Studia Islamica 107 (2012) 1-37 brill.com/si State, Land Tax and Agriculture in Iraq from the Arab Conquest to the Crisis of the Abbasid Caliphate (Seventh-Tenth Centuries) Michele Campopiano Centre for Medieval Studies, University of York, UK Introduction As several historians of the Middle East have noted, a signifijicant decline in tax revenue occurred in Iraq from the time of the Arabic conquest to the rise of the Buwayhids (945).1 The decline in tax revenue is a crucial element in the study of the social, political and economic history of the Medieval Middle East, because State income relied heavily on land tax. A decline in tax revenue from Iraq would have deeply afffected the stability of the Islamic Caliphate, since the Caliphate relied heavily on tax income from the Mesopotamian plain.2 A decline in tax revenue could be seen as a conse- quence of inefffijicient tax collection, of weaker control on tax collectors or of changes in tax assessments, but it could also indicate a decline in agricul- tural production. The scholars who have investigated this topic, including Robert McC. Adams, Eliyahu Ashtor and David Waines, however, have offfered only par- tial explanations for the reasons behind the decline in tax revenue. Adams 1 I would like to thank Professor Michael G. Morony and Professor Michael L. Bates for their suggestions. R.McC. Adams, Land behind Baghdad. A History of set- tlement on the Diyala plain (Chicago and London, 1965), 71-102; E. Ashtor, A Social and Economic History of the Near East in the Middle Ages (London 1976), 36-60; D. Waines, “The Third Century of Internal Crisis of the Abbasids”, Journal of the Economic and Social History of the Orient 20, 1977, 282-306. 2 Waines, “The Third Century”; P. Christensen, The Decline of Iranshahr. Irriga- tion and Environments in the History of the Middle East, 500 bc to ad 1500 (Copen- hagen, 1993), 86-9; H. Kennedy, “The Decline and Fall of the First Muslim Empire”, Der Islam 81, 2004, 3-29; in particular 11-5. DOI: 10.1163/19585705-12341234 2 M. Campopiano / Studia Islamica 107 (2012) 1-37 analysed an impressive data set generated by excavations in the Diyālā region of Iraq, which clearly show the decline in the number of tilled sur- faces in Iraq (particularly in the second half of the 9th century). He tried to link the decline in the size of the tilled surface with an inefffijicient water- supply administration, but it is hard to fijind any analysis of the conditions of land tenure and tax administration in his work. He also did not have the benefijit of a set of data on the prices of agricultural products, which were clearly connected to agricultural output. Ashtor realised that the decline in agricultural output was linked to the problems of land tenure and adminis- tration in the Arab Caliphates, and how this determined the social and political conditions that led to an ecological disaster which changed the face of the Near East in the Middle Ages. However, an extensive analysis of the elements of land tenure and tax administration that determined this evolution in Iraq cannot be found in his work. Waines linked the shrinkage of tilled surface area with the fall in tax revenue, and tried to relate them to the political crisis of the Abbasid Caliphate in the 3rd century ah. However, like Adams, Waines does not provide an analysis of the land tenure and its connections to tax collection and land administration. Furthermore, no link is established between the decline in tax revenue and the levels of food prices and wages, assuming that a decline of tax revenue automatically means a decline in agricultural production. In this paper I will show that decline in tax revenue was mainly due to the fall in soil output. I will demonstrate that the fall in soil output, political administration, land tenure and the methods of tax assessment were deeply interrelated. I will endeavour to explain how land tenure, tax administra- tion and political and social structure contributed to the decline in agricul- tural output, especially in food crops, and why the decline in tax revenue was an important indicator of this economic crisis. In particular, I will focus on the incentives and disincentives related to tax collection and adminis- tration and, in general, the role of the State in agricultural activity. Taxation was just another form of surplus extraction and reflected the relative power of diffferent social and political groups. Diffferent social categories were able to take advantage of this form of surplus extraction and influence reforms in tax assessment. It is also extremely important to understand the role of State investment in agricultural activity. The creation and maintenance of a good irrigation system, for example, was very important for Iraqi soil and was mainly implemented by the State.3 The State (an anachronistic term, 3 Adams, Land, 85; Ashtor, A Social and Economic History, 45.  M. Campopiano / Studia Islamica 107 (2012) 1-37 3 employed hereafter for simplicity’s sake) also had considerable influence on the structure of land tenure. In this paper I will focus on these multiple aspects. I will begin the analysis by refijining the data on tax revenue used by pre- vious historians, relying on primary sources for correction. I will conclude with an analysis of two important indicators: real grain prices (the main product of Iraqi agriculture) and military wages. The diachronic changes in the levels of these two indicators will provide us with: 1) another clue regarding the decline in agricultural production (increasing grain prices) and 2) a measure of the decline in the purchasing power of the State (increasing military wages). The Arab Caliphates did not radically change the system of taxation that they inherited from their pre-Islamic predecessor, the Sasanian Empire.4 Changes in religion and the ruling classes did not modify the nature of these Middle Eastern