Mohamed Talbi.
The aim of the present study is to underline the dominant characteristics of the agricultural sector in Ifrīqiya in the 3rd/9th century, while attempting simultaneously to extrapolate the social profile of the people who, on different levels, made their living by tilling the soil. Our sources include legal texts, which, contrary to their unjustified reputation as theoretical and unhelpful works, make a significant contribution towards confirming, enriching, and focusing our knowledge.
The greatest jurist of medieval Ifrīqiya, Saḥnūn (160-240/777-854), admits that his knowledge about the legal status of the lands of Ifrīqiya and the Maghrib is, generally speaking, uncertain. Is he referring to the lands classified as 'unwa (taken by force) or ṣulḥ (governed by the terms of an agreement)? Had the original inhabitants kept the ownership of their land by converting to Islam? Nothing certain is known about this, he tells us1—that is, nothing that could be used in theoretical discussions between legal experts. For we do know that in general terms, the distinction between the classification 'unwa or şulḥ was difficult to make precisely in every part of the region, and that the practical results of this classification cannot be easily accounted for in their entirety. But the problem is very clear for Ifrīqiya. It did not present itself in these terms from the Middle Ages on, not even on purely theoretical and legal grounds.
In fact, in Ifrīqiya as elsewhere, the fate of land was determined by the fortunes of conquest and the decisions taken by military leaders with regard to current conditions. Much land was naturally left to its original owners. But many other areas, which had been abandoned because of military operations, or had been taken away from their owners because of their behavior during the fighting, were awarded to the victors as a recompense for their courage and to keep up their zeal. It is necessary to emphasize that these awards—despite the tales which were preserved concerning the allocations made in the time of the Prophet—did not in fact obey any strict or well-defined legal rule, but were essentially governed by the immediate situation. It was Ḥassān ibn al-Nu'mān who, after his definitive victory (78/697-98),2 drew up a system of land classification. In particular, he introduced the concept of the kharāj, either negotiated or imposed on the independent individuals who had been allowed to maintain the ownership of their land,3 and established an administration which had the task of organizing it. He went on to distribute land to the Arabs, naturally, and also to their allies, the Berbers.4 The acreage thus distributed, while certainly large, must have been very variable both in quality and quantity, depending on the rank, services rendered, social status, and ethnic background of the beneficiaries. But unfortunately we do not have any detailed information on this subject. Mūsā ibn Nuṣayr, who replaced Ḥassān and completed the pacification of the country, was notable for his policy of violence and plunder. He undoubtedly embarked on new distributions of land on which, however, we do not have any precise or direct information. As we shall see later, he rewarded his close collaborators abundantly, and also appropriated a lion's share for himself, an act which in the end led to his downfall.5
The land register of Ifrīqiya emerged profoundly changed from the tumult of the Islamic conquest, which was particularly long, drawn-out, and difficult. It must have assumed from this time on the configuration that it was to keep grosso modo until at least the end of the 3rd/9th century.6 In particular, it is quite definitely as a result of these gifts, agreed upon by Ḥassān lbn al-Nu'mān and Mūsā ibn Nuşayr, that the sector of very large Muslim properties came into being, of which there is still evidence in the 3rd/9th century. The only question which will always remain unresolved, because of the nature of our sources, is that of the size of this sector as a percentage in relation to the whole of the economic sphere. Let us examine the references which bear witness to the existence of this sector; all of them are indirect, thus voiding the possibility of deliberate falsification, and hence all the more worthy of belief.
One of the most explicit texts is that preserved by Mālikī. This is the biography of Abū 'Abdullāh Muḥammad ibn Masrūq (d. beginning of the 9th c.) who, on his father's death and following a mystical conversion, abandoned his whole inheritance to lead a life of prayer and total poverty.7 We learn in this connection that his father Masrūq had been one of the prefects of Mūsā lbn Nuṣyr in the Maghrib. Masrūq owned, on the road to Sousse, an entire village to which he had given his name, al-Masrūqīn,8 as well as several other hamlets (manāzil).9
The fortune of a holy tābi'ī, Abū'l-Mughīra 'Abdullāh ibn al-Mughīra ibn Abī Burda al-Qurashī,10 who was the qadi of Ifrīqiya from 99/717-18 to 123/741—that is, until the moment when, in the midst of the Khārijite expansion, Kulthūm ibn 'Iyāḍ replaced 'Ubayd Allāh ibn al-Ḥabhāb as the head of Qayrawān—seems also undoubtedly linked with Mūsā ibn Nu#1E63;yr. Abū'l-Mughīra owned two estates which, judging by their names, each included a hamlet or a village: Qaṣr Mughīra and Qaryat al-Mughīriyīn.11 These two estates must have been left to him by his father, a well-known tābi'ī who, we are told "had participated in the conquest of the Maghrib and Spam with Mūsā ibn Nuṣayr,12 and had also taken part in the siege of Constantinople."13
The case of 'Alī ibn Aslam, a contemporary and foster son of Saḥhnūn, is no less significant. He was the grandfather of Abū Isḥāq al-Jabanyānī, to whom Abū'l Qāsim al-Labīdī had devoted a work of manāqib,14 that is, a saint's life. 'Alī ibn Aslam had, so we are told, built the cathedral mosque of Sfax on his own; he had aiso surrounded the town with a wall of unglazed bricks (ṭūb), and on the coast he had erected guard houses (maḥris) to which he had given his name (Maḥris 'Alī) before they were dubbed al-Maḥris al-Jadīd.15 'Alī ibn Aslam could allow himself such lavishness for the benefit of his city. In fact, "he was master of a large fortune, it is said, and of numerous hamlets (manāzil)—such as Jabanyāna and others as well—where the agricultural holdings were enhanced by wonderful residences (ribā'). In Sfax itself, he had many other agricultural holdings of a similar kind."16 This huge fortune, which was based on the ownership of several hamlets and villages—including Jabanyāna (Djebeniana), which still exists—had its origins, we are told, in the awards made to the Arab conquerors. The family of 'Alī ibn Aslam, which could be traced back to the Bakr ibn Wā'il, a branch of the Rabī'a, was in fact part of the ahl al-khiṭaṭ, who were beneficiaries of gifts of land.17 In addition, to complete the picture, it should be noted that Ibn Aslam possessed a mosque in Qayrawān, and a residence (rab') which gave its name to a whole section of the town, Darb Aslam.18
We shall see later that Abu Muḥhriz (d. 214/829), who was grand qadi of Qayrawān, also owned numerous manāzil, that is, properties large enough to include several hamlets.
Another corroborating piece of evidence is afforded by the story of Manṣūr al-Ṭunbudhī,19 who is better known because of the important part he played in the great revolt of the jund against Ziyādat Allāh (201-23/817-38). Manṣūr, who had been made responsible for the government of Tripoli before he took on the leadership of the rebellion, owned a veritable fortress at Ṭunbudha—hence his name al-ṬTunbudhī—as well as several other hamlets in the area around Tunis.20
Al-Ṭunbudhī was naturally not the only lord of the country to possess vast estates covering a number of hamlets. He had many emulators in the Aghlabid kingdom, and they presented a real threat to the ruler's power. These country squires, who had held their enormous estates since the time of the conquest, were meant in theory to supply the jund, the armed forces which guaranteed the Arab presence in the Maghrib. Under Ibrāhīm II (261-89/875-902), they had long since ceased to fill this role, and had rather become a cause of turmoil, insubordination, fiscal revolt, and (if the occasion was ripe) a threat to public order. In the context of his policy of reducing the power of the nobles, the amir decided to tame them. The extent of the rebellion21 that he provoked throughout the country (280/893), except in the Sahel and the province of Tripoli, provides us with information about both the importance and the establishment of the great estates. These constituted real fiefdoms scattered across the rich agricultural areas, and were sometimes defended by veritable fortresses in cut stone, as in the case of Ibn Abī Aḥmad,22 lord of Bāshshū at the entrance to Cap Bon.
Unfortunately, we are given hardly any information at all about the possessions of the amirs. But we have every reason to suppose that they were the greatest property owners in their kingdoms. On occasion they showed little restraint in confiscating the large properties which aroused their envy, which shows that they owned some. Thus in 275/888-89 Ibrāhīm II appropriated Ibyāna by force. Ibyāna was a large agricultural property, the size of a village, lying five kilometers southeast of Tunis in the plain of Mornag.23 A telling anecdote reveals, moreover, that this amir had a wonderful palm plantation in the south.24 A relative of the ruling family. Ya'qūb ibn al-Maāwho in 232/846-47 had played an important part in the reconciliation of Muḥammad I (226-42/841-56) with his brother Abū Ja'far Aḥmad (who had usurped power), had, we are told, "abandoned the Black and renounced the world,"25 turning to mysticism (nusk). His mysticism and his renunciation of the world had not prevented him, however, from owning numerous estates (ḍiyā')26 in the Jamma area, where the fortified town of Mahdīya would be built in 3 00/912-13.27 Another relative of the dynasty, Ibn Ṭālib (217-75/832-88), who was qadi of Qayrawān twice before being stripped of his possessions by Ibrāhīm II and tortured to death, also had a fine fortune, valued at 80,000 dinars,28 based on the cultivation of huge rural estates.29 At a slightly earlier time than the one under study, we are told that the Fatimid caliph had sizable sugar cane plantations in Sicily, and that the faqīhs, who were particularly uncooperative, boycotted the product as a matter of conscience.30
Thus we do have reliable information provided on a large scale about the existence of great estates,31 covering one or several villages. Medium- and small-sized properties also existed. In the Sahel and in the province of Tripoli they even predominated. These two regions were effectively spared by the great rebellion of the jund (209-12/821-27) in the reign of Ziyādat Allāh I, and by the uprising of the serfs (280/893) in the reign of Ibrāhīm II, which is a sure sign that the overall situation was unfavorable to the great landlords, who were capable of flouting authority. One can therefore surmise that the large estates, while not wholly missing—the cases of Ibn Masrūq and of 'Alī ibn Aslam may be recalled—did not predominate. In these traditionally peaceful plains, had Ḥassān ibn al-Nu'mān, and Mūsā ibn Nuṣayr after him left the ownership of the land, with a few exceptions, in the hands of independent farmers, and had they more or less preserved the former status quo? This must be assumed. There is, in fact, no mention of any establishment of the jund.
We have some fairly precise information about the extent of the relatively medium-sized properties in the Sahel. We know, for instance, that Saḥnūn, who had experienced some hardships in his early years, to the point of being hampered in pursuing his education,32 finally found himself, by dint of sacrifices and hard work,33 owner of a property which ensured him an annual revenue of 500 dinars,34 and where 12,000 olive trees could be counted,35 many of which he had plained himself.36 The property of his friend 'Abd al-Raḥīm al-Zāḥid (d. 246-47/860-61) had 17,000 olive trees.37 The areas involved can be estimated at approximately 120 and 170 hectares. Considering the size of the great estates, these properties, which weie certainly comfortable, appeared medium-sized.38
Very small-scale holdings, barely large enough to support the peasants who farmed them, were obviously in existence too. Their importance as a percentage of all cultivated land is impossible to determine. What is certain is that they supported large numbers of needy peasants. Baqīya, the brother of the famous ascetic Buhlūl ibn Rāshid (d. 183/799), was undoubtedly one such small landholder. We encounter him disturbing a pious gathering around his brother, who refused to discuss worldly matters, while Baqīya himself persisted in talk of rainfall and agriculture.39 Rabāḥ (d. 172/788-89), an ascetic no less famous than Buhlūl, lived meagerly off the income from a tiny plot of land. He was not even able to offer a little milk to the laborers called in for the harvest.40 The holy man who, in the Sahel, had offered hospitality to the no less holy Wāṣil ibn 'Aṭā' (d. 252/866), was also undoubtedly a needy peasant living poorly.41 Thus we can be certain that the small holding was often so tiny that it hardly supported its owner. But unfortunately it has been impossible to find any precise information anywhere on the actual size of these small individual plots.
