife whose name is unknown, purchased a qīrāṭ in each of Shawā Sallant and the nearby villages of Shawā and Kafr ‘Abd al-Mu’min during 1802-1813.55 In the accounts of multazims laying claim to their ūsya land after the cadaster of 1813-14, she is shown as holding four qīrāṭs of Mīt Khamīs and an unspecified share of Shawā Sallant. In 1814 Khadūja received and apparently Consolidated portions of the rizqa land held by her father and grandfather in Mīt Khamīs: twenty-three feddans designated for the support of a hospital (bimaristān), plus two-thirds of a garden.56 Her niece Baihāna (fl. 1813-45) acquired seventeen feddans of the same rizqa in Mīt Khamīs and the remaining third of the garden. She too was listed as the multazima of an unspecified share of Shawā Sallant after 1813.57

34To note that iltizāms and other offices became in effect inheritable and that this implies the hereditary holding of them is one thing. It is another thing to demonstrate that there was hereditary holding in actuality, but this is one of the important conclusions that can be drawn from the history of the Bayklī/ Talkī and Farḥātī lineages, who maintained a long-term interest in iltizām rights and rizqas in at least four nearby villages. It seems likeiy that further research would reveal a similar pattern of behavior among other iltizām and office-holding households, including the great households of Cairo. Hathaway has noted that «grandees and, indeed, entire households were commonly called by the names of the [village] tax farms upon which their wealth was based,» and interprets this in terms of the importance of the income generated by tax farms for the households.58 Households that became known by the name of a certain village or district very likeiy did so because of a long-term association with it, in the form of hereditary possession of tax-farming rights.

MARRIAGE STRATEGIES

35Of equal interest are the patterns of marriage that appear in these lineages. Marriage within the extended household lineage, including members of slave origin and free birth alike, may be regarded as endogamy. After the death of Amīr Yūsuf al-Bayklī, for example, his widow, whose name we do not know, was married by Yusuf s former slave ‘Alī Afandī ‘Azabān.59 Some years later Fāṭima, the daughter of Yūsuf and his widow, was married by one Ḥusayn, a former slave of‘Alī Afandī ‘Azabān.60

36Although the court records permit us to trace household and family lineages, only rarely do they indicate boundaries of households, as factions and residential groups, in a given period. This is one of those rare occasions. We may infer from the first of these two marriages that ‘Alī Afandī ‘Azabān succeedcd Amīr Yūsuf al-Bayklī as the head of his bayt, for marriage to his master’s widow was a common way of legitimating the succession. Among the leading amīrs in Cairo, «when a grandee died, one of his clients typically married the widow or chief concubine. This practice was part of the process of taking over the house of the patron.»61 A union such as this conferred legitimacy upon the successor: the late master’s widow represented the continuity of the household lineage, and as the daughter or former slave of another prominent grandee or grandee’s wife she might also have influential connections with other prominent houses. In this system of succession and legitimation, the master’s widow seems to have remained an unmarried matriarch only in unusual situations, such as when she had an adult son or was very advanced in age.62

37The marriage of Fāṭima, daughter of Yūsuf al-Bayklī, to Ḥusayn, the former slave of ‘Alī Afandī, appears to have been intended to strengthen the ties between a leading member of her stepfather’s personal retinue and the al-Bayklī family. One can infer that as head of the bayt, ‘Alī Afandī arranged the latter marriage, intending to establish Ḥusayn as his successor. Among elite households, including the imperial household, male non-kin were incorporated by marriage to daughters or concubines of the head.63

38Toward the end of the eighteenth century another case of endogamy occurred among the slave and kin «descendants» of Amīr Khalīl Farḥātī. The marriage of ‘Alī Jalabī ‘Atā’ (fl. 1798-1802), the former slave of Muḥammad Farḥātī al-Kabīr, to Fāṭima (fl. 1802), the daughter of Amīr Yūsuf al-Farḥātī, mimicked the familiar father’s-brother’s-daughter pattern of marriage (Fig. 2). That is, the groom in this case was the slave of the son of Khalīl Farḥātī, while the bride was the daughter of the slave of Khalīl Farḥātī. Though we lack any further information about ‘Alī Jalabī and Fāṭima, one may infer that their marriage served to incorporate ‘Alī into the household of Fāṭima’s father, Amīr Yūsuf, probably as the successor of the latter. Evidently ‘Alī could not expect to succeed his own master, Muḥammad al-Kabīr, as head of his household, due to the presence of the latter’s adult son Maḥmūd (d. bef. 1811).