Empires, which were typifijied by centralised govern- ments ruled by elites who lived on the surplus extracted from the land. Taxation on the land was the main source of revenue for the State centra- lised. According to juridical sources, Kharādj, a tax imposed on the land of conquered populations who had not accepted Islam before the conquest and had not signed a special agreement (ṣulḥ), was levied on most of the conquered areas.5 Another kind of tax assessment should also be remem- bered—the tithe (ʿushr), which was assessed on the land of believers. This impost, set at 10% of production in the case of naturally irrigated land, could be reduced by half in the case of artifijicially irrigated land.6 Since 4 M. Khan Fateh, “Taxation in Persia. A Synopsis from the Early Times to the Conquest of Mongols”, Bulletin of the School of Oriental Studies 4/4, 1928, 723-43; M.G. Morony, Iraq after the Muslim Conquest (Princeton, 1984), 99-124; A. Gari- boldi, Il regno di Xusraw dall’anima immortale. Riforme economiche e rivolte sociali nell’Iran Sasanide del VI secolo (Milan, 2007), 178-9. 5 For a general survey of these topics, see: D.C. Dennet, Conversion and the Poll Tax in Early Islam (Cambridge, Mass., 1950); F. Løkkegaard, Islamic taxation in the classic period with special reference to circumstances in Iraq, second edition (Phila- delphia, 1978); H.M. Tabūtabāʾī, Kharāj in Islamic law (London, 1983), A. Oran, S. Rashid, “Fiscal Policy in Early Islam”, Public Finance 44, 1989, 75-101. See also: V. Lagardère, “Structures étatiques et communautés rurales: les impositions légales et illégales en al-Andalus et au Maghreb (XIe-XVe)”, Studia Islamica 80, 1994, 57-95. 6 See for example Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj. Part seven and excerpts from Abū Yūsuf ’s Kitāb al-Kharāj, translated and with introduction and notes by A. Ben Shemesh (Leiden-London 1965), 37 (English translation), 123 (Arabic text. From here onwards Kudama Ibn Djaʿfar, Kitāb al-kharādj, ed. Ben Shemesh). 4 M. Campopiano / Studia Islamica 107 (2012) 1-37 conversion to Islam and purchase of land by Muslims could cause the removal of kharādj, the authorities tried to establish kharādj as a tax assessed on the land itself, independent from the religious beliefs of the owner. In this form, kharādj difffered from what it is usually called djizya, the capitation, which was assessed on the non-Muslim population, even if the two words were sometimes used inter-changeably.7 The juridical trea- tises dealing with taxation and administration were often called Kitāb al-kharādj, the book of kharādj, and this reflects the general importance of this impost.8 This paper will focus on what was called the Sawād, the rural area of Iraq, extending roughly from Takrīt to ʿAbbadān and from Hulwān to Ḳādisiyya.9 This area is important for several reasons. Firstly, it was the rural region of Baghdad, the new capital city of the Caliphate in the Abbasid period. The area experienced a high rate of urbanisation under the Abbasid Dynasty,10 and the urban population obviously had to be fed. Sawād land is also of interest due to its particular status in the Caliphal taxation system. Accord- ing to many Arab historians, ʿUmar I (634-644) decided to not distribute the land of as-Sawād among the conquerors as he wanted to turn it into a trust to benefijit the whole umma.11 Instead of distributing the land among the Arab conquerors, he left it to its inhabitants, imposing kharādj upon them. Thus, he immobilised the revenue of the land and its inhabitants, creating a rent for public necessities. Ḳudāma Ibn Djaʿfar (end of the   7 See, for example, ash-Shaybānī, Kitāb as-siyyar al-kabīr (el Cairo, 1957), 15.   8 See, for example, Abū Yūsuf, Kitāb al-kharādj, translated and with intro- duction and notes by A. Ben Shemesh (Leiden-London, 1969. from here onwards: Abū Yūsuf, Kitāb al-kharādj (transl. Ben Shemesh)); Abū Yūsuf, Kitāb al-kharādj (El Cairo, 1886. Arabic text); Ḳudama Ibn Djaʿfar, Kitāb al-kharād (ed. Ben Shemesh). See: R. Verrier, Introduction à la pensée économique de l’Islam du VIIIe au XVe siècle (Paris, 2009), 67-9.   9 Morony, Iraq, 137; see Yāḳūt ar-Rūmi, Kitāb muʿdjam al-buldān, 5 vol. (Beirut 1955-7), vol. III, 272. 10 Adams, Land, 89-97; I.M. Lapidus, “The Evolution of Muslim Urban Society”, Comparative Studies in Society and History 15/1, 1973, 21-50. 11  Aḥmad Ibn Yaḥyā Ibn Djābir al-Balādhurī, Kitāb futūḥ al-buldān (Liber expug- nationis regionum, auctore imámo Ahmed Ibn Jahja Ibn Djabir al-Beládsorí ), ed. M.J. de Goeje (Leiden, 1866). It also useful to read the fijiscal rescript attributed to ʿUmar II included in the Sīra ʿUmar Ibn ʿAbd al-ʿAzīz of Ibn ʿAbd al-Hakam: H.A.R. Gibb, “The Fiscal Rescript of ʿUmar II”, Arabica 2/1, 1955, 1-16, p. 5. See also A.N. Poliak, “Classifijication of Lands in the Islamic Law and Its Technical Terms”, The American Journal of Semitic Languages and Literatures 57/1, 1940, 50-62.  M. Campopiano / Studia Islamica 107 (2012) 1-37 5 9th fijirst half of the 10th century) wrote that ʿUmar made the Sawād and Egypt fayʾ 12 and turned it into a mawḳūfa for the benefijit of all Muslims.13 I won’t discuss in this paper the concept of fayʾ and its evolution, we should simply note that the revenue of fayʾ land was supposed to benefijit the whole Muslim community. This fact was later used as a justifijication for limiting land sales in as-Sawād. Al-Māwardī (974-1058) wrote that the as-Sawād land could not be sold and that only “what was created on it of plant and building” could be transferred.14 Though these all appear to be later theori- sations, they help us to understand the importance of as-Sawād for the Caliphate, especially under the Abbasid when Iraq became the centre of the Empire. 