In the rural scene of Ifrīqiya during the 3rd/9th century, particularly striking is the extreme inequality which characterizes it. The huge estate, taking in one or several villages, bordered on the wretched, minute property. This inequality was also reflected in the methods of farming. We have evidence that the large- and medium-sized properties were cultivated mostly by slaves; only the cultivation of the small holdings really necessitated calling upon a free labor force. Several converging pieces of information eliminate all doubt on the subject. But before taking a closer look at this information, the political and social conditions which had committed the Ifrīqiyan economy to this path should be called to mind.
Slavery. Of all the provinces conquered by the Arabs, the Maghrib was the chosen land of slavery. Mūsā ibn Nuṣayr had dazzled his superiors in the Orient by his supply of slaves.42 Under his rod, it was the slaves who had carried out the huge works necessary for the construction of the arsenal in Tunis, modeled on the one in Tyre.43 The Maghrib seemed fated to fill local needs, as well as those of the Orient, for jawārī (slave girls) for the palaces and for slave labor for the economy. It was even mooted that the Maghrib should specialize in providing a regular and permanent supply of slaves, which, from 123/740 onward caused a series of revolts44 that were set off by the austere and egalitarian Khārijite ideology. When local sources dried up, they were compensated for in the nick of time by expeditions against the Mediterranean islands. These expeditions culminated in the conquest of Sicily, begun in 212/827, after a respite of half a century.45 This is how, for example, the monk Bernard, who was on a pilgrimage to the Holy Land with two companions, was able to observe in about 866 at Tarento, then in the hands of the Aghlabids, what he estimated to be 9,000 slaves, who had come from Benevento and were loaded on six vessels ready to sail toward Ifrīqiya and other Muslim ports.46 It should be added that buying from Christian dealers was a far from minor source of supply of slave labor.47 Ifrīqiya, until the end of the 3rd/9th century, did not want for slaves.
It is well known that these slaves were used largely for domestic chores. This custom continued, after a fashion, in the upper middle classes until quite recently, that is, up until the abolition of slavery. There is much less information about the part played by the slaves in the economic life of the country,48 both in general and with specific reference to agriculture. All the evidence we have assembled points convincingly to the conclusion that the role of slaves was of capital importance in the agricultural sector in Ifrīqiya during the 3rd/9th century. At this time, Ifrīqiya's agriculture, at least that of the larger estates, could make use of an enormous production force of cheap labor, which did not demand, apart from the cost for the upkeep of human machines, any specific distribution of wages. Moreover, the basic investment was not burdensome. After the successful campaigns in Sicily, the slaves were sold at very low prices.49 But the large landowners, who often were also men of the sword, did not even always have to invest money to acquire them. Many were able, as a result of victory, to procure slaves without opening their purses.50
Masrūq, who as we recall had held high office under Mūsā ibn Nuṣayr, had certainly not acquired his slaves with his own money. The following account reveals how his son dealt with them after his own conversion to mysticism: "'Alī ibn Muṭṭalib tells us: after his conversion he was seen to visit the villages (qarya) that his father had left him, one after the other. Their inhabitants, all those who were in residence, came out to meet him and said, 'We are your slaves ('abīduka). Everything in this village belongs to you.' He replied, If what you say is the truth, you are free, and your goods are your own property to keep.' Thus he stripped himself of everything he had inherited from his father and kept absolutely nothing."51
This text is clarity itself and leaves no doubts about the farming methods of the great estates. Other information is given in connection with one of the very first qadis of Qayrawān, Abū Khālid 'Abd al-Raḥmān ibn Ziyād ibn An'um al-Ma'āfirī al-Sufyānī (74 or 75-156 or 162/693 or 694-772 or 779).52 He was born in Ifrīqiya. It can be deduced, given his date of birth, that his father was a member of the jund, and as a result could not have been forgotten by Ḥassān ibn al-Nu'mān in the distribution of land. 'Abd al-Raḥmān followed in his father's footsteps on other fronts. We find him a prisoner in Byzantium. Freed through payment of a ransom by Abū Ja'far al-Manṣūr, he was noted at the Baghdad court for his pro-Umayyad leanings and his religious intransigence. In fact, he had in the interim acquired a solid reputation as a strict traditionalist. He then sought and was accorded the right to return to his family in Ifrīqiya, where he owned (obviously inherited from his father) a large estate (ay'a). This estate had been farmed by slaves ('abīd) under the supervision of the best qualified (akfa'u, aqwamu) among them. It is even stated—and this offers useful ethnic information—that one of the slaves was a robust, hearty, fair-haired man.53
Abū Muḥriz Muḥammad ibn 'Abdullāh ibn Qays al-Kinānī54 (d. 214/829) was also qadi of Ifrīqiya, a role in which he had succeeded Ibn Ghānim (d. ca. 190/806). He also belonged to the class of the jund—his uncle 'Abd al-'Azīz ibn Qays was commander of the guard (shurṭa) of 'Abd al-Raḥmān ibn Ḥabīb (127-37/7 4 5-55)55—and unlike his colleague and rival Asacl ibn al-Furāt, the hero of the conquest of Sicily, he was very rich. So when he was named qadi he proceeded (in order to discourage in advance the accusations of having become rich after assuming his new office) to give a kind of public accounting of what he owned. It is quoted in the Ṭabaqāt of Abū'l-'Arab:56
"When he was named qadi, he gathered all his slaves and all his flocks (kulla 'abdin lahu wa māshiya). He presented them to the people and said, 'Here is everything that I possess'—there was a crowd of slaves there ('abīdan kathīra) and all kinds of flocks—and he added, 'I have presented all my goods so that you can acquaint yourself with them on sight. If my goods, which you have here before your eyes, show any increase whatsoever, you can accuse me of corruption.' Then he had all of these goods brought to his villages (manāzil), where they had been before."
Another piece of evidence is furnished by the above-mentioned uprising, fomented in 280/893 by the great "feudal lords" against Ibrahim II. That year, writes ibn 'Idhārī, "the people of Tunis, Cap Bon, Laribus, Béja, and of the country round Qammūda rose up against him, and took as their leaders the men of the jund and of other social classes. The reason for all this was that the sultan Ibrāhim al-Aghlab had taken away their slaves and their horses, and had done them other injustices as well."57 The slaves mentioned were probably used to cultivate their masters' estates. As these masters sometimes maintained, as we have seen, well-fortified castles, and as the slaves could easily exchange the pitchfork for the sword when the occasion demanded, the amir's concern is understandable. The great agrarian landowners, thanks to the numerous slaves who worked on their farms, as well as to the horses standing in rows in their stables, represented not only an enormous economic force, but also imposed severe limitations on the amir's power, especially, no doubt, in connection with taxes. The "other injustices" referred to in the text were certainly fiscal in character.
It should be emphasized that it was nor only the great landowners who used slaves to cultivate their huge estates. The more modest farmers also had their own slaves. For example, Saḥūn, who did not shrink from personally putting his hand to the plow at times, usually employed slave labor for the cultivation of his lands. An anecdote recounted by 'Abd al-Jabbār ibn Khālid illustrates this point: "We were going to take lessons from Saḥnūn at his home in the Sahel. One day he appeared before us with a plowshare (miḥrāth) over his shoulder, and a yoke of oxen in front of him (alzawj). He announced to us: 'My slave (al-ghulām) has had a fever since yesterday. When I have finished (plowing), I will come and give you my course.' I said to him: 'I could go and do the plowing for you, and you can give your course to my friends. When I return, I will study under your guidance what I have not understood.' And so it was. On my return he gave me his meal: an oat cake and old oil."58
Had Saṣnūn only one slave for all the work on his property? This is quite possible if one reflects on his extremely humble beginnings. But it is also probable that the story, which has a certain theatrical quality, was arranged to take on an edifying character. At all events, it is an accurate reflection of the economic customs of an age when, even on modest farms, the agricultural development was traditionally bound up with the use of slave labor. The exact number of slaves is impossible to state, and necessarily varied depending on the size of the estate.
In accordance with a tradition which had its roots in pre-Islamic times, minding the flocks was the slaves' special privilege, on occasion one accorded to free men, as can be seen in the following anecdote. Ibn Ṭālib, as we have seen, was, on more than one count, a rich and powerful man who took particular care of his reputation as a karam:59
One of his friends tells how he had gone out with him after a shower. He was mounted on an Egyptian donkey. On the way, he came across some puddles. A young boy who was minding some sheep rushed forward, seized the donkey's bridle, and led him across the puddles. "Who is your master?" Ibn Ṭālib said to the young slave afterward. "So-and-so," he replied. Ibn Ṭalib stopped at a mosque and said to the young slave, "Go and fetch him." The master came to him. "How much did you pay for this young slave?" Ibn Ṭālib asked him. "Ten dinars," he replied. "Here they are," said Ibn Ṭālib. "Let him go free, and it is to you that he will henceforth be bound by the ties of employment (wa walā'uhu laka)." The sum was counted out on the spot, and the bond freeing him was drawn up. Then Ibn Ṭālib said to the former owner, "From this moment on you must pay a salary to this young boy for herding your flocks." A salary of two dinars a year was arranged. "Do not leave your master," Ibn Ṭālib said lmally to the young boy, "and do not forget us either, for we will help you."60
Ibn Ṭālib, whom we see here generously freeing the slaves of others to keep his image untarnished, naturally had his own land cultivated thanks to a plentiful supply of slave labor. We learn that to free himself from a vow, linked to the trials and tribulations of his life as qadi and his wrangles with those in power, on one occasion he had to send away his wife and sell all his slaves.61 Although our sources are silent on this point, we may assume that he must immediately have bought others so as not to let his estates go untended. The available evidence combines compellingly to indicate that in the final analysis slaves formed an integral part of the agrarian scene in Ifrīqiya during the period under investigation. The only exceptions to the rule were the indigent country folk, such as Baqīya or Rabāḥ, whom we have already met. To harvest his plot of ground, the latter called upon seasonal workers (to all appearances freemen) who boasted of the strength of their arms.
Patterns of Landholding. The agrarian scene in Ifrīqiya during the 3rd/9th century thus seems to be a mosaic made up of patterns of holdings of very different sizes: huge estates and more or less comfortable middle-class properties, cultivated by a large slave labor force, at little expense, bordering on a mass of small parcels of land, on which large numbers of humble, needy peasants busied themselves. The enormous diversity of this agrarian mosaic was also reflected in the way of life. If the humble peasants were tied to their plots on which they painfully subsisted, the large and medium farmers divided their lives between their estates and the city, where they of ten spent the betrer part of their time. We have seen that Masrūq had given his name to a whole section of Qayrawān where his residence stood. 'Abd al-Raḥmān ibn Ziyād ibn An'um also lived in the capital, at Bāb Nāfi';62 similarly, the qadi Abū Muḥriz lived in the Zuqāq ibn Dīnār.63 Manṣūr al-Ṭunbudhī, whose estates were in the Tunis area, was governor of Tripoli; Ya'qūb ibn al-Maḍā' lived at court; Saḥnūn was the most famous master in Qayrawān, where he ended his life as qadi; and Ibn Ṭālib led an opulent life, also in the capital.