39It may appear that the two Fāṭimas above had little choice but to be «inherited» along with leadership of the households of Yūsuf al-Bayklī and Yūsuf Farḥātī respectively.64 To be sure, the position of elite women of slave origin was defined mainly through relations of marriage or concubinage. Yet to a large extent the same can be said of their male counterparts: marriage to a woman «related» to the master or his wife by blood, concubinage, or service in the inner household was the principal way in which male slaves were integrated into a bayt upon manumission. Marriage to a woman of standing from another household was the way in which alliances were made. Wives such as these were desirable and necessary because of the status that marriage to them conferred upon their husbands, because of the alliances that marriage to them cemented, and because of the wealth they controlled. Yūsuf al-Bayklī’s daughter Fāṭima, for example, was her stepfather’s creditor when he mortgaged four qīrāṭs of the village of Dibū ‘Awwām in 1745.65

40Thus the above cases of endogamy in the al-Bayklī/Talkī and Farḥātī households seem to be capable of explanation in terms of two fundamental aspects of the household system of the eighteenth century. The first is the reproduction of households, which involved not only the purchase of slaves but the incorporation of them and other male non-kin through marriage. The second aspect is succession, which is closely related to reproduction.

41A frequent explanation for endogamy in the Islamic Middle East is that it is one way of preventing the fragmentation and eventual loss of the property of a family, especially landed property. There is indeed a correlation between systems of inheritance in which multiple heirs of both sexes receive shares, as in the Islamic rules of inheritance, and a preference for endogamy.66 However, this explanation does not suffice for the above instances of endogamy because the Islamic rules of inheritance did not apply to the principal assets possessed by elite households. Moreover, when the Islamic rules of inheritance did apply, they could be sidestepped.

42There were a number of options avallable to households to enable them to control the intergenerational transmission of resources without having to resort to endogamy. For example, one could dispose of legal property (milk) by founding a waqf endowment and designating the officers and beneficiaries as one wished, and/or one could bequest up to one-third of an estate. Bequests by grandees must have been rare, however, since many died suddenly and violently, and if they had no surviving children a sole widow would legally inherit a fourth of the estate. According to the Islamic rules of inheritance, former slaves and other non-kin are not eligible to inherit, and so it is true that marriage to the widow of a grandee would give his successor access to the widow’s property. Nevertheless, the principal assets of a household were the tax farms and offices it controlled. Though in effect inherited, as was noted earlier, legally such assets were not property and not subject to the Islamic rules of inheritance, hence they were not necessarily in danger of fragmentation and loss through inheritance and exogamous marriage by women.

43Moreover, tax farms and offices (as well as property) were routinely transferred inter-vivos by sale or gift. In addition to arranging for suitable marriages for their manumitted slaves, the elite habitually resorted to inter-vivos gifts in order to set them up with a residence and retinue of their own. Thus the successor who married his master’s widow and «opened» his house normally would have been the head of his own sub-household/retinue and in control of a certain number of villages and offices for some time beforehand.

44Toward the end of the eighteenth century and especially in the early nineteenth century the strategies of marriage and heirship pursued by the surviving «old houses» changed, due to their changing circumstances. In the second half of the eighteenth century they must have become progressively less wealthy than the leading buyūt and much less capable of reproducing themselves in the customary way, by purchasing and incorporating slaves. In the early nineteenth century, Russian expansion ended the trade in central Asian slaves, and the reforms of Muḥammad ‘Alī deprived the surviving households of much of their remaining income from tax farms and offices. Whether due to straightened circumstances, disruption of the slave trade, or other factors, by the end of the eighteenth century the al-Bayklī/Talkī household lineage appears to have consisted entirely of kin, and similarly, the Farḥātī lineage appears to have included only one person of slave origin, ‘Alī Jalabī ‘Atā’. In other words, Figures 1 and 2 suggest that these households (and others like them) experienced a transformation in their mode of reproduction: in the eighteenth century they reproduced themselves by purchasing and incorporating slaves, but by the nineteenth century they had become households largely or even purely of kin.