1. Decline in Tax Revenue and Agricultural Output Tax revenues from Iraq probably declined after the Arab invasion. Towards the end of the Sasanian Empire, fijigures relating to tax revenue seem to have been higher than in the Arabic period.15 It is said, for example, that under Kawād (488-531), the revenue of as-Sawād was 150,000,000 dirhams mithāḳīl, the equivalent of 214,000,000 dirhams in the 9th century, according to the rate given by Ibn Khurdādbih.16 According to the estimates of Altheim and 12 F. Løkkegaard, “Fayʾ ”, in Encyclopaedia of Islam, 2nd Edition, 12 vol. (Leiden, 1960-2004. From here onwards EI(2)), vol. II,869-870; Løkkegaard, Islamic Taxa- tion, 38-72. 13 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj, (ed. Ben Shemesh), 135 (Arabic Manu- script). See also P.G. Forand, “The Status of the Land and Inhabitants of the Sawād during the fijirst two centuries of Islām”, Journal of Economic and Social History of the Orient, 14, 1971, 25-37, 29-30; see Abū ’l-Ḥasan al-Māwardī, al-Aḥkām al-sulṭāniyya wa-al-wilāyyā al-dīnīyya (Beirut, 1980), 163-4 and 219. Some Arabic sources claim that part of the land had been appropriated by Muslim soldiers early on, therefore ʿUmar asked them to return it to the Muslim community, M.G. Morony, “Land- holding in Seventh Century Iraq: Late Sasanian and Early Islamic Patterns”, in The Islamic Middle East, 700-1900, ed. by A.L. Udovitch (Princeton, 1981), 135-175, 154. 14 Al-Māwardī, Al-aḥkām sulṭāniyya, 219. 15 Adams, Land, 71. 16 Abū ’l-Ḳasim ʿUbayd Allāh Ibn Khurdādbih, Kitāb al-masālik wa ’l-mamālik (Liber viarum et regnorum auctore Abu ’l-Kasim Obaidallah Ibn Abdallah Ibn Khordadhbeh et Excerpta e Kitab al-Kharadj auctore Kodama Ibn Dja⁠ʾfar), ed. M.J. De Goeje (Leiden, 1889), 14 (Arabic text); this fijigure is reported also by Abū ’l-Ḳāsim Ibn Ḥawḳal al-Naṣībī in his Kitāb ṣurat al-arḍ. Opus Geographicum, vol. I (Leiden, 1938), 234; see Christensen, The Decline, 88. 6 M. Campopiano / Studia Islamica 107 (2012) 1-37 Stiehl, in the eighteenth year of the government of Husraw II (607), the revenue from central and southern Iraq was 168,000,000 dirhams in the old coinage, and at the end of his reign 240,000,000 dirhams.17 These fijigures seem to be quite high compared with the following ones. ʿUmar Ibn al-Khattāb, according to several geographers, such as Ibn Khurdādbih and al-Muḳaddasī (second half of the 10th century), collected 128,000,000 dir- hams, through having 500,000 dhimmī who had to pay djizya. Since djizya was assessed at between 12, 24 and 48 dirhams per head, using the medium taxe rate we can estimate that 12,000,000 dirhams were collected from the djizya (although usually the poll taxe and the land tax were paid together, as a single contribution), making the value of the crops in the Sawād 116,000,000 dirhams.18 A lower fijigure of 100,000,000 dirhams was occasion- ally reported for the value of the kharādj under ʿUmar Ibn al-Khattāb. Three explanations can be supplied for this discrepancy: these two fijigures may refer to two diffferent periods of ‘Umar’s reign, the lower fijigure may refer only to the Sawād al-Kūfa,19 or the lower fijigure may refer only to the kharādj while the higher fijigure might also include other forms of revenue.20 Muʿāwiya (661-680) collected 120,000,000 dirhams from the crops of the Sawād alone. I assume that this fijigure refers just to Sawād al-Kūfa, the area extending from Kaskar to the Zāb River and from Hulwān to Ḳādisiyya,21 since the districts of Tigris and Hulwān are listed apart.22 Another 20,000,000 dirhams was collected from Hulwān and 10,000,000 from the district of 17  F. Altheim-R. Stiehl, Ein asiatischer Staat. Feudalismus unter den Sasaniden und ihren Nachbarn, vol. I (Wiesbaden 1954), 41; Adams, Land, 70-71; estimates based on Ibn Khurdādbih; Kitāb al-masālik wa ’l-mamālik, 14-5 (Arabic text). See also Ibn Ḥawḳal, Kitāb ṣurat al-arḍ, 235; Morony is very sceptical about the liability of these estimates; Iraq, 120. 18  Ibn Khurdādbih, Kitāb al-masālik wa ’l-mamālik, p. 14 (Arabic text); Shams ad-Dīn ʿAbd Allāh Muḥammad al-Muḳaddasī, Kitāb aḥsan at-taḳasīm fī maʿrifa al-aḳalīm, 133 (see also al-Muqaddasī, The best Divisions for Knowledge of the Regions, translated by B.A. Collins (Reading, 1994), 121); Ibn Hawḳal, Kitāb ṣurat al-arḍ, 234. 19  “The income from Sawād al-Kūfa amounted, one year before ʿUmar’s death, to 100,000,000 of dirham”; Abū Yūsuf, Kitāb al-kharādj, (transl. Ben Shemesh), 80; Abū Yūsuf, Kitāb al-kharādj, 15 (Arabic text). 20 Al-Balādhurī, Kitāb futūḥ al-buldān, 270. 21  Morony, Iraq, 154. 22 See Ahmad Ibn Abī Yaʿḳub Al-Yaʿḳubi, Ta⁠ʾrīkh, ed. by M.Th. Houtsma, 2 vol. (Leiden, 1883), vol. II, 277-8.  M. Campopiano / Studia Islamica 107 (2012) 1-37 7 Tigris. These high fijigures were reached, in part, due to the threat of dis- missal made against the governor of the province.23 We should also remember that in 698 a new dirham was introduced, lowering the weight of the average coinage from about 3.98 g to 2.97 g.24 However, we should remember that at this stage a large part of the kharādj was collected by assessing a fijixed amount of dirhams on each djarīb. There- fore, the absolute fijigure would not have changed very much after the coin- age reform. Figures for tax revenue under al-Hadjdjādj (695-714) seem inconsistent, ranging from 18,000,000 to 40,000,000 dirhams.25 ʿUmar II (717-720) collected 124,000,000 dirhams. Judging from the sources, this fijig- ure probably also includes the poll tax, and thus a subtraction of about 12,000,000 dirhams provides a better estimate of the land tax.26 Under the Umayyad, the revenue remained fairly stable, without any noticeable increase. These fijigures have already been adjusted for tax col- lection expenditures and other normal expenditures.27 Morony is sceptical about the possibility of comparing Sasanian fijigures and Arabic ones. He explains, for example, that there was no provincial division exactly equiva- lent to the Sawād in the Sasanian period.28 He also shows that the total revenue of 150,000,000 Kawād is said to have raised can be related to 23 Adams, Land, pp. 84-85; Ahmad Ibn Abī al-Yaʿqūbi, Ta⁠ʾrīkh, vol. II, 277. Yāqūt ar-Rūmi writes that Ziyād Ibn Abī Sufyān, governor of Iraq (665-673) collected 125,000,000 di dirham, his son ʿUbayd Allah († 685) 135,000,000, Kitāb muʿdjam al-buldān, vol. III, 274. 