How did all these great and middle-sized landowners supervise their farms from a distance? First, depending on the constraints imposed by their urban activities, by visiting their estates on a more or less regular basis. The richest, like Masrūq, had residences (rab', pl. ribā') on their estates—really villas in the Classical style, or ranches, if one can hazard a comparison with certain aspects of modern life—always ready to receive them. Some, like Saḥnūn, personally supervised the cultivation of their estates, and gave up a good deal of their time to this, at least at certain seasons of the agricultural year. But this form of administration, which necessitates frequent journeys, and long spells on the spot, is not conceivable in the case of all the great landlords. Most often landowners like Ibn Ṭālib had wakīls, stewards to whom they merely sent written instructions.64 Ibn An'um received news of his farms as far away as the Orient.65
The 3rd/9th century in Ifrīqiya was one of stability. Up until the overthrow of the Fatimids, the agrarian structures of the 3rd/9th century, which continued those established from the beginning of the 2nd/8th century, experienced no important fluctuations. Thanks to peace, which followed the troubles and great social eruptions of the previous century, these structures were even able to produce a rich yield. The outlines of the estates and the methods of farming showed a great stability, and the modifications, when thev did occur, must have been in the direction of concentrating existing arrangements, favored, as we shall see, by the law. If a certain social mobility is apparent at the level of relatively medium-sized properties (illustrated by the case of Saḥnūn, who had risen from quite humble beginnings), the great estates—apart from very exceptional cases such as Ibn Masrūq, who stripped himself of his goods and gave the proceeds to his freed slaves—cannot have undergone any appreciable modification or changed hands. This would of course not include cases of violence, ot which Ibyāna is a typical example. Everything encourages us to deduce that the social stratum of great landowners, which was preserved by heredity and intermarriage, had jealously guarded all its privileges during the whole of the 3rd/9th century. The policies of Ibrahim 11, who aimed at reducing the power of the great lords without really coming to grips with the problem,68 had only undermined the structure on which the regime rested, thus indirectly leading to its downfall and the accession of the Fatimids.
After the warlike adventures of the 2nd/8th century and the beginning of the 3rd/9th, this class of great landowners led a gilded, outrageously opulent existence both in the country and in the city. We are told that before his conversion Ibn Masrūq "deflowered a virgin every day,"67 and if Ibn Ṭālib brought his friends very pretty and expensive jawārī,68 one can imagine that he did not deprive himself either. Aḥmad ibn 'Alī ibn Ḥumayd69 (d. between 251 and 261/865-75) kept open house at Qayrawān, had his furniture (including two tables of pure crystal) brought from the Orient, and was a generous benefactor whose gifts in kind demanded the mobilization of more than twenty mules (naturally, coming from his own lands) to bear loads of wheat, oil, and honey. His table constantly overflowed with the most costly dishes. Is it surprising that Fatimid propaganda, which promised an equal share for all in the Sun of God, symbolized by the Mahdi, should have found ready cars, despite its heretical character?
These agrarian and social structures, marked by the great inequalities that we have mentioned above, nonetheless functioned properly all through the 3rd/9th century. There were certainly some uprisings, but these were relatively small in number and impurtance, except for the great rebellion ot the jund m the reign of Ziyādat Allāh I. No revolt of the peasants is mentioned; the social order, as it had evolved, was accepted. Individual injustices were denounced, there was a demand for qadis of integrity, but the social pyramid was not called into question and the overall division of wealth remained unchallenged. We shall see that respect for the law, which was presented as divine, had much to do with this acceptance of the established order. It was not disputed until the end of the century, when in the mountains of Kabylie unrest was fomented by Shī'ī missionary work.
Agricultural Production. Agricultural production had been almost constantly maintained at a highly satisfactory level. It was supported by a sizable irrigation system which, despite the total silence of our sources, must have been maintained by a plentiful and inexpensive supply of slave labor.70 During the entire century only three shortages of food are recorded,71 all of them clustered between the years 260 and 268/873 and 881. Only one took a really serious turn, and this is attested to by only a single source—the Nihāya of Nuwayrī. This was because an Ifrīqiyan fleet carrying a cargo of grain had been captured by the Byzantine navy.72 In general, the only result of the few famines of the 3rd/9th century was an increase in the cost of wheat. The cost of transportation, and speculation, which was not curbed by law, easily explain the sudden rise in prices. It was customary in such situations for holy men like Buhliil to sell their supplies and share the misfortune of all in their anxiety to be charitable.73 The soil of Ifrīqiya, as it was divided up and farmed in the 3rd/9th century, fed its population easily. Its production was even in surplus, as a general rule, and if there were difficult situations from time to time as a result of abnormal fluctuations in the weather, it cannot be said that people died of hunger.
Natural and Political Disasters. At the end of the next century, famine, the real famine which kills, decimated the population. Let us read the description of an eye-witness, Raqīq (d. after 418/1027-28):
In 395 (1004-5), a great calamity tell on Ifrīqiya. Human misery showed its face in broad daylight; the poor perished; the rich were ruined; prices went up and foodstuffs disappeared. The country people deserted their regions; most of the houses were abandoned and there was not even anyone left to inherit them. To all tfie.se calamities were added epidemics (wabā') and the plague (ṭa'ūn), which carried off the greater part ot the population, both rich and poor. One hardly saw people busy except caring for or visiting the sick, performing the last rites for a dead person, following a funeral procession, or returning from a burial. (In Qayrawān) the bodies ol the destitute (al-ḍu'afā) were gathered at Bāb Sālim. Communal graves were dug for them, and in each grave a hundred corpses or more were buried. The number of dead in the ranks of the upper classes (ṭabaqāt ai-nās), among the scholars, the merchants, the women and children,, was so high that only their Creator—the Most High!—could number them. The mosques in Qayrawān were deserted, the public ovens and baths were still. People were even reduced to burning the doors of their houses and the rafters from their roots. Many city dwellers and country folk emigrated to Sicily, and people went so far in these circumstances as to: pay 2 dirhams for a pomegranate tor a sick man, and 30 dirhams for a chicken. It is even said that in the country people ate each other. Such is the account given by Abū Isḍāq al-Raqīq.74
It would not seem that this picture has been much exaggerated. From 395/1004 onward, one disaster followed another.75 The years 409/1018-19, 413/1022-23, 425/1033-34, 432/1040-42, 447/1055-56, 469/1076-77, etc., can all be edged in black. In the fatwās of the faqīhs, there is talk only of importing grain. H. R. Idris emphasizes that "it is difficult to admit that all these 'responsa' only involve bad years."76 Certainly Rome's famous granary could no longer even feed its own population, although the population was necessarily reduced after the dreadful famine of 395/1004-5. To remedy the situation, reserves of gold, already depleted after the demands made upon them when Mu'izz left for Cairo, were exhausted.
The reign of Bādīs (386-406/999-1016) ended badly. We fully share Jean Poncet's opinion that the high point of the dynasty should be situated before rhe year 395/1004,77 which suddenly revealed a situation that for numerous reasons worsened progressively. Soon after the ruinous campaigns in the central and furthest Maghrib, civil war broke out. The death of Bādīs, in the middle of a battle against his uncle Ḥmmād, definitively accomplished the division of the kingdom. The political collapse which this division completed had been preceded by economic collapse, as revealed in the fearful famine of 395/1004-5. S. D. Goitein notes that from this time on the trade of Ifrīqiya had begun to decline.78 What is more, the merchants began to leave the country with their wives and children to establish themselves in Egypt.
The crowning event was the Hilalian invasion (448/1052), which neither those in power nor the enfeebled country could strangle, overpower, or channel. Without going so far as to describe it as a historic catastrophe, it can be said that the Hilalians had delivered a severe blow to an economy that was already seriously ailing. They precipitated the transformation of an economy, traditionally agricultural and nrban but undoubtedly in a state of crisis, into a largely pastoral economy, with all the political and economic consequences that such a transformation implies. For the peaceful life of the city dweller, they substituted the adventurous existence of nomads, whose warlike tendencies weighed heavily on the fate of the country for centuries. When the Sulaym were invited to settle in the country at the beginning of the 6th/13th century, in the hopes that their presence would improve the situation, they only served to aggravate the already serious problems. When they fought one another, it was to the detriment of the ailing country. This was the real Hilalian catastrophe. This was the ill which paralyzed the country and made its economic, social, and political recovery impossible. Contrary to the opinion of Jean Poncet,79 who cannot have direct access to the literature of that period, people of that era were fully aware that they were facing a real and unprecedented disaster in the country's history. To convince oneself of the depths of their distress, and of the misfortunes which scattered them on the roads in total destitution (when they were able to save their lives), it is only necessary to read the poignant lines of one Ibn Sharaf (d. 460/1068).80
Time did not heal the wounds. Anarchy became a permanent feature of the country, and it is only with historical perspective that Ibn Khaldūn, who cannot be accused of subjectivity—he knew the nomads well, since he had taken refuge with them on certain occasions during his stormy life and even had a deep appreciation for many of their rough-hewn qualities—measured the full breadth of the tragedy. The effects of the "catastrophe" grew, in fact, as time went on. Insecurity became the order of the day, making remedial efforts useless and progressively sapping people's energy. Many of the peaceful peasants, infected by the disease, had to change their way of life. In short, a new kind of thinking sur faced, one strongly influenced by the nomadic way of life. How can this transformation, which profoundly modified the social structures of a very urbanized country, not be described as a catastrophe?
This process, which had developed with an implacable logic, like an irreversible mechanism, had distant origins. In it should be sought one of the causes of the decadence of lfrīqivan civilization, a decadence that resulted in present-day underdevelopment. Behind this evolution, the origins of the collapse of the agricultural and social structures inherited from the 3rd/9th century should be examined. These structures had not been replaced soon enough by others better adapted to the new circumstances, ones able to effect efficiently a transition from the old to the new. Therein lies a line of research that deserves to be more thoroughly explored; here it can only be sketched out in outline.
The middle of the 4th/10th century saw an important modification in the balance of power in the Mediterranean. In 350/961 Nicephorus Phocas seized Crete and sounded the death knell for the supremacy of Arab fleets in the Mediterranean.81 There were no more conquests. From then on, Ifrīqiya could no longer procure slaves in large enough quantities and at a low price. There was no comparison with earlier times, when they were captured on the spot by the tens of thousands. Penury followed plenty.82 One could still buy slaves from Italian merchants; the energetic measures promulgated by the doges and Pope Gregory VII (1073-85) to curb this trade prove that it still persisted.83 But, like the trade in strategic materials, it was an illicit trade, condemned by the political and religious authorities,84 and therefore somewhat risky and wearing. It could not take the place of the abundant resources of the 2nd/ 8th and the 3rd/9th centuries and ensure a regular and economically competitive supply of slaves. It was, at best, a palliative.
Under these circumstances, it appears that the agrarian structures inherited from the past were seriously affected by the shortage of slave labor, especially from the second half of the 4th/10th century onward. The Fatimids, who did not live up to the revolutionary promises of their propaganda, had not overthrown everything. But the political debility that marked the end of their rule in Ifrīqiya progressively deprived the soil of Ifrīqiya of "serfs." It was the large-and medium-sized farms which particularly felt the effects of the new situation. The upkeep of the irrigation system, which compensated for the lack of rain and alleviated the effects of periodic droughts, must also have been neglected or abandoned. A change in the methods of farming was necessary to adapt to the new situation, but it did not come soon enough or on a large enough scale to maintain the volume of yields and avoid the collapse of agricultural production. This resulted in a steady drain on gold reserves, and in frequent severe famines as well.85
Why this decline? Were there not enough free men available on the job market and capable of replacing the declining supply of slave labor by putting their energy to good use in the agricultural sector? In other words, was the demographic situation unfavorable, on account of the sapping of manpower by wars which had succeeded one another since the accession of the Fatimids?86 Was the old capital in slaves depleted, and could it no longer be renewed in sufficient quantity? Did the large- and medium-sized estates lie partially or totally fallow because of this? Was this before the Hilalians came and settled, and transformed so many plots of land into pasturage? We can offer no certain replies based on the broad documentary evidence. But we are inclined to think that the process that had transformed Rome's famous granary, with a reduced population, to a country of endemic famines can be explained to a great extent, by the drying up of the sources which had provided Ifrīqiya with slaves for a long time.