45Muḥammad ‘Alī’s reforms also reduced the range of options formerly avallable to households in managing the transmission of property, tax farms, and offices to succeeding generations. In 1813-14 many multazims were permitted to retain some of their ūsya land and were promised a stipend in compensation for their losses. The holders of rizqa land who could verify their claims were also promised a stipend as partial compensation for the taxation of this land, and both kinds of stipend seem to have been paid regularly from no later than the early 1820s.67 Under the new regime the stipends and the ūsya land were subject to the Islamic rules of inheritance. However, they could not be transferred by gift, nor could ūsya land be sold, except to the state. Now, it seems, the principal sources of wealth remaining to the «old houses» were much more likeiy to be divided among multiple heirs of both sexes.68

46In light of these changes, endogamy within the «old houses» may have acquired more of the meaning assigned to it by modern scholars, and the apparent increase in its frequency may not be a coincidence. Starting in either the late eighteenth century or early in the nineteenth century, endogamous marriages of kin occurred in three successive generations within the al-Bayklī/Talkī lineage, and a case of double endogamy occurred within the Farḥātī lineage.

47Raqiyya Khātūn (d. bef. 1812), daughter of Ismā‘īl Aghā, one of Yūsuf al-Bayklī’s slaves, married Muḥammad (d. bef. 1805), son of Amīr Ḥasan, the slave of Ismā’īl Odabāşā, who was the slave of Amīr Muḥammad al-Bayklī. Their son Ḥasan Talkī (fl. 1805-12) married ‘ā’isha (fl. 1825), daughter of Muḥammad Afandī, who was the son of a slave of Amīr Muḥammad al-Bayklī. The son of Ḥasan and ‘ā’isha, Muḥammad Ḥasan Talkī, married and divorced Zahra, whose maternal grandfather was Muṣṭafā Rajab al-Bayklī. The son of the latter couple was the same Ismā’īl Talkī mentioned earlier, who retained shares of the iltizām of Mīt Khamīs and of the niẓāra of the Mosque of Muḥammad Katkhudā.

48The relationship of Amīr Ḥasan Farḥātī (fl. 1759-1802) and Muḥammad Farḥātī al-Saghīr to the rest of the Farḥātī lineage was not explicit in any of the evidence found, but was inferred on the basis of names (indicated by the question marks in Fig. 2). Amīr Ḥasan appears to have been the father of‘Abd al-Raḥmān Ḥasan Farḥātī and Muḥammad Ḥasan Farḥātī, and Muḥammad Farḥātī al-Saghīr («the Younger») was probably the son of Muḥammad Farḥātī al-Kabīr («the Elder»). ‘Abd al-Rahmān’s first marriage was to Umm Hān, the daughter of Muḥammad Farḥātī al-Saghīr and probably his cousin. She was a multazima of Shawā Sallant, and sold a qīrāṭ of the iltizām rights in that village to Khadūja, the daughter of Muḥammad al-Kabīr in 1802.69 His second marriage to Zahra of the al-Bayklī/Talkī lineage produced three children before their divorce: Muḥammad, Ḥasan, and Gülsün, who were listed as multazims of Shawā Sallant after 1813. Muḥammad outlived his brother and sister, and thus inherited from his great-aunt Khadūja, daughter of Muḥammad al-Kabīr. Her estate included iltizām rights in the village of Shawā Qulūnjīl, near al-Manṣūra.70 After divorcing ‘Abd al-Raḥmān Ḥasan Farḥātī, Umm Hān, daughter of Muḥammad al-Ṣaghīr, married his apparent brother Muḥammad, i.e. another cousin. Their daughter Gülsün Khātūn (fl. 1802-36) held six qīrāṭs of the iltizām rights in Mīt Khamīs and unspecified shares in the villages of Shawā and Shawā Sallant.71 Individuals. Legally it could be endowed as waqf, though few multazims seem to have done so, perhaps because Muḥammad ‘Alī discouraged that.

49Keeping in mind the fragmentary nature of the evidence and hence of this picture, one can infer that these endogamous marriages were motivated primarily by a desire to maintain the economic status of the household by keeping such offices, lands, and stipends as held by Ismā’īl Talkī, Muḥammad al-Farḥātī, and Gulsun Khātūn within their respective lineages. The inference is based, first, on the disappearance of slaves from within these households, so that endogamy can no longer have served to integrate non-kin; and second, on the new system of inheritance affecting these households, in which endogamy seems to have been one of the few ways in which a lineage could avoid the fragmentation of offices, lands, and stipends.