24 Ph. Grierson, “The Monetary Reforms of ʿAbd al-Malik: Their Metrologi- cal Basis and Their Financial Repercussions”, Journal of the Economic and Social History of the Orient 3/3, 1960, 241-64; Ashtor, A Social and Economic History, 83; H. Gaube, Arabosasanidische Numismatik (Braunschweig, 1973), p. 63. Arabic authors are conscious of this change. Abū Yūsuf, for example, shows us that the dirham in use at the time of ‘Umar had a higher value than the dirham used in his time (at the end of the 8th century), since the weight of a dirham was one mithqāl, Abū Yūsuf, Kitāb al-kharādj (transl.. Ben Shemesh), 107; Abū Yūsuf, Kitāb al-kharādj, 63 (Arabic text). Al-Māwardī says that the weight of ten dirham is equal to that of seven dirham mithāḳīl; al-Māwardī, Al-aḥkām as-sulṭāniyya, 195. 25 Al-Balādhurī, Kitāb futūḥ al-buldān, 270; Ibn Khurdādbih, Kitāb al-masālik wa ’l-mamālik, 14 (Arabic text). 26 Adams, Land, p. 85; E. Ashtor, A Social and Economic History, 63; Ibn Khur- dadbeh, Kitāb al-masālik wa ’l-mamālik, 14 (Arabic text). 27 M. Shimizu, “Les fijinances publiques de l’Etat Abbaside”, Der Islam 42, 1965, 1-24, 9, A. von Kremer, Über das Einnahmenbudget des Abbasiden-Reiches vom Jahre 306 H. (918-919) (Wien, 1887), 9. 28 Morony, Iraq, 121-3. 8 M. Campopiano / Studia Islamica 107 (2012) 1-37 al-Māwardī’s claim that the land tax base in the 6th century was 150,000,000 djarīb.29 I think that, despite the care required when establishing any com- parison with Sasanian tax revenue, it is still possible to compare the Persian level of tax revenue with the Arabic one. Continuity in administration often survived at the district level.30 Arab geographers more or less knew to which area they were referring when they used words such as ‘Sawād’ and ‘Iraq’ because they were aware of which districts and sub-districts they were referring to. Ibn Khurdādbih, for example, who provided details of the tax revenue for Kawād’s kingdom, lists all the districts contributing to the fijinal revenue in his time, as Ḳudāma also does. After the fijirst half century of Abbasid rule, the trend seems to have been a steady decrease in revenue. In the lists of revenue included in al- Djahshiyārī’s Kitāb al-wuzarāʾ wa al- kuttāb it can be seen that under Hārūn ar-Rashīd the tax collected from the crops of as-Sawād was 80,780,000 dirhams and 14,800,000 dirhams were collected from other sources. This fijigure refers only to the Sawād al-Kūfa. If, in addition to the value of the crop from the Sawād, we add the revenue from Hulwān, Kaskar and Kuw- war Didjla (the lower Tigris valley), which is included in the later revenue list of the Sawād in Ḳudāma’s and Ibn Khurdādbih’s works, we get a total fijigure of 117,980,000 dirhams of land-tax revenue.31 Ṣāleḥ Aḥmad el-ʿAlī also considers another revenue estimate that could refer to the second year of the kingdom of ar-Rashīd (787-8). This list calculated around 87,860,000 dirhams of tax revenue from Sawād al-Kūfa, 11,640,000 from Kaskar, 20,000,000 and 20,000 dinars (that is 240,000 dirhams) from Kuwwar Didjla, and 4,000,000 dirhams from Hulwān, in total 123,740,000 dirhams.32 These two lists could be two diffferent versions of the same budget. However it is clear that both lists refer to the years of the reign of ar-Rashīd before 803, 29 Morony, Iraq, 120. See also Yāqūt ar-Rūmi, Kitāb muʿdjam al-buldān, vol. III, 272-3. 30 Morony, Iraq, 163-4. 31  The 14,800,000 dirhams collected from other sources is not included; Muḥammad Ibn ʿAbdūs al-Djahshiyārī, Kitāb al-wuzarāʾ wa al- kuttāb, ed. by M. al-Saqqa-I. al-Abyārī-A. Shalabī (El Cairo, 1938), 281-3; A. von Kremer, Über das Budget der Einnahmen unter der Regierung des Hârûn al-Raśíd (Wien, 1887), 4; D. Waines, “The Third Century”, 284-6. 240 riṭl of clay and 200 cloaks were also collected. 32 Ṣ.A. el-ʿAlī, “A New Version of Ibn al-Muṭarrif ’s List of Revenues in the Early Times of Harun al-Rashīd”, Journal of the Economic and Social History of the Orient, 14, 1971, 303-10.  M. Campopiano / Studia Islamica 107 (2012) 1-37 9 because both refer to Yaḥyā Ibn Khālid, the vizier who fell into disgrace in that year. In 819 (204 of hegira), according to Ḳudāma Ibn Djaʿfar, the rev- enue of Iraq was 177,200 kurr of wheat, 99,721 kurr of barley and 8,095,800 dirhams.33 His list does not include the district of Hulwān. The amount of wheat listed is possibly the result of an error in the manuscript tradition. Ḳudāma calculated a fijinal revenue of 108,457,650 dirhams on the basis of a wheat price of 1,367 dinars for 100 kg and a barley price of 0.68 dinars for 100 kg, if the fijigure of 177,200 kurr of wheat was correct the fijinal revenue would be 144,127,025.85 dirhams (one kurr = 2,925 kg). Both von Kremer and De Goeje propose that it should read 117,200 kurr of wheat.34 Ḳudāma gives the amount of revenue for this year because it is the fijirst one available in the State archives, as previous records were burnt in the year 197 of hegira. As was mentioned earlier, according to Ḳudāma the total revenue of as-Sawād seems to have been 108,457,650 dirhams. However, he clearly overestimated the value of the crops, since he calculated them according to the prices in use in his time, at the beginning of the 10th cen- tury. Furthermore, the data for the district of the Tigris are taken from the year 260 of the hegira, not 204 like the other data, but this should not sig- nifijicantly afffect the fijinal result. He also wrote that the ṣadaḳa of the area of al-Baṣra amounted to 6,000,000 dirhams, making the total sum 114,457,650 dirhams.35 At the time of Ibn Khurdādbih (middle of the 9th century), the revenue of as-Sawād amounted to 70,650 kurr of wheat, 112,050 kurr of barley or rice, and 11,848,840 dirhams.36 The fijirst interesting thing to note is the abrupt fall in wheat production. Rice and barley were more robust grains and capable of growing on highly saline soils. If this amount is added into the total amount—in dirhams—of tax revenue, the fijinal result is 113,097,760 dirhams (calculating grain prices in dinars and using an exchange rate for dinar/dirham of 1=20). It should be remembered, however, 33 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj, ed. M.J. De Goeje, 239-40. 