It is certain that this agricultural landscape, formed by the socio-political situation of the 2nd/8th century and the 3rd/9th century had not changed soon enough. The capital amassed by the landed upper classes enabled them to survive despite the diminution of the revenues from their land, which resulted from the increasingly serious poverty of their slave laborers. In general, the reserves of gold that the country still had at its disposal must have worked against agriculture in the end, precisely by blocking the remodification of the ancient patterns of cultivation. On no level of society did anyone clearly feel a pressing need to maintain the old levels of production at all costs. Certain brakes within the social structure must have been against it. The great farmers, who were the most affected by the crisis, must have sought compensation by moving their capital into other fields and speculating on the market. To the socioeconomic changes demanded by the new situation, they preferred the easy solutions offered by the relative abundance of gold, such as the palliative of imports, which moreover had the advantage of ensuring high profits to speculators. H. R. Idris notes that in the fatwās, "the only talk was of contracts of qirāḍ in cash for buying grain."87
The fact is that the warning bells sounded by famines, although they decimated the population, did not provoke a healthy shock to the system. There is no indication that the deep origins of the malaise were clearly understood. The real solutions, those capable of ensuring the full cultivation of the land following new formulas, were perhaps seen, because of deep-seated inhibitions, as too adventurous or impossible to put into practice. Probably all these factors, and others as well, worked simultaneously and in the same general direction.
Apparent Prosperity. Paradoxically, this fundamentally unhealthy economic situation was masked by a false affluence, undoubtedly due to the huge profits it brought to some speculators on the black market of the period. Al-Mu'izz ibn Bādīs wasted a lot of money, and the waste, praised by blind and self-interested acolytes, was seen as a new zenith.88 Finally, Ḥaydarān (443/1052) suddenly revealed and simultaneously precipitated total collapse. And the Hilalians, by spreading a pastoral form of farming that was extensive and did not require much labor, rendered the changes that had not taken place earlier quite impossible to achieve.
This gloomy picture of the Ifrīqiyan economy before and after the Hilalian invasion contrasts strongly with the ill-distributed but genuine prosperity that the country experienced in the 3rd/9th century. Seen from later centuries, the Aghlabid Ifrīqiya of Abū al-Gharānīq (250-61/864-75) took on the colors of a lost Eldorado, and its prosperity became proverbial. "Today," wrote the Spaniard Ibn al-Khaṭīb in the 8th/14th century, "people at home say, when they quote a proverb about a peaceful reign, and when they want to describe a state as just and prosperous: It is the reign of Abū'l-Gharānīq."89 Apart from some rare and transient climatic accidents, the agrarian structure had not only allowed the fulfillment of local needs, but had also ensured the production of exportable surpluses which were sources not of inflation and rising prices but of real wealth.
Ifrīqiya at that time had a fruitful trade with all its neighbors: Christians, Muslims, and sub-Saharan blacks. Neither barriers of religion nor war had halted trade with Sicily and Italy, which had been traditional since ancient times. As in the past, Ifrīqiya continued to increase its wealth by sending its surplus oil to these areas,90 and the merchants were able to travel more or less freely in both directions, despite the hostilities. Ifrīqiyan oil also went to the Maghrib and Egypt.91 In a good year Ifrīqiya also exported its surplus grain, as in ancient times. The cargoes of several ships carrying wheat, which were unloaded by Ibn al-Ḥayyāq in Alexandria toward the end of the eighth or the beginning of the ninth century, came from Ifrīqiya.92 Ifrīqiya also exported the products of its industry and its mines.93 The caravans that crossed the Sahara probably did not carry food products; they set off, rather, with manufactured items, especially small trading goods, and brought back ivory, gold, and, of course, slaves who found employment by the tens of thousands in the armies of the amirs,94 and who seem also to have been particularly appreciated in the oasis of Jerid.95 The Qayrawānian Saqan al-Ṣā'igh96 manufactured chains of gold-plated copper which were sent to the country of the Blacks (Bilād al-Sūdān), and 'Alī ibn Ḥumayd,97 the vizier of Ziyādat Allāh I (201-23/817-38), had built up a very confortable fortune acquired largely in the ivory trade.98
Law, to which we now turn, had never divorced itself from the economic life of the country, at least during our period and in the years that immediately followed.
Mālikī Sunnism. The first concern of the lawyers was to set a limit to all the disputes on the rights of landlords to their properties. The property law in Ifrīqiya, precisely because of its unique legal basis in the eyes of Muslim law through the act of fatḥ, that is the "opening" of the country to Islam, ran the risk of provoking interminable disputes. We have seen that Saḥnūn admits openly that nothing certain is known about the theoretically legal status of the properties in Ifrīqiya. In these circumstances, a decision to examine the title of every single person to their property, and to keep the book of complaints and claims open, under the pretext of review and of justice, would have led to an endless process of quibbling and corruption and wrecked the country's economy in the name of justice and due process. Wisely, Ifrīqiyan Mālikī law gave its support to maintaining the status quo ante, that is, to economic and social stability. "In the final analysis and after careful study," Saḥnūn pronounces, "we think that it is right to confirm the status quo hallowed by centuries of occupation. In particular, unless it is a question of multiple, contradictory accounts, or of estates taken over by force, or whose owners have been subject to measures allowing arbitrary expulsion, one must confirm the occupants in the ownership of their land, in accordance with the well-known system of endowments after deduction of one-fifth, which we know, from the repercussions that have come down to us across the centuries, to have been effectively applied. In fact, there does not exist in Ifrīqiya an estate whose neighbors cannot agree on the explanation of its founding and origins. And it is actually very rare that this kind of thing escapes the notice of the occupants and their neighbors."99
This text shows clearly that the estates alluded to were those which had been awarded to the conquerors by virtue of the act of conquest. It undoubtedly concerns the great estates, above all, which were objects of envy for those in power, that is, the amirs, as is evident from the appropriation of Ibyāna, mentioned above, by Ibrāhīm II in 275/888-89. Other important estates must have changed hands prior to this date (Saḥnūn, it should be remembered, died in 240/854) by the same methods of violence, that is by the use of extortion and expulsion, to which reference is made in the text, and for which methods alone a procedure of review and eventual re-establishment of the former owners in their lost rights is allowed. By the mouth of the most eminent jurist of the age, the law had in short declared itself on the side of order and social and economic equilibrium by confirming the accomplished fact. The social and political structures inherited from the fatḥ, from the conquest, were legalized and perpetuated. The law was thus brought, by virtue of the sacred principle of respect for private property, to align itself in fact with the large landowners, to whom it offered the guarantees of law against all forms of dispute, expropriation, and pillage. It could not be otherwise in the context of the period, for any other decision would inevitably have led to anarchy. The law opted for order and social tranquillity. Sunnism was not Khārijism; it was not in its nature to be revolutionary.
One should, however, be explicit and avoid any misunderstanding. If this option, through fear of adventurers, in fact led to a kind of sanctification of private property, whatever its dimensions and origins might be, it did not on this account put the law at the service of those in power and the lawyers in their pay. Many indeed were the qadis, such as Asad ibn al-Furāt, Saḥnūn, and 'Īsā ibn Miskīn,100 who defended their independence against all forms of pressure, including those that caine from the court, and brought all their weight to bear, with integrity, courage, and obstinacy, on the side of the weak and oppressed.
The law, in short, sought to be equal for all, even when it upheld situations marked by profound inequalities. Its provisions with regard to agrarian structures, born of half a century of war prolonged by a series of uprisings, violence, and different upheavals, are a mine of information both on these structures and on the organization of agricultural life in general.
The Ḥanafī School. It should first be made clear that we only have a single voice at our disposal, that of Mālikī Sunnism. The 3rd/9tb century in Ifrīqiya was marked by a great burgeoning of legal documents. We should remember in passing that this activity was not under state control, since in principle the state did not legislate. In these conditions, the first school to dominate the legal scene was that of the Ḥanafī discipline. The jurists who professed Ḥanafism were members of the majority until the middle of the 3rd/9th century at least,101 and the qadis, who always applied the solutions of their own school, were at first essentially chosen from among their ranks. It is therefore regrettable that we possess no ḤHanafī legal text dating back to the 3rd/9th century in Ifrīqiya, not even the Asadīya,102 which did however represent a Ḥanafī-Mālikī compromise and of which only a few pages still survive. Aside from the legal activities of the Ḥanafīs, which must have found expression through numerous provisions relating to agricultural law, the legal activities of the Mu'tazilites103 should also be noted. If they did not have a large hearing in the crowd, they had the court on their side, and doubtless' many elite groups as well. As they had supplied numerous qadis to the Aghlabids, they must have exercised a considerable influence on everything having to do with justice, including the agricultural sector, which is our primary interest here. Finally, the Ibāḍīs should not be overlooked. It is true that they were not represented by a single qadi in Ifrīqiya in the 3rd/9th century, and consequently never had the opportunity to put their ideas into practice, at least officially. But they had nevertheless been active,104 and they had been able to expound their theories freely for a long time, even in the Great Mosque of Qayrawān, before Saḥnūn, who was named qadi in 234/849, broke up their circle.105 They must have influenced all forms of activity, at least in the heart of their own communities, which were capable of organization in order to avoid practically all official jurisdiction.106 This means that agricultural activities must have had to bend in turn, or sometimes even simultaneously, to legal provisions,107 which were rather different, and varied in their degree of flexibility or severity, and which authorized or forbade certain forms of transaction or cultivation depending on the beliefs of the school that was in power.
The Mudawwana. Of all these provisions, we possess only those codified in the Mudawwana. A unique document of its kind, the Mudawwana was also up until then an insufficiently exploited document. The text is obviously not accommodating; it belongs to that category of documents that historians have still not learned to manipulate properly, and toward which, because of the reputation they have gained for being theoretical and unrelated to reality, they retain a large measure of suspicion. This suspicion, particularly in the case of the Mudawwana,108 is absolutelv unjustified. It is closer in fact to a code of jurisprudence than a legal encyclopedia. It is a huge compendium of aqwāl, that is, replies or solutions, to real problems that cropped up in concrete form in daily life. Suḥnūn, and all his disciples after him, had drawn on the Mudawwana to resolve suits that had nothing imaginary about them. The historian would therefore be quite wrong to despise them. Our only regret is not to have other similar anthologies at our disposal reflecting the leanings of other legal schools.
We shall examine in turn the treatment in the Mudawwana of the shufa, the qisma, the 'arāya, and the minḥa, and other questions connected with farming the land.
The shufa is the right of pre-emption, exercised in principle only on real estate.109 "There can be no shufa except when it is a question of the sale of houses, lands, palm-trees, or other categories of tree."110 Certain things are specifically excluded from the shufa, such as flocks, ships, fabrics and all consumer goods in general. Under certain circumstances, fruit not yet gathered can nevertheless be the object of a shufa,111 but never harvests of grain, or mills,112 even those installed on water courses.
The shufa is also exercised on goods that are held in common and are indivisible.113 The process is set in motion when one or several parts of the property are put up for sale. This situation can occur in two cases: when an associate wants to leave an association, or when a legatee wants to realize his inheritance. The shufa then works at favor of the partner or partners remaining, who can exercise their right of pre-emption and have priority in buying the part of the property that is the object of transfer. This prerogative is exercised by all those who have rights, including the dhimmīs.114 But all the shafī do not enjoy the same degree of priority. In the case of inheritance, in particular, it is the legatees who are the closest relatives to the deceased who have first call,115 and that does not always happen without debate, as situations can be very complicated.
Questions of price arise too and often occasional multiple disputes, since fraudulent transactions are not uncommon. As a general rule, it is the statements of the interested third party which are held to be reliable,116 and the shafī, if he wishes to exercise his right, must buy at the price agreed to by the latter. But one must also take the statements of the seller into account, and eventually have recourse to investigations and assessments.117 It is easy to imagine the interminable proceedings that all this can produce.
The shufa must also be exercised to include the whole of the part of the property which is up for sale. The shafī' could not claim his right over a part of that property detached from the whole. Saḥnūn asks Ibn al-Qāsim the following question on this point:
Let us look at the case of three men who own (shares): the first in a palm grove and lands, the second in a village, and the third in houses. The three sell the entirety of their shares, as a unit, to a third party. The shafī' for the village, the palm grove, and the houses in question is the same man. He proposes to take (the share) constituted by the palm grove, which is up for sale, at the price suggested, excluding the village and the houses. The third party buyer calls upon him to buy it all, or to abandon it all.118
What should be done in this case? Ibn al-Qasim does not have a ready answer. Such a situation, at the period when he had absorbed the teaching of Mālik, had probably never come up in Egypt or Medina. He therefore reasons by analogy, to come to the conclusion that the shafī' must take everything or leave everything.