CONCLUSION

50This paper has explored two aspects of the reproduction of elite households in eighteenth-century Egypt, the material and the social. The material reproduction of these households was based, in part, on the hereditary control of tax farms, endowments, and offices. Their social reproduction entalled the incorporation of non-kin, who often joined the household as slaves, through endogamous marriage. Like households of kin, these elite households were not stable in their composition, but constantly changing with the addition of new members and the fissioning of sub-households, as the senior members of both sexes «graduated» in the sense of being manumitted, married, and set up with their own residences and retinues.72

51In Egypt, the era of the politics of households was brought to an end by Muḥammad ‘Alī, who used military force to destroy the great households of Cairo and whose reform of the land and endowment regimes deprived the remnants of the household system of a viable material base. Through the mid-nineteenth century the surviving «old houses» lived on in attenuated form, becoming mainly households of kin due to their inability to acquire and incorporate non-kin as members, as they had formerly done.

52The «old houses» continued to practice endogamy, though the meaning of endogamy was now different from what it had been in the eighteenth century. Whereas it had been a means of legitimating succession, incorporating non-kin, and building alliances, it seems to have become to a much greater extent a means by which a male lineage avoided the fragmentation and loss of property and stipends through inheritance and exogamous marriage by women. What were the implications of such a change for elite women?

53Under the old household system the marriage of women as well as men was carefully controlled, for marriage was crucial to the reproduction of a household. For a man, an appropriately chosen wife was desirable and necessary due to the status she conferred upon him, the real as well as symbolic relations with other households she represented, and the wealth she controlled. In other words, if the status of female non-kin was determined by marriage or concubinage, the male non-kin within a household also acquired full membership in it and access to its resources through endogamous marriage. In the nineteenth century, on the other hand, endogamy among the «old houses» appears to have been motivated mainly by a desire to keep females’ shares of the family property within the male lineage. In neither period were women or men free to choose their spouses, the decisions being made by their guardians and patrons in the interest of the status of the household and its reproduction. However, if indeed the main purpose of endogamy in the nineteenth century changed - that is, if the aim was to insure control of the property of women (and younger males) by the head of the household, then this would certainly imply a loss of autonomy and status for women, compared to the eighteenth century.

54This conclusion has to be qualified in two respects. First, it must be regarded as tentative, a hypothesis, indicating one direction in which research might go in pursuing the question of the position of elite women in nineteenth century Egypt. If the demise of the politics of households is a plausible factor in the apparent decline of their status, the new system of property relations that arose, and a corresponding change in strategies of marriage and heirship, may be of equal importance.73 Second, this conclusion applies only to those households that held ūsya land and stipends in the early nineteenth century, i.e. the surviving «old houses» and the households of many officers in the Pasha’s pre-niẓām army, who acquired tax farms before 1813. This is because the Islamic rules of inheritance were applied to ūsya land and stipends, and so endogamy was one of the few ways in which the division of those assets could be avoided.

55Yet the practice of endogamy by the old multazim families may have anticipated the behavior the new elite elements that arose in the nineteenth century: the Ottoman-Egyptian («Turco-Circassian») elite as well as the native elite. The former became a propertied class as a resuit of privileged grants of land by Muḥammad ‘Alī and his successors to themselves, their family, and various officers and officiais. Along with the remaining parcels of ūsya land, these properties came under the rubric of ‘ushr land in 1854, and they too were subject to division according to the Islamic rules of inheritance.74 The majority of the native elite came from the families of rural notables (village shaykhs and ‘umdas, or headmen), who held the lion’s share of the remaining agricultural land, known as athar or kharaj land. The latter category of land was always potentially divisible among male and even female heirs, but during 1858-81 it also was made subject to the Islamic rules of inheritance. While the behavior of the Ottoman-Egyptian elite has not been studied in this respect, rural notable families clearly responded to the changing economic and legal environment in devising strategies to avoid the division and loss of their land.75

56The end of the politics of households in Egypt was not the end of household politics. Muḥammad ‘Alī destroyed the old household system, but in consolidating his position he built up his own ruling household. The nature of the «household government» of Muḥammad ‘Alī and his immediate successors has been discussed already by F. Robert Hunter,76 and is outside the boundaries of the present study. Suffice it to note that even that great «modernizer,» Khedive Ismā‘īl, insured the loyalty of his high officiais by marrying them to royal concubines and including a grant of endowed land in the package.77 This method of rule would not have been unfamiliar to the sultans of previous centuries.

Fig. 1: The Descendants of Amīr Muḥammad al-Bayklī, Katkhudā of the ‘Azabān, 1694-1845

Sources: Registers of Mahkamat al-Manṣūra; al-Jabartī, 1880.

Fig. 2: The Descendants of Amīr Khalil Farḥatī, Jurbajī of the Tufenkjian, 1759-1846

Sources: Registers of Mahkamat al-Manṣūra.