34 Von Kremer, Über das Einnahmebudget, 43, Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj (ed. M.J. De Goeje), 180. 35 Adams, Land, p. 85; Ashtor, A Social and Economic History, 42. 36 A. von Kremer, Culturgeschichte des Orients unter den Chalifen (Wien 1875- 1877), vol. I, 291; Ibn Khurdādbih, Kitāb al-Masālik wa ’l-Mamālik, 8-14 (Arabic text). For the district of Hulwān, the text reports an income of 1,800,000 dirham. De Goeje proposes to follow Djahshiyārī and Ibn Khaldūn who have 4,800,000. I don’t agree with him because this fijigure is based on an earlier estimate of the tax revenue, therefore I read 1,800,000 dirham; Ibn Khurdādbih, Kitāb al-masālik wa ’l-mamālik, 10, n. 8. 10 M. Campopiano / Studia Islamica 107 (2012) 1-37 that these two sums are overestimated as they are based on prices at the beginning of the 10th century (see Appendix). Scholars have also proposed estimating the tax revenue for the year 893 (280 hegira) at 75,000,000. This estimate is based on a tax-farm leased in that year that included about half the Sawād.37 If we look at the budget for 306 hegira (918-19) under al-Muḳtadir, it can be seen that the amount of revenue for all Iraq is 1,547,734 dinars, which equals only 30,954,680 dirhams. If the customs, revenues from markets, mints, and other kinds of imports and duties are subtracted and the land revenue from Hulwān (kharādj e diyyāʿ revenues) is added,38 the result is more or less the fijinal amount of Sawād land-tax income, 1,367,705 dinars. This fijigure more or less equals 27,354,100 dirhams, an abrupt fall compared with the previous fijigures of tax revenue.39 As previously mentioned, the decline in imperial revenue could also have been determined by an inefffijicient tax collection system. Ḳudāma, for example, thought that the decline in imperial revenue in his time was due to the weakness of the administration; many people simply refused to pay the taxes.40 Numerous tax reductions and tax exemptions could be granted by the government and the fijiscal prefects (ʿummāl) could profijit from the weakness of central power in order to increase their personal income.41 As was explained before, Waines, in his paper on the decline of the Abbasid Empire, explains the phenomenon by means of the decline in agricultural output: if the land-tax revenue decreased, it was because the land was pro- ducing less or less land was being tilled. His analysis seems correct, espe- cially if two other phenomena are taken into consideration. Firstly, a large part of the tax income was collected, from the end of the 8th century, as a share of the land production (according to a system called muḳāsama, which will be described later) and this system clearly links agricultural out- put and state revenue. The second relevant element is the rise in prices from the end of the 8th century to the middle of the 10th century. Wheat was the most important 37 H. Busse, “Das Hofbudget des Chalifen al-Muʾtadidbillah (279-289/892-902)”, Der Islam 33, 1967, pp. 11-36. Waines, “The Third Century”, p. 286. 38 von Kremer, Über das Einnahmenbudget, 29. 39 von Kremer, Über das Einnahmenbudget, 26 (Arabic source) and 32 (German translation). 40 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj, 249 (Arabic text). 41  H.Q. es-Samarraie, Agriculture in Iraq During the 3rd Century ah (Beirut, 1972), 185-91.  M. Campopiano / Studia Islamica 107 (2012) 1-37 11 component in the diet of Middle Eastern populations.42 It is difffijicult to avoid the impression that real grains prices are increasing between the beginning of the Abbasid period and the fijirst half of the 10th century, whereas real wages are declining. That points to a decline of grain produc- tion in the Sawād. Between the second half of the 8th century and the beginning of the 10th century, nominal prices of wheat increased around ten times (from 0.112 dinars for 100 kg to 1.36 dinars).43 In the same period, the wages of mounted soldiers (who had a very high bargaining power) increased between two and six times (according to the diffferent sources available, see Appendix). Nominal wages increased, but real wages declined: if we want to calculate the real price of wheat in terms of wages, we can see how the price of 100 kg of wheat would have represented at the end of the 8th century only 1.723% of the monthly wage of a horseman, while at the beginning of the 10th century it would have been as much as 10.9% of a monthly wage of a cavalier. The increase in prices could in the long term be the efffect of the devalu- ation of coins. It can be very risky to estimate the efffect of the devaluation of coins in determining inflation, since every analysis has to be based on the specimens which are still available. In Ehrenkreutz’s analysis of several specimens of mintless dinars from the Umayyad and early Abbasid period, we can see how their standard of fijineness and weight do not show any con- sistent decline, with the exception of several specimens from the period of the civil war between al-Amīn and al-Ma⁠ʾmūn (810-813).44 From the reign of al-Ma⁠ʾmūn onwards, the inscription on the dinars mention the name of the mint where the coins were struck, but we cannot see any relevant sign of decline in the standard of fijineness of the dinars struck in Baghdad and Samarra before the rise of the Buwayhids.45 Furthermore, the standard of 42 D. Waines, “Cereals, bread and society. An essay on the stafff of life in medi- eval Iraq”, Journal of the Economic and Social History of the Orient 30, 1987, 255-85. 