It is obvious from what precedes that the shufa, to which long exegeses are devoted in the Mudawwana—a sure sign of the important part that it played at that period—worked against the dismemberment of property. It even favored its concentration in the same hands. In doing this, it operated in favor of the maintenance and reinforcement of the great estate, which often exceeded the dimensions of a village, which confirms what we had noted elsewhere. The shufa heats witness also to the existence of common and indivisible family estates, farmed no doubt under the guidance of ihe eldest in the family, or of the most eminent personage of the gens. Thanks to the possibilities offered by the shufa, the family inheritance was defended against the intrusion of strangers, the share of those selling out being bought as a first option by those who had priority, and were the most qualified among the close relatives. In the case of lands that were not inherited, but rather acquired and farmed in common, it offered the holders of the parts, always by the mechanism of priority rights in buying, the chance of enlarging their assets as departures cut down on the number of associates. This process, which at the term of its evolution ended by concentrating the property in the hands of one individual, could not but favor the constitution of great estates, even where they had not existed previously.
The possibilities offered by the shufa were all the more real since the conditions governing its application clearly favored the shafī'. The latter could, in fact, exercise his right as soon as an agreement of sale was made, regardless of whether or not this sale was carried out.119 And above all, he had the option of retaining this right and of using it a year, or even longer, after the execution of the sale.120 The buyer of one part of an estate that was indivisible and held in common was thus never certain of a deal, particularly if, at the moment of closing the sale, a possible shafī' was absent. In this case, the latter could always, in fact, use his absence as an argument to call the sale into question to his benefit, however much time had elapsed.121 He could equally use his ignorance of the sale as an argument with the same success. In practical terms, the delay allowed before the expiration of pre-emptive rights was so long, so imprecise, and so subject to dispute that it always left the door open for the return of a parcel of land that had fallen into the hands of strangers to the bosom of the family circle unless one could prove that the shafī' had made a tacit renunciation of his right of shuf'a.122
It also happened that the part that was sold underwent more or less important changes in the hands of its new owner. If the latter, for example, had sown, he had the right to harvest without paying rent to the shafī, who for his part had retained the option of bringing the shufa into play. He retained this option even if the new owner had planted the land in question with palm trees or other fruit trees, but was obliged, in addition to the reimbursement of the purchase price, to pay damages to the dispossessed buyer. In short, it is clear that the legislator had wished, depending on the case, to guarantee the integrity of the family inheritance, or when it was a question of common enterprise having an association as its basis, the interests of the founding partners.
But the mechanism of the shufa, because of the real threat it posed to the validity of transactions, could be a very serious source of instability and could dangerously affect economic activity. So the legislator also provided a proper remedy to this drawback. Mālikī law, as it is expressed in the Mudawwana, affirms the fact quite forcefully: "No shufa for anything which has been subject to a previous division."123 It is enough, therefore, to arrange in advance for the division of property originally held as common and indivisible, to escape the shufa and to be able eventually to dispose of it in full security and without risk of dispute.
The qisma, the sharing of property, habitually poses numerous thorny questions that it would be tedious to consider in their entirety here. We should simply indicate that Saḥnūn dedicates two voluminous chapters of his Mudawwana to it124 and content ourselves with some examples that confirm the features of the agricultural scene in Ifrīqiya that we have already identified.
Question: Consider the case of two brothers who inherit a village including houses, trees, and bare terrain (arḍ bayḍā'). The two brothers want to divide it between them. How should it be done?
Answer: The houses must be divided following the procedure already indicated elsewhere in such cases. The same applies to the land; one must proceed in the way I have already indicated to you on this matter. Question: Can you recall to me the procedure in connection with the bare terrain?
Answer: You survey the lands and estimate their quality, their market value, and their proximity to each other. Afterwards you divide them, taking all these criteria into account, and give each one his share, grouped in the same area. If, on the other hand, the land in question presents differences that are too marked, each of the two legatees will receive his share of each part separately. The same thing happens when houses or palm trees are involved.
Question: How can the criterion of proximity of one parcel of land in relation to another be estimated?
Answer: Mālik has not given us any definite guideline on this subject. Ibn al-Qāsim added subsequently: I think nevertheless that one can consider a mile, or something equivalent to it, as a close distance when orchards (al-ḥawā'it) or land is involved.
Question: And the trees belonging to a village inherited by the two brothers: how would Mālik divide them between them? First, in the case where the trees are of different kinds—apple trees, pomegranate trees, peach trees, citron trees and other fruit trees—and are mixed in the same plantation; then when they are grouped by species in separate orchards?
Answer: On this specific point, I heard nothing from Mālik's lips. But I myself think that if the trees are mixed, as you say, in a single plantation (ḥā'it), the division should be accomplished by evaluation, and by giving to each one a distinct part in the same place. If, on the other hand, several separate orchards (ajinna) are involved, with apple trees set apart, and pomegranates and other fruit trees all grouped together separately, and it each orchard can be divided between the inheritors, then the division should be done by valuation, and each one is given, in each orchard, the part which is his.125
This passage is not the only one where the thorny problem raised by the division of the huge estates, as large as a village and larger, is brought up. Saḥnūn returns to the problem elsewhere.126 He also considers the division of stock and slaves,127 a division that raises less problems than those of the lands to which the animal and human stock is bound.
Can one oppose the division? No. Even if the individual, among several legatees or associates, wishes it; even if the land is too small and its division renders it impossible to cultivate.
Question: When a small plot of land is involved, belonging in common to several persons, and if the part which is due to each of these people, once the plot is divided, is so small that it is of no use, should one, in Mālik's opinion, proceed to the division or not?
Answer: Mālik said: one must proceed to the division, even if some peopie are opposed to it. As soon as a division is asked for, even if the demand is formulated by only one person, one must proceed with it.128
The division of land—whether it concerns huge estates or little plots, which stood side by side in Ifrīqiya in the 3rd/9th century—often raised complex legal and technical problems. And these divisions should be, in the case of litigation, says Saḥhnūn, in the exclusive jurisdiction of the qadi, outside the authority of the prefect of police (Ṣāḥib al-Shurṭa),129 unless the latter is acting under the orders of the qadi. One should also employ experts ( qussdām),130 whose compensation is most often covered in practice by the beneficiaries.
Small and large landowners, for differing reasons, often found themselves in situations that did not allow them to cultivate their properties personally. Then they employed musāqāt, sharecropping. In theory, the contract of musāqāt cannot be agreed upon, if the etymological sense of the expression is respected, unless there is irrigation. In fact, this kind of contract can be extended to dry cultivation—such as olive trees and cereals—which need the care that justifies it.131 This contract of cultivation generally leaves the tenant (al-'ārnil) with half the crop. But this proposition can vary depending on the wish of the parties.132 Basically, everything depends on the assets which the owner of the land and the tenant respectively bring to the agreement. The latter can bring just his labor, but can also provide implements and stock, especially animals and slaves. At the time when the agreement is signed, all these points must be covered in detail.
Question: A tenant signs a sharccropping contract (musāqāt) for palms or other trees. Are all the labor costs the tenant's responsibility, in Mālik's doctrine?
Answer: Yes, unless there are already animals and slaves in the orchard concerned who are already engaged in its cultivation. In this case, there is no objection to their continuing to do so.
Question: If on the one hand the tenant (al-musāqī) makes their maintenance of the orchard a condition of the contract, and if on the other the owner of the land wants to withdraw them, has the latter, in Mālik's opinion, the right to effect this withdrawal?
Answer: Mālik says on this point: When the contract is drawn up and the terms are fixed, the owner of the land does not have the right to withdraw them. Nor can he stipulate, "I give the orchard on lease (musāqātan) on the condition of withdrawing my slaves and animals." But if he withdraws them beforehand and subsequently offers the orchard for rent, there is no objection to his acting in this way.133
This passage, and others too,134 confirms the use on a large scale of slave labor in Ifrīqiyan farming. This raises numerous legal problems. Should the owner or the tenant, depending on the case, replace slaves who have died during the term of the contract?135 Who is responsible for the upkeep of these slaves?136 Can they be supported on the farm given for rent? In general, it is the responsibility of the tenant to meet the needs of animals and slaves used in farming whether they are provided by himself or the landowner. And it can even happen that a slave (al-'abd al-ma'dhūn lahu fī'l-tijāra), taking the role in some way of supervisor, or of deputy, acting on the authority of his master, himself signs sharecropping contracts.137 This method of farming must have suited all those who could not personally attend to their lands, especially rich merchants who had invested a part of their profits in real estate, and many of the great landowners whose occupations in town demanded much of their time.
The 'arāya,138 and the minḥa, which the Mudawwana discusses at some length, must also correspond to certain kinds of farming, although this is never clearly stated. The 'arīya or 'āriya (pi. 'arāyā) is certainly presented as a gift (ḥiba). In theory, it consists in offering the produce of a certain number of orchard trees to a third party, before the fruit is ripe and the harvest gathered in. The owner nevertheless retains the option of buying back the subject of the gift from the beneficiary, by offering him, after valuation, the equal value of this gift in kind or in cash. The minḥa139 is a kind of 'arīya that is specially concerned with livestock: camels, sheep, and cattle. It involves offering a third party, for a limited period, the interest in a certain number of animals, an interest which can be bought back before the expiration of the contract. The 'arīya and the minḥa no doubt covered certain agricultural practices. They must have constituted a kind of payment, in kind and at an appointed time, for services rendered, with the landowner always retaining the option of recovering his goods if there was an upturn in his liquidity or his reserves.
Agricultural work also naturally demanded the use of hiring or renting land and hiring animals or services. The land was divided into two categories: what was irrigated and what was not. In general, the rent of irrigated land was clue immediately upon the signing of the contract,140 since no climatic risks were involved. The rent on land given over to dry farming, on the other hand, was only payable once the capacity of labor and seeds were insured.141 For these lands, the renting, moreover, could only be firmly agreed on for one year, and in case of natural disaster—drought, flooding, locusts, devastation due to the passing of armies, plant diseases, etc.—total or partial refund of the rent was guaranteed to the tenant, depending on the scale of the damage.142 The sensitivity of Muslim law to unforeseen circumstances and risks is noticeable in all these provisions. It should also be said that a plot of land rented out for one kind of cultivation could not be used for another kind, which would cause greater impoverishment of the soil.143
It was also customary to hire animals: a cow, for example, to carry out the plowing or to benefit from her milk,144 or a stud to cover females.145 All these details lead us to the heart of life as it was led by small landowners, who were needy, without capital, and did not have the means of production or the means of developing their little plots of land.
Many services were also hired from freemen or slaves. These services were often paid in kind. For example, the harvest was insured by a share of the yield, sometimes even half.146 These practices are still current today. The services of a shepherd were hired for a year147 in return for payment in cash or kind.148 Many slaves were also hired to cover the needs of agricultural life,149 and it sometimes happened that they were given once, for life,150 just like a house. A slave could he hired for carrying a harvest151 or for other jobs.152 These slaves often worked hard during the day and spent part of the night grinding the grain.153 They seem to have been especially prized as shepherds,154 which confirms what we have seen elsewhere. They were hired from their masters and entrusted with herding of the flocks. And to make this hiring less burdensome, they could be authorized to gather together the animals belonging to several landowners. Naturally, small landowners were involved. In the case of negligence, it was the master of the hired slave—had the hiring been done without his knowledge—who was responsible for loss and had to pay compensation. Apart from cases of real negligence, properly ascertained, the responsibility of the shepherd was never binding. In particular, he was never responsible for losses due to theft or to any other accident. In this regard, and unless there was proof to the contrary, the shepherd's statements were believed. All these methods of farming—which naturally did not concern either the large- or medium-sized farmers, who had far more extensive possibilities for developing their estates—offer us insights concerning this class of small farmers, peasants without land or stock, making a meager living by using all kinds of expedients, which increased their dependence on those who possessed, apart from the land, the means of production. More specifically, they owned animal and human resources offered for hire on occasion.
The agricultural landscape in Ifrīqiya during the 3rd/9th century, as revealed to us essentially through the complementary accounts of biographical and legal sources, was undoubtedly marked by very profound and flagrant inequalities. It was the landscape of a society which was ethnically and economically very heterogeneous. Did these inequalities have their origins in the policies followed from the Muslim conquest onward? Or had these policies merely prolonged existing structures by replacing, on the estates which had fallen vacant because of war, the former farmers by others chosen from among the victors, who had to be rewarded for services rendered on the battlefield? In other words, did the large estates of the 3rd/9th century, those of the amirs and big landowners, merely mask or prolong the ancient imperial estate, or the former latifundia? This question cannot be answered with any degree of certainty until the preceding period of the Muslim conquest, which is the most obscure period of Ifrīqiyan history, has been more thoroughly examined and illuminated.155
Thanks to the abundance of slave labor, the agrarian structures of the 3rd/9th century in Ifrīqiya had functioned perfectly and ensured the production of surplus goods for export, especially as far as oil was concerned. Prosperity was real. It was certainly not equally so for everyone. But if inbreeding, favored by the law, seems to have been the rule as far as the great estates were concerned, Ifrīqiyan society of that period was still not totally restrictive. It often happened that people of most modest origins were able to acquire very comfortable estates, and it is certain that social mobility was even more noticeable and more widespread in the third sector, which we have barely touched on. Wealth, if it was ill-distributed, was no longer the exclusive privilege of the descendants of the illustrious veterans.
Why did this prosperity finally disintegrate? Why was the end of the next century marked by a catastrophic increase in the amount of land lying fallow and the beginning of an era of famines? There are many reasons. But we believe that one must look for them particularly in the changes that occurred in the balance of power in the Mediterranean from the middle of the 4th/10th century onward. This balance was upset in favor of Byzantium, and of Christianity in general. This rupture, in making slaves a much rarer commodity, must have had serious repercussions on the Ifrīqiyan economy, and especially on the agricultural sector, which was not able to change in time and adapt itself to the new situation. Why did the change not take place? Perhaps because the heavy tax burden imposed by the Fatimids, or more particularly by the Zirids156—paying tribute to Cairo and involvement in ruinous expeditions in central and furthest Maghrib—made this change impractical. The huge number of salaries that had now to be included in the cost of farming, when added to the taxes, probably ruined the profit margin on land, at least the mediocre areas. In certain cases it had certainly become more profitable not to cultivate it. The state could have reacted to the situation by introducing lower taxes, and by extending other incentives to agriculture. But it seems that the state showed total ineptitude. Perhaps too, the palliatives of imports, facilitated by the relative abundance of gold, had masked the real problems and the proper solutions all the more easily, since speculation was thriving on it on every level. The demographic situation was probably not favorable either, and became increasingly unfavorable after the crises of 395/1004-5. The third sector, traditionally overdeveloped, absorbed a labor force that could otherwise have filled the manpower requirements of agriculture and taken up where the disappearing slaves left off. Briefly, several factors worked in the same direction. The decline of agriculture, which had no doubt begun from the middle of the 4th/10th century onward, reached undeniably catastrophic proportions after the great famine of 395/ 1004-5. Finally, the Hilalians gave it the death blow. They found mam plots of land lying fallow, and thus had no difficulty in accentuating the movement, and, above all, in giving it an irreversible character and in committing the country to a largely pastoral trend that showed itself to be harmful to the country's evolution on every level. And thus it is that agriculture, which in the 3rd/9th century was a source of power and immense wealth, became the sole resource of the weak in the time of Ibn Khaldūn. The latter entitled the eighth section of Chapter 5 of his Muqaddima: "On agriculture, the means of subsistence for humble folk and peaceful country people."157 What a road had been traveled since the time when Manṣūr, lord of Ṭunbudha, made the throne of the Aghlabids tremble!
1. See Abū Ja'-far al-Dāwudī, Kitāb al-amwāl, partial trans, H H. Abdul-Wahah and F. Dachraoui as "Le régime foncier en Sicile au Moyen Ạge," Études d'orientalisme dédiées à la mémoire de Levi-Prouençal (Paris 1962), II 405-9, -428-29. See also H. Mones, "Al-Tanẓīm al-idārī wa'l-mālī li-Ifrīqiya wa'l-Maghrib khilāla 'aṣr al-wulāt," Majallal kullīyat al-ādāb (Kuwait), I (1973), pp. 94-97.
2. The chronology is somewhat uncertain: see EI2. III, 271. s.v. Ḥassān ibn al-Nu'mān.
3. In the sources we find, for example, waḍa'a al-kharaj 'alā 'ajam Ifrīqiya (be imposed the kharāj on the natives of Ifrīqiya), according to Ibn 'Abd al-Ḥakam, Futūḥ Ifrīqiya wa'l-Andalus, partial trans. A. Gateau as Conquête de l'Afrique dti Nord et de l'Espagne (Algiers 1947), pp. 80-81: but on the other hand, wa-ṣālaṣa 'alā al-kharāj, wa-katabahu 'alā 'ajam Ifrīqiya (he negotiated the kharāj and made a pact on this issue with the natives of Ifrīqiya), in the version recorded by Ibn 'Idhārī, Bayān al-mughrib, ed. G. S. Colin and É. Lévi-Provençal (Leiden 1948-51), I. 38. The first statement suggests an 'unwa system of land, the second a ṣulḥ system. Such are the problems underlying the uncertainty conceded by Saḥnūn. It should be noted, however, that Ibn 'Abd al-Ḥakam had a definite bias toward solutions of force, which explains his interpretations of the facts. This example is consciously chosen in view of the precise legal implications clearly underscored by R. Brunschvig in his article, "Ibn 'Abd al-Ḥakam et la conquête de l'Afrique du Nord par les Arabes: étude critique," AIEO, VI (1942-47), pp. 108-55. The version preserved by Ibn 'Idhārī seems, on the other hand, to be more in line with the general policy of Ḥassān, who sought to conciliate the natives, and is thus the one I prefer. It should be borne in mind that this policy, like that of Abū'l-Muhājir Dīnār in the past, aroused displeasure in high places and resulted in Ḥassān's recall and replacement by Mūsā ibn Nuṣayr, who introduced a more severe policy.
4. Mālikī, Riyāḍ al-nufūs, ed. H. Mones (Cairo 1951), I, 36.
5. Ibn 'Abd al-Ḥakam, Futūḥ, pp. 110 f; Ibn 'Idhārī, Bayān, I, 44-46.
6. The accession of the Fatimids (296/909) must have led to profound changes, with all those who had actively supported the Aghlabids losing their estates. The upheaval certainly became even more marked later under the Zirids. Whatever traces remained of the old system disappeared under the impact of the Hilalian invasion (443/1052).
7. Mālikī, Riyāḍ, I, 126-27; Ibn Nājī, Ma'ālim al-īmān (Tunis 1902-3), I, 245-46.
8. Al-Masrūqīn still exists, not far from the railway station of Sidi El-Hani, on the road from Qayrawān to Sousse. The ethnic surname derived from it, al-Masrūqī, is still in use.
9. See below, n. 16.
10. Abū'l-'Arab, Ṭabaqdt, ed. Mohammed ben Cheneb (Paris 1915), p. 234; Mālikī, Riyāḍ, 1, 81 f.; Ibn Nājī, Ma'ālim, I, 210 f.; Ibn 'Idhārī, Bayān, I, 49; Khayr al-Dīn al-Zinklī, Al-A'lām (Cairo 1954-59), VIII, 198.
11. Mālikī, Riyāḍ, I, 81; Ibn Nājī, Ma'ālim, I, 211. Ibn Nājī adds that his descendants had continued to own important property in Qayrawān until the town was ruined.
12. Mālikī, Riyāḍ, I, 81; Ibn Nājī, Ma'ālim, I, 197. After the assassination of Yazīd ibn Abī Muslim (103/721-22), it was suggested to him that because of his rank and the esteem in which he was held, he should provisionally assume power in Qayrawān while awaiting the nomination of a new governor (Ibn 'Idhārī, Bayān, I, 49).
13. Constantinople was besieged in 50/670 and in 96-98/714-16. It is not possible to say with certainty in which of these two sieges Mughīra Id, about 105/723) had taken part.
14. Ed. and trans. H. R. Idris as Manāqib d'Abū Isḥāq al-Jabnnyānī (Paris 1959).
15. In his Kitāb al-mamālik wa'l-masālik (trans. M. de Slane as Description de l'Afrique septentrionale, repr. Paris 1965, p. 20 of the Arabic text, p. 46 of the translation), Bakrī cites six Maḥrīs in the Sfax area, none of which can definitely be identified with the one that concerns us here. It is dangerous under these circumstances to identify Maḥris 'Alī or al-Maḥris al-Jadīd with the present town of Mahares (35 km south of Sfax), as Idris has done; see Manāqib, p. 198, n. 9. 'Iyāḍ (see Biographies Aghlabides, ed. M. Talbi, Tunis 1968, p. 320) and Tijāanī (Riḥla, Tunis 1958, p. 69) are of no help, since they simply reproduce what Labīdī says.
16. Labīdī, Manaqib, p. 3. Idris' translation (op. cit., p. 198) here does not take into account the actual situation. The word manzil, used alone or ill connection with the name of a tribe or person, always has the meaning of "village" or "town" in Tunisia (El-Menzel, Menzel Bou Zelfa, Menzel Djemil, Menzel Témime, and more recently Menzel Chaker and Menzel Bourguiba). The mention of Jabanyāna (Djebeniana), a town that still survives on the coast 35 km north of Sfax, as being among the manāzil belonging to 'Alī ibn Aslam, eliminates any further doubt as to the meaning of this word. It should be noted here that in the Madārik of 'Iyāḍ_ad (see Talbi, Biographies Aghlabides, p. 320), one reaḍs Hāssana instead of Jabānyana, surely because of a mistake by the copyist. The meaning of rbā', the plural of rah', still remains to be determined. This word is applied to comfortable houses and the space around them. Ibn Manẓūr (Lisān al-'arab, Beirut 1956, VIII, 102) provides several expressions which confirm this sense, hence the translation above.
17. īaīldl, āqndqib, p. 2. The translation IdrisIdt is (op. cit., p. 196) cannot be accepted here. The question is not one of high office, but rather of gifts of land; for this interpretation, which is more in keeping with the context, see Iban Manẓūr, Lisān, VII, 288 f.
18. See Labīdī, Manāqib, p. 2 (trans, p. 197), and n. 4.
19. On him see TalbL É.'Emitat Aghlabide (Paris 1966), pp. 170-99.
20. See Ibn al-Ābbar, Kitāb al-ḥulla al-siyarā, ed. H. Monés (Cairo 1963), II, 382.
21. On this uprising of the feudal lords, see TalbL'Émiratrat Aghlabide, pp. 290-96.
22. TalbL'Émiratrat Aghlabide, pp. 294 f.
23. Op. at., p. 285.
24. Maālikī Riyāḍ, I, 382-83.
25. Ial-AbbārbaḤ ullatla, I, 182, 183; see also TalbL'Émiratrat Aghlabide, p. 230.
26. Al-Qāḍī al-Nu'mān, Iftiāḥ al-da'wa, ed. W. al-Qīaḍī (Beirut 1970), pp. 90 f.
27. I'Idhārīari, Bayān, I, 169; A. Lézine, Mahdiya: recherches d'aéologiegisl hlamique (Paris 1965), pp. 16 ff.
28. Talbi, Biographies Aghlabides, p. 215. To evaluate the differences in wealth, it may be recalled that Saḥnūn's annual revenue, which was said to be comfortable, amounted to 500 dinars.