43 See Appendix. 44 A.S. Ehrenkreutz, “Studies in the Monetary History of the Near East in the Middle Ages”, Journal of Economic and Social History of the Orient 2, 1959, 128-61, 135-43; see also M. Broome, A Handbook of Islamic Coins, London 1985, pp. 29-32; M.L. Bates, “The ‘Abbasid Coinage System, 833-946”, Presented at the Annual Meeting of the Middle East Studies Association, 1996 (available on http:// www.numismatic.org/./collections/abbasid.html); J.L. Bacharach, “Thoughts about Pennies and other Monies”, Middle East Studies Association Bulletin, 35, 2001, 2–13 (availableonhttp://fp.arizona.edu/mesassoc/bulletin/Pres%20Addresses/Bacharach .html). 45 Ehrenkreutz, “Studies”, 144-5. 12 M. Campopiano / Studia Islamica 107 (2012) 1-37 fijineness was increased under the reign of al-Ma⁠ʾmūn.46 An analysis of silver coins shows that their weight tended to vary more, with diffferent types of silver coins having diffferent weights (often depending on the province where they were struck) and it is very likely that silver coins were circu- lated by weight. Dirhams struck between the 7th and the 10th century were almost always of high quality silver.47 All these elements show that the efffect of the devaluation of coins in determining price increases can be neglected. The relationship between the decline in grain production and the fall in tax revenue is not only confijirmed by price data, but reveals that the decline was probably even sharper. As a large part of the tax revenue was collected in kind, the increase in grain prices suggests that the fall in agricultural production during the 9th century was very considerable. The increase of nominal (but not real) wages, especially military wages, also shows how the ‘purchasing power’ of the State income dramatically declined (see Appendix). 2. Land and Taxation Policies As we have seen, after the Arab conquest some land was considered only liable to pay the tithe (ʿushr, amounting to 10% or sometimes 5% of the harvest). Tithe-payers were often Arab landlords who received large land grants from the political authorities.48 Therefore, it was necessary to com- pensate for the fijinancial losses caused by the introduction of the tithe by 46 T. El-Hibri, “Coinage Reform under the ʿAbbāsid Caliph Al-Ma⁠ʾmūn”, Journal of the Economic and Social History of the Orient, 36, 1993, 58-83, in particular 72. 47 W. Hinz, Islamische Masse und Gewichte (Leiden, 1955), 1-4; J. Walker, A cata- logue of the arab-Sasanian coins: Umaiyad governors in the east, Arab-Ephthalites, ʿAbbăsid governors in Ṭabaristăn and Bukhără (London, 1967), CXLVI-CLIV and 1-174 (catalogue); A.A. Gordus, “Non-Destructive Analysis of Parthian, Sasanian, and Umayyad Silver Coins”, in Near Eastern Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles, ed. by D.K. Kouymjian (Beirut, 1974), 141-62; M.L. Bates, “Islamic Numismatics”, Middle East Studies Association Bulletin 12/2, 1978, 1-16; 12/3, 1978, 2-18; 13/1 1979, 3-21; 13/2, 1979, 1-9; Id., “History, Geography, and Numismatics in the First Century of Islamic Coinage”, Revue Suisse de numismatique 65, 1986, 231-62; Broome, A Handbook, p. 28-29. 48 See, for example: M. Morony, “Landholding and Social Change: Lower al-ʿIraq in the Early Islamic Period”, in Land Tenure and Social Transformation in the Middle East, ed. by Tarif Khalidi, (Beirut, 1984), 209-22.  M. Campopiano / Studia Islamica 107 (2012) 1-37 13 raising tax rates on kharādj-payers, taxpayers liable to pay the heavier land tax. Tax assessment on kharādj land mainly followed a system called ʿala ’l-misāḥa: from a fijixed portion of land (usually one djarīb) a fijixed amount of money and/or crops was collected. The tax was also often levied on untilled land. According to many of our sources, the impost assessed varied with the kind of crop. The tax was usually levied in cash and the rate depended on the crop. Normal tax rates under the Sasasians included, for example, one dirham per djarīb for wheat and barley, 5 or 6 for rice, 8 for vines, 7 for clover, one dirham for 6 fruit trees, and one dirham per 4 or 6 date palms (depending on the kind of palm).49 For the period follow- ing the Arab conquest the tax rates frequently reported in the sources were: 12 dirhams for a djarīb of olives, 10 dirhams for vines, 8 or 10 dirhams for date palms, 3 or 8 dirhams for clover, 6 dirhams for sugar cane, 5 dir- hams for sesame and cotton, 4 dirhams for wheat and 2 dirhams for barley.50 We are told that on wheat and barley ʿUmar Ibn Khaṭṭāb levied either 4 and 2 dirhams respectively, or one dirham and one ḳafīz. Some sources, such as al-Balādhuri’s Kitab futūḥ al-Buldān and al-Māwardī’s Al-aḥkām al-sulṭāniyya, afffijirm that tax assessments also depended on diffferent levels of soil fertility and the proximity to markets and water courses.51 Under Arab rule the tax rate was subject to small fluctuations, but did not change radically over time (except when tax was assessed according to the system called muḳāsama, as will be seen later on).52 It should also be remembered that the weight of the dirham changed at the end of the 7th century, from about 3.98 g to 2.97 g. This would have reduced the taxes paid in silver, but without returning them, in most cases, to Sasanian levels.53 Initially, the tax was also assessed according to a smaller djarīb compared with the size in 49 Morony, Iraq, 100. 50 Occasionally diffferent rates are found: for example, al-Muḳaddasī wrote that in Iraq a djarīb of palms was taxed at 8 dirhams; Ibn Khurdadbeh also speaks of 8 dirhams for palms, 6 for vines and 6 for the medic. It should be remembered that there are no offfijicial documents, just geographical and juridical treatises; Løkkegaard, Islamic Taxation, 102-103; Shams ad-Dīn ʿAbd Allāh Muḥammad Al-Muḳaddasī, Kitāb aḥsan at-taḳasīm fī maʿrifa al-aḳalīm, 133; Abū ’l-Ḳasim ʿUbayd Allāh Ibn Khurdādbih, Kitāb al-masālik wa ’l-mamālik, 14 (Arabic text). 