29. Mālikī, Riyāḍ, I, 380; Ibn Nājī, Ma'ālim, II, 109.
30. MālikīikRiyāḍyad, quoted by M. Amari, Bibhoteca Arabo-Sicula (Arabic text, Leipzig 1857), Appendix, p. 3.
31. This phenomenon is certainly not exclusively peculiar to Īfriqiya. Research in the Orient could lead to similar findings. Note, for example, that "the rich village of Rawān, near Balkh, belonged to Yaḥyā ibn Khālid personally" (EI2, I, 1033, s.v. Barāmika).
32. MālikīikRiyāḍyad, I, 254; Talbi, Biographies Aghlabides, pp. 87 ff.
33.T 'ba,hi. Biographies Aghlabides, pp. 97, 123, 126, 127.
34. Op. cit., p. 163.
35. Op. cit., p. 198.
36. Op. cit., p. 128.
37. MālikīikRiyāḍyad, I, 328; Talbi, op. cit., p. 163.
38. To clarify our ideas here, it should be mentioned that landowners holding 120 or 170 hectares are not uncommon even today, and that Tunisian olive groves, which have doubled in number since, independence, comprise some 55 million trees. It would seem that in the 3rd/9th century the trees were even more numerous. Saḥnūn gives us an example of the enormous effort of planting, but the owners of middle-sized properties naturally did not all go in for this activity only.
39. Mālikī, Riyāḍ, I, 213; Talbi, Biographies Aghlabides, p. 31.
40. Mālikī, Riyāḍ, I, 215. Rabāḥ died at the age of 38: on him see Mālikī. Riyāḍ, 1, 210-22; Abū'l-'Arab, Ṭabaqāt, pp. 45-52.
41. 'Ya\bi, Biographies Aghlabides, p. 108.
42. Ibn 'Abd al-Ḥakam, Futūḥ, p. 86; Ibn 'Idhārī, Bayān, I, 40, 43; Talbi. L'Émirat Aghlabide, pp. 32-35.
43. M. Talbi, "Dar Ṣinā'at Tunis," in Al-Sha'b, April-May 1967, pp. 22 ff.
44. Talbi, L'Émirat Aghlabide, pp. 35-41.
45. Op. cit., pp. 386-89, 416 ff.
46. G. Musca, L'emirato dt Bari, 847-871 (Bari 1964), pp. 70 f. Musca believes that the number 9,000 is exaggerated, particularly because of the capacity of the ships, which, even if the slaves were packed in like sardines, could only carry 1,500 each. Note, on this point, that Balawī, the author of Tāj al-mafriq (MS preserved in the BN in Tunis, no. 15060, fol. 24r), tells us that he had embarked at Tunis for the Orient on 17 Rabī' II 737/23 November 1336, on a sailing ship carrying 1,000 other passengers. Had the capacity of ships grown so much since the earlier Middle Ages? H. Ahrweiler (Byzance st la Mer, Paris 1966, p. 413) points out that in the ninth and tenth centuries, the same expression, for instance dromon or chélandia. can indicate a small, medium-sized, or large ship of the same class. She concludes that "it is therefore pointless to try to estimate the dimensions and crew serving on a dromon-chélandia" She does note, however, that "the largest important battle ships had two rows of oars . . . and a hundred oarsmen, as well as a crew of 200." We may conclude that large boats capable of carrying 1,000 men were in existence from the High Middle Ages onward.
47. R. S. Lopez, Medieval Trade in the Mediterranean World (London 1955), p. 34; see also Talbi, L'Émirat Aghlabide, p. 532, and n. 4.
48. Yves Lacoste thus writes, for example, with regard to medieval North African society: "Nor is it a question of a slave society; slaves are numerous there, but they hardly participate at all in production." See his Ibn Khaldoun, naissance de l'hisloire, passé du tiers-monde (Paris 1966), p. 37.
49. Talbi, L'Émirat Aghlabide, pp. 440. 462. For this period we have been able to find some precise details on prices. A young shepherd slave cost ten dinars (Talbi, Biographies Aghlabides, p. 214); a jāriya, or slave girl, 40 dinars (ibid., p. 219); a very beautiful young jāriya. 80 dinars (ibid., p. 216). For the sake of comparison, we should here add that at precisely the same time a modest one-room lodging cost 20 dinars, and that a barley loaf (sult) was worth one dirham (ibid., p. 219). These prices roughly correspond to those found for the same period in the Orient by F.. Ashtor, Histoire des prix et des salaires dans l'Orient médiéval (Paris 1969), pp. 56, 58, 87 f., 90 f.
This information does not seem wholly convincing (see below, n. 60). It is obvious, however, according to the different accounts found in the Orient as well as in the Muslim West, that a good slave was not worth more than a good mule. When slave labor was plentiful, for instance after military victories or successful pirate raids, a slave was even cheaper. This is a constant that can be verified with confidence in more recent times. As noted for another area of Muslim Africa: "Oversupply cuts prices. In the glutted market of Bagirmi (at the southern end of Lake Chad), in 1878, old men could be purchased for two to three dollars; women, old or young, for five; children of six or eight could be exchanged for an ordinary Bagirmi or Bornu shirt, which in Kuka cost three-quarters of a dollar. A comparable glut was recorded after a successful campaign by Samori's troopers in the west, at about the same time. Six chickens were sold for one slave, a sheep for three slaves, a cow for ten—but these prices may have been the consequence also of a shortage of food on that occasion." See Allan G. B. Fisher and Humphrey j. Fisher, Slavery and Muslim Society in Africa (London 1970), p. 165.
When a large supply of slaves was available in the markets of Ifrīqiya in the earlier period under discussion, and for similar reasons, the price situation cannot have been very different. On this point cf. also J. Mathiex, "Trafic et prix de l'honime en Méditerranée aux XVIIe et XVIIle siècles," Annales E. S. C., IX (1954), pp. 157-64; Lucette Valensi, "Esclaves Chrétiens et esclaves noirs à Tunis au XVIIle siècle," Annales E. S. C., XXII (1967), pp. 1267-88. It is a pity that in these two articles no prices for other commodities are offered for the sake of comparison. The prices given are meaningless without a scale of reference for evaluating their actual significance.
50. An example may be found in Talbi, L'Émirat Aghlabide, p. 522.
51. Mālikī, Riyāḍ, I, 126.
52. On him, consult Abū'l-'Arab, Ṭabaqāt, pp. 27-33; Mālikī, Riyāḍ, I, 96-103; Ibn al-'Imād, Shadharāt al-dhahab (Cairo 1931-32), I, 240, and Ibn al-Athīr, Kāmil (Cairo 1965), V, 315; VI, 12, 59.
53. Abū'l-'Arab, Ṭabaqāt, p. 28; Mālikī, Riyāḍ, I, 100.
54. On him see Abū'l-'Arab, Ṭabaqāt, pp. 84 f.; Mālikī, Riyāḍ, I, 189-96; Ibn 'Idhārī, Bayān, I, 97, 104; pseudo-Raqīq, Ta'rīkh Ifrīqiya, ed. M. Kaahi (Tunis 1968), pp. 124, 181, 232 f. This last work is wrongly attributed to Raqīq. On this point, see M. Talbi, "Un nouveau fragment de l'histoire de l'Occident musulman," Cahiers de Tunisie, XIX (1971), nos. 73-74, pp. 19-52.
55. Pseudo-Raqīq, Ta'rīkh, p. 124.
56. See p. 84; cf. also Mālikī, Riyāḍ, I, 193.
57. Ibn 'Idhārī, Bayān, I, 123.
58. Talbi, Biographies Aghlabides, p. 97, See also p. 127, where there is another reference to Saḥnūn's slave. On Saḥnūn's disciple 'Abd al-Jabbār, who tells this anecdote, see Abū'l-'Arab, Ṭabaqāt, pp. 145 f.; Mālikī, Riyāḍ, I, 366-70; Talbi, Biographies Aghlabides, pp. 294 f.; Ibn Nājī, Ma'ālim, II, 124.
59. This karam is illustrated by numerous traits mentioned in his biography; see Talbi, Biographies Aghlabides, pp. 214-25; Mālikī, Riyāḍ, I, 378-82. Ibn Ṭālib was so careful in cultivating his public image that he was suspected of having designs on the throne. This, among other things, was a factor leading to his downfall; see Abū'l-'Arab, Ṭabaqāt, pp. 145 f.
60. Talbi, Biographies Aghlabides, p. 214. We are told another similar anecdote in which Ibn Ṭālib again frees a young shepherd; see Biographies Aghlabides, p. 215; Mālikī, Ii 379.
The salary promised to the young freed man seems particularly low. Indeed, in the same biography (Biographies Aghlabides, p. 219) we learn that a barley loaf (suit) was worth one dirham. If the exchange rate of the dinar is set at a maximum of 1/20 to the dirham, the annual wage of the shepherd would not be more than the value of 40 loaves of poor quality. True, this can be explained by the bond of goodwill, meaning that the young man is still dependent for his upkeep on his former master, who is now his mawlā. Such may well have been the case. And the price of the slave? Ten dinars or 200 dirhams; that is, 200 loaves of bread. We know for a fact that the price of slaves was not very high. But a slave for a mouthful of bread! This seems rather far-fetched, and would indicate that the information on prices in our sources, since it does not stand cross-checking, is weak and of limited reliability. Such material must therefore be handled with great caution. For a comparison with some prices reported by Ashtor, see his Histoire des prix, pp. 50, 81.
61. Taibi, Biographies Aghlabides, p, 212; Mālikī, Riyāḍ, I, 378.
62. Mālikī, Riyāḍ, I, 97.
63. Op. cit., I, 189.
64. Taibi, Biographies Aghlabides, pp. 216, 218; Mālikī, Riyāḍ, I, 380; Ibn Nājī, Ma'ālim, II, 109.
65. Mālikī, Riyāḍ, I, 99.
66. Talbi, L'Émirat Aghlabide, pp. 281-97.
67. Mālikī, Riyāḍ, I, 126.
68. Talbi, Biographies Aghlabides, pp. 216, 219.
69. Talbi, Biographies Aghlabides, pp. 322 f.; Mālikī, Riyāḍ, I, 312, 388.
70. See M. Solignac, Recherches sur les installations hydrauliques de Kairouan et des steppes tunuiennes du Vile au Xle siècles (Algiers 1953).
71. Talbi, L'Émirat Aghlabide, pp. 169, 287-90.
72. Op. cit., p. 532.
73. Talbi, Biographies Aghlabides, p. 30; Mālikī, Riyāḍ, I, 213.
74. Ibn 'Idhārī, Bayān, I, 256 f. See also Ibn Nājī, Ma'ālim, III, 190, 191; Ibn al-Athīr, Al-Kāmil fīl-ta'rīkh (Beirut 1966), IX, 185.
75. H. R. Idris, La Berbérie Orientate sous les Zindes (Paris 1962), I, 149, 161, 227, 274, 293, 340, 350, 355; II, 658.
76. Idris, Zirides, II, 663. The material in the following pages is also of interest in this regard.
77. "Le inythe de la 'catastrophe' hlialienne," Annales E. S. C., XXII (1967), pp. 1102, 1105. Even if the Hiialian invasion had been made possible only by the economic and political collapse that preceded it, there is still no justification for denying its catastrophic character. One must avoid going from one extreme to the other. In this regard, see H. R. tdris, "L'invasion hilalienne et ses conséquences," Cahiers de civilisation médiévale, III (1968), pp. 353-71—a reply to Poncet's article. Cf. also Claude Cahen, "Quelques mots sur les Hilaliens et le nomadisme," JESHO, XI (1968), pp. 130-32; M. Brett, "Ifrīqiya as a Market for Saharian Trade," Journal of African History, III (1969), pp. 347-64; J. Berque, "Les Hilaliens repentis, ou l'Algérie rurale au XVe siècle d'après un manuscrit jurisprudentiel," Annates E. S. C., XXV (1970), pp. 1325-53; idem, "Du nouveau sur les Banū Hilāl?" Studio, Islamica, XXXVI (1972), pp. 99-113.