51  Aḥmad Ibn Yaḥyā Ibn Djābir Al-Balādhurī, Kitāb futūḥ al-buldān, 271; Abū ’l-Ḥasan al-Māwardī, Al-aḥkām al-sulṭāniyya, 220-1; on this problem, see also Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj, 121 (Arabic manuscript). 52 Morony, Iraq, 103. 53 For example, paying one dirham per djarīb in the Sasanian period would have meant to give 3.98 g. of silver to the State. With Arabic tax rates, after the 14 M. Campopiano / Studia Islamica 107 (2012) 1-37 Sasanian times: 1050 m2 compared with 1592 m2, which implicitly increased the tax rate by about 33%. However, use of the larger djarīb was restored, probably under the governor Ziyād Ibn Abī Sufyān (665-673).54 Poll tax also increased: Sasanian rates were 4, 6, 8 or 12 dirhams per year, Arab rates 12, 24 or 48 dirhams per year according to personal wealth. This tax was often collected at the same time as the land tax, so it represented yet another burden for the peasants.55 The social and institutional framework of land tenure is very relevant to tax administration. To understand the nature of land tenure in the Sawād between the 7th and 10th centuries is a very delicate issue, due both to the scarcity of sources and the conceptual difffijiculties of legal theory. Several theories have been formulated, such as van Berchem’s idea that most of the as-Sawād ultimately belonged to the State, and Lambton’s approach, which tends to treat Sawād land as divided into individual properties.56 This prob- lem cannot be solved in this paper. However, it is necessary to try to under- stand some relevant issues that link land tenure to tax collection. Village estates were the normal agricultural and taxation units at the time of the Arab conquest, while at the end of the Sasanian Empire small- holdings were often absorbed by large estates, which were very common in the Sawād and along the Tigris. Under the Persians, such estates were usu- ally apportioned offf from crown lands as grants. The Arabs confijiscated the crown lands, but left many of the land grants to Persian noblemen.57 After the Arab conquest Muslim landlords acquired estates in the Sawād of Kūfa and existing landlords converted to Islam. It has been argued that the Islamic conquest may have led to the partial emancipation of peasants, leading peasant-cultivators to become the proprietors of some of these weight of the dirham had been changed, the taxpayer would have paid 11.88 g. of silver for wheat and 5.94 for barley; Ashtor, A Social and Economic History, 83. 54 Morony, Iraq, 103, Hinz, Islamiche Masse, 55 and 65-66. 55 Morony, Iraq, 107-8. 56 M. van Berchem, La propriété territoriale et l’impot foncier sous les premiers chalifes. Études sur l’impôt du kharāg (Genève, 1886); A.K.S. Lambton, Landlord and Peasant in Persia. A Study of Land Tenure and Land Revenue Administration, second edition (Oxford, 1969). Morony, “Landholding” 135. More generally, see also W. Schmucker, Untersuchungen zu einigen wichtigen bodenrechtlichen Konsequen- zen der islamischen Eroberungsbewegung (Bonn, 1972); M.G. Morony, “Grundeigen- tum im frühislamischen Irak”, in Jahrbuch für Wirtschaftsgeschichte.Sonderband 1987 Das Grundeigentum in Mesopotamien, ed. by B. Brentjes (Berlin, 1988), 135-47. 57 Altheim-Stiehl, Ein asiatischer Staat, p. 141; Morony, “Landholding”, T. Daryaee, Sasanian Persia: The Rise and Fall of an Empire (London, 2009), 29 and 147-8.  M. Campopiano / Studia Islamica 107 (2012) 1-37 15 lands.58 Nevertheless, sources show that a concentration of the land into the hands of Muslim elites occurred after the Arab conquest.59 We are informed of attempts to enforce the rule that kharādj land should not be sold to Muslims, both under ʿUmar Ibn al-Khaṭṭāb (634-644) and under the last of the ʿUmayyads.60 Yaḥyā Ibn Ādam wrote that ʿUmar forbade the acquisition of kharādj land (arḍ al-kharādj)61 and Abū ʿUbayd said that the purchase of the land owned by the ahl adh-dhimma (non-Muslim poll- tax payers) was forbidden.62 Some anecdotes also confijirm that Muslim landlords were purchasing land after the conquest. For example, Yaḥyā Ibn Ādam reports that a dihḳān approached ʿAbd Allāh Ibn Masʿūd, a Compan- ion of the Prophet, asking him to buy his land. The Arab accepted, provided that the old landlord would continue to pay kharādj on it.63 Finally, Muslim jurists tried to show that kharādj was imposed on the land in order to justify the idea that if the landlord was Muslim but the land was kharādj land the heavier tax had to be paid anyway. Ḳudāma Ibn Djaʿfar later wrote that there was no dispute among the jurists that if somebody of the ahl al-ʿanwa (the ‘conquered’ population under Muslim rule) accepted Islam he should nevertheless pay kharādj.64 Abū Yūsuf (†798) also wrote that no one had the right to turn kharādj land into ʿushr land.65 This also seems to be the policy pursued by the authorities. ʿUmar II (717-720) afffijirmed that conversion to Islam exempted individuals from the poll tax 58 Morony, “Landholding”, 157 and 165; B. Johansen, The Islamic Law on Land Tax and Rent, (London-New York-Sydney, 1988), 11-2. Johansen writes: “By transform- ing land into a commodity and the land tax into a proof of property, the Hanafijite legal position on kharādj stands in the way of the attempt to consider the peasant as a serf bound to the soil”. 59 J. Banaji, “Aristocracies, Peasantries and the Framing of the Early Middle Ages”, Journal of Agrarian Change 9, 2009, 59-91, 78-86. 60 Morony, “Landholding”, 139-40 and 155; Lambton, Landlord, 22-3; Haque, Landlord and peasant in early Islam: a study of the legal doctrine of “Muzāra⁠ʾa” or sharecropping (Islamabad, 1977), 216-218. 61  Yahyā Ibn Ādam, Kitāb al-kharādj (English translation), 49; Yahyā Ibn Ādam, Kitāb al-kharādj (Arabic text), 39. 62 Abū ʿUbayd, Kitāb al-amwāl (Bayrūt, 1989), 157. 63 Yahyā Ibn Ādam, Kitāb al-kharādj (English translation), p. 49; Yahyā Ibn Ādam, Kitāb al-kharādj (Arabic text), 39. Edition: Le livre de l’impot foncier de Yaḥyā Ibn Ādam, ed. Th.W. Juynboll, (Leiden, 1896). 64 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj (English translation), 26. 65 Abū Yūsuf, Kitāb al-kharādj (English translation), 3; Abū Yūsuf, Kitāb al-kharādj, 49 (Arabic text). 16 M. Campopiano / Studia Islamica 107 (2012) 1-37 but not from the kharādj, whereas the mawālī (who were Muslim converts and clients of the Arabs) asked to pay ʿushr.66 The Caliph also stressed that the property of converted people would remain fayʾ if they accepted Islam after the conquest.67 Therefore kharādj continued to be paid on kharādj land, although other lands were only liable to pay ʿushr. Fiscal agents usually tried to deal with responsible individuals within the village or district in order to help them collect taxes and mediate between the peasantry and the bureaucracy.68 These people were usually larger landlords, recruited from the general class of tunnāʾ, large or medium- sized landlords.69 There is no doubt that these groups took advantage of their position by trying to increase the burden on the peasantry, or by lending money to the farmers at a high interest rate. The Syriac Chronicle of Zuqnīn, written at the end of the 8th century at the monastery of Zuqnīn, near Amida (nowadays Diyarbekir on the Tigris in Turkey),70 provides a gloomy picture of the situation: “When they were assessed for seventy thousand, they themselves demanded a further three (thousand), not spar- ing or saying: ‘such and such a village is broke and cannot bear (such a burden)’ ”.71 These men were responsible for tax collection in their villages and in this way part of the tax burden could be shifted onto their tenants, even if, according to legal theory, tax payment proved land ownership. Legal theory usually put the burden of kharādj on the shoulders of the landlords,72 while payment of ʿushr was often shared with the tenant, but there is evidence that in some cases the tenants paid kharādj.73 In the 66 A.A. Duri, “Notes on taxation in Early Islam”, Journal of the Economic and Social History of the Orient 17, 1974, 136-44, in particular 138. 67 Gibb, “The Fiscal Rescript”, 3. 68 C. Cahen, “Fiscalité, propriété, antagonismes sociaux en Haute-Mésopotamie au temps des premiers Abbasides”, Arabica 1, 1954, 136-152; Løkkegaard, Islamic Taxation, 166-8; Morony, “Landholding”; Idem, “Landholding and Social Change”, 210. 69 Løkkegaard, Islamic Taxation, 166-8; Shimizu, “Les fijinances publiques”, 15-6. 70 Mistakenly attributed in the past to Dionysius of Tell-Mahre. The Chronicle of Zuqnīn. Parts III and IV. ad 488-775, Translated from Syriac with Notes and Intro- duction by A. Harrak (Toronto, 1999), 1. See also W. Witakowski, The Syriac Chron- icle of Pseudo-Dionysius of Tel-Maḥrē (Uppsala, 1987). 71  The Chronicle of Zuqnīn, 267-8; see also 244-5. 72 Johansen, The Islamic Law, 7-19. 73 Morony, “Landholding”, pp. 149-53; M. ʿAbdul Jabbar, “Agricultural and irriga- tion labourers in social and economic life of ʿIrāq during the Umayyad and ʿAbbāsid Caliphates”, Islamic Culture 47, 1973, 15-31.  M. Campopiano / Studia Islamica 107 (2012) 1-37 17 anecdote reported by Yaḥyā Ibn Ādam mentioned above, the Companion ʿAbd Allāh Ibn Masʿūd bought the land of a dihqān provided that the old landlord still paid kharādj on it.74 This anecdote shows that sometimes the Persian landlords were reduced to being the tenants or stewards of new masters and that payment of kharādj did not always burden the landlord. The fijirst people to benefijit from tax collection were, of course, the tax collectors who were supposed to rely on the villages’ resources while they were collecting taxes around the country and they often collected their wages directly from the imposts. There is no reason to doubt that the demands of the jurists (for example Abū Yūsuf in the late 8th century), to appoint rightful and honest people to the position of tax collector and to enquire constantly into their behaviour, had some real foundation.75 At lower levels, it seems that there was little control over tax-collectors. We read in the Chronicle of Zuqnīn: “because they trusted that they did not have to answer to anyone, (they) mercilessly and pitilessly determined (the tax) and imposed it on the people”.76 In addition to the kharādj land, post-conquest Iraq was also marked by the presence of crown land (sawāfī al-ustān). It consisted of all the land owned by the Persian royal family, their supporters, the lands of the fijire temples, and also the royal mints, mills and cisterns, etc. The peasants residing on these lands were bound to pay a tax /rent proportional to the yield.77 Since the reign of Muʿāwiya, the caliphs had tried to extend the estates under their direct control by reclaiming Sasanian crown lands and swamp lands, especially in the area of al-Batāʾih (the famous Big Swamp).78 The Chronicle of Zuqnīn also shows that the agents of the crown lands tried to arbitrarily extend crown land by confijiscating properties that were not registered in the census, the cadastral assessment for the land tax.79 Sawāfī lands were also granted to Muslim conquerors, following a general process that led to the authorities granting large estates to their 74 Yahyā Ibn Ādam, Kitāb al-Kharādj, p. 49. 75 Abū Yūsuf, Kitāb al-Kharādj (transl. Ben Shemesh), 107; Abū Yūsuf, Kitāb al-kharāğ, 63-4 (Arabic text). 76 The Chronicle of Zuqnīn, 267. 77 Morony, Iraq, 104-5. 78 I.M. Lapidus, “Arab Settlement and economic Development of Iraq and Iran in the Age of Umayyad and Early Abbasid Caliphs”, in The Islamic Middle East 700- 1900: Studies in Economic and Social History, ed. by A.L. Udovitch (Princeton, 1981), pp. 177-208