78. A Mediterranean Society, I: Economic Foundations (Berkeley 1967), p. 32.
79. "No contemporary attributed the same importance to them," he writes, and to this adds, "Travellers and chroniclers alike at the end of the eleventh century in no way confirm such a picture." See "Le mythe de la 'catastrophe' hilalienne," p. 1100. These statements can only be explained by ignorance of the sources. It would be interesting to discover which travelers Poncet was referring to. Ibn Ḥawqal had visited the country in 340/951-52; Bakrī (d. 487/1094) never set foot there, and in any case he most often reproduces the accounts of Warrāq (d. 363/973-74). As for Idrīsī (b. 493/1100, d. 550/1156 or 560/1166), he was not alive at the time of these events and usually paints a black picture of the country.
Some of our information we owe to Ibn Rashīq (d. 456/1064), Ibn Sharaf (d. 460/1068), and Ibn Sa'dūn (d. 485 or 486/1092-93), all eyewitnesses, as well as to Abū'l-Ṣalt (d. 529/1135), who had settled in Mahdlya in 506/1112-13 and gathered much of his information from there. Their works are almost certainly lost, but of course such was not always the case. Later compilers, particularly Ibn 'Idhārī, who was writing in 706/1306-7 and who quotes his sources, and Ibn Khaldūn (732-808/1332-1406), who does not quote them, did not create works out of their imaginations. Their works are essentially the compiled testimony of eyewitnesses. They have, to be sure, omitted much, and that is to be regretted; but they undoubtedly invented nothing. They had no reason to do so.
80. Ibn Bassām (d. 542/1147), Al-Dhakhīra, ed. Ṭaha Ḥusayn (Cairo 1939-45), IV, i, 177-84. Ibn Rashīq too has devoted a famous poem to the razing of Qayrawān (Dīwān, Beirut n.d., pp. 204-12; partially reproduced by H. H. Abdul-Wahab in Bisāṭ al-'aqīq, 2nd ed., Tunis 1970, pp. 72-73), but it is too mannered. The sorrow that shines through the poem of Ibn Sharaf is much more genuine and more poignant. His description of the haggard refugees covered with rags, of decimated families, of weeping widows and orphans, and the scenes of horror and violence, convey better than any of the chronicles the enormous extent of the suffering and the vast scope of the disaster.
81. Sec Archibald R. Lewis, Naval Power and Trade in the Mediterranean, A. D. 500-1100 (Princeton 1951), pp. 183-224.
82. To clarify this point and illustrate the reversal of the situation, it should be noted that in Morocco in 1708 only 800 Christian slaves are mentioned in the records, and 660 in 1723, as against 9,000 Muslim slaves in Malta in 1749. The situation was identical in Tunisia m the same period. See Mathiex, "Trafic et prix de l'homme," pp. 157-64.
83. A. Schaube, Handelsgeschichte der romanuchen Völker des Mittelmeer-gebiets his zum Ende der Kreuzzüge (Munich and Berlin 1906), p. 23.
84. As W. Hevd writes, "Charlemagne, and after him the popes Zacharias and Adrian I, took severe measures to end this horrible trade"; see his Histoire du Commerce du Levant au Moyen-Âge (Amsterdam 1959), I, 95.
85. A faqīh of the period, Ibn al-Ṣā'igh, had noticed the danger of this drain of gold reserves and had warned against it. See Idris, Zirides, II, 666-67.
86. The contemporary birth-control practices—including abortion—should also be recalled here. They were especially in demand at limes of crisis, as they also served to maintain the market value of slave girls. Not every jāriya was suited to become umm walad. See certain fatwās on this point in Idris, Zirides, II, 658 f.
87. Idris, Zirides, II, 663.
88. Idris gives Chapter 111 of his work (I, 127-203), a section dealing with the reign of Mu'izz. from 407/1016 to 442/1051, the title "L'Apogée." Poncet criticizes him, in the article already mentioned above, for allowing himself to be taken in by the acolytes of Mu'iz/., who were interested parties and committed partisans. The chapter in question would have been much more illuminating if its guiding line of research had been the desire to understand how the Hilalian "catastrophe" had been made possible.
89. A'māl al-a'lām, partial ed. H. H. Abdul-Wahab in Centenario Avian (Palermo 1910), II, 439. A new edition of the text has been prepared by A. M. al-'Abbādī and M. I. Kattānī under the title Al-Maghrib al-'arabī fīl-'aṣr al-wasīṭ, al-qism al-thālith min Kitāb a'māl al-a'lām (Casablanca 1964).
90. Talbir L'Émirat Aghlabide, pp. 530-36. On the oil trade in classical times, see H. Camps-Fabrer, L'Olivier el l'huile dans l'Afrique romaine (Algiers 1953). The major part of the forest of olive trees in the Sahel had been destroyed by the nomads (see Tījānī, Riḥla, p. 65; reproduced by al-Wazīr al-Sarrāj, Ḥulal, ed. M. H. al-Hīla, Tunis 1970, p. 322). Some years previously, the traveler Abdarī, who visited the country in 688/1289, noticed the same thing. Oil had started to be imported into the Sahel, he wrote; see his Al-Riḥla al-maghribīya, ed. M. al-Fāsī (Rabat 1968), p. 237.
91. Bakrī, Masālik, pp. 20 (Ar. text), 46 (trans.). As mentioned above, Bakrī generally reproduces Warrāq.
92. Mālikī, Riyāḍ, I, 127. The anecdote that tells of these exports is found in the biography of an Ifrīqiyan ascetic, Abū 'Īsā Marwān ibn 'Abd al-Raḥmān al-Yaḥṣubī, a contemporary of Buhlūl (d. 183/799), who lived in Alexandria. It is told by another Ifrīqiyan faqīh, Yahyā ibn 'Umar (d. 289/902).
93. Ibn Ḥawqal, Ṣūrat al-ard (Beirut n.d.), p. 95. The author had visited Ifrīqiya in 340/951-52, as he himself states in this work (pp. 83, 96).
94. Talbi, L'Émirat Aghlabide, pp. 136-38, 150, 153, 187, 189, 251, 284-85, 292-93, 304, 349, 522, 679.
95. Op. cit., p. 193.
96. Mālikī, Riyāḍ, I, 117; Talbi, Biographies Aghlabides, p. 32.
97. On him, see Talbi, L'Émirat Aghlabide, pp. 223-26.
98. Mālikī, Biyāḍ, I. 388; Talbi, Biographies Aghlahides, p. 322.
99. Dāwudī, Kitāb al-amwāl, p. 409.
100. On these three qadis, see Talbi, L'Émirat Aghlabide, index.
101. Op. cit., p. 233.
102. On the composition of this work, see M, Talbi, "Kairouan et le Malikisme espanol," Études d'orientalisme dediées à la memoire de Lévi-Provençal (Paris 1962), I, 320-24.
103. Talbi, L'Émirat Aghlabide, index. The Mu'tazilite qadi Ibn Abī'l-Jawād, who was dismissed in 232/846-47, had by this time held the post for 18 years.
104. J. Schacht, "Bibliothèques et manuscrits abâḍites," Revue Africaine (1956), p. 381, indicates an Ibāḍī Mudawwana attributed to Abū Ghānim Bishr ibn Ghānim al-Kburāsānī (ca. 200/815), whose authenticity is however questionable.
105. Talbi, Biographies Aghlahides, p. 104.
106. Talbi, L'Émirat. Aghlabide, index, s.v. Ibāḍites.
107. We know that Saḥnūn, the head of the Mālikī school in his time, had come to the qaḍā' thanks to Sulaymān ibn 'Imrān, the head of the Ḥanafī school, which was then in the majority. And in practice he divided his duties with the latter, allowing him the right to pronounce judgment following the principles of his school. See Talbi, L'Émirat Aghlabide, p. 233; idem, Biographies Aghlahides, p. 102.
108. On the circumstances of its composition, see Talbi, "Kairouan et le Malikisme espagnol," pp. 325-30. The work attracted little attention from researchers. G. H. Bousquet, however, gave an analysis published between 1958 and 1962 in La Revue Algérienne, Tunisienne et Marocaine de legislation et jurisprudence, and especially in the Annales de l'Institut des Etudes Orientates d'Alger. On the basis of Lhis analysis, he subsequently published an index in Arabica, XVII (1970), pp. 113-50. It should be pointed out that this index, which the author considers to be a "fairly complete" one in which "the leanings of each author necessarily come out," has no headings for "agriculture" or even "culture," while the one for "beating" figures in its proper place.
109. Al-Mudawwana al-kubrā (Cairo a.h. 1323), V, 399-46l.
110. Op. cit., V, 402, the section entitled "Mā lā taqa'u fīhi al-shuf 'a." See also op. cit., V, 423.
111. Op. cit., V, 426-31, the section entitled "Mājā'a fī'l-shf'a fī'l-thamara."
112. Op. cit., V, 432.
113. Op. cit., V, 424.
114. Op. cit., V, 399.
115. Op. cit., V, 399-401.
116. Op. cit., V, 404-6. 409-10.
117. Op. cit., V. 409.
118. Op. cit., V, 414.
119. Op. cit., V, 421.
120. Op. cit., V, 404, 418-19.
121. Op. cit., V, 418.
122. Op. cit., V, 419.
123. Op. cit., V, 424.
124. Op. cit., V, 462-97, 498-532.
125. Op. cit., V, 465. The section from which this passage is taken is entitled "Mā jā'a fī qismat al-qurā wa-fīhā dūr wa-shajar."
126. Op. cit., V. 464, 468, 474, 488.
127. Op. cit., V, 472, 486, 488-89.
128. Op. cit., V, 516.
129. Op. cit., V, 488-89.
130. Op. cit., V, 518-19.
131. Op. cit., V, 17-18, the section entitled "Musāqāt al-ba'l"; and that entitled "Musāqāt al-zar'".
132. Op. cit., V, 2-3.
133. Op. cit., V, 3-4. We know that many Ifrīqiyans frequented Mālik's court and that he held them in the greatest esteem. Some of them, like the qadi Ibn Ghānim, continued to consult him by letter on a regular basis. It is therefore possible that all these questions may have been put to Mālik by the Ifrīqiyans, in the presence of Ibn al-Qāsim, who preserved and transmitted his master's replies.
134. Op. cit., V, 4-6, 10-11, 16.
135. Op. cit., V, 4.
136. Op. cit., V, 6.
137. Op. cit., V, 16.
138. Op. cit., IV, 258-69.
139. Op. cit., IV, 265-66.
140. Op. cit., IV, 541.
141. Op. cit., IV, 528-29.
142. Op. cit., IV, 530-31; V, 25-39.
143. Op. cit., IV, 538.
144. Op. cit., IV, 298.
145. Op. cit., IV, 427.
146. Op. cit., IV, 459-61.
147. Op. cit., IV, 403.
148. Op. cit., IV, 406-7.
149. Op. cit., IV, 433-435-36.
150. Op. cit., IV, 266.
151. Op. cit., IV, 431.
152. Op. cit., IV, 434.
153. Op. cit., IV, 434-35.
154. Op. cit., IV, 436-41.
155. On the High Empire, one can consult, among others, H. d'Escurac-Doisy, "Notes sur le phénomène associatif dans le monde paysan à l'époque du Haul-Empire," Antiquites Africaines, I (1967), pp. 59-71; on the Low Empire, see Cl. Lepelley, "Déclin ou stabilité de l'agrictilture africaine au Bas-Empire," Antiquités Africatnes, I (1967), pp. 135-44. See also Christian Courtois, Les Vandales et l'Afrique (Paris 1955). Charles Diehl, L'Afrique Byzantine (Paris 1896), is of no help at all for the subject that concerns us here.
156. For example, the amount exacted for the year 366/976-77. See Ibn 'Idhārī, Bayān, I, 230.
157. Ed. Beirut 1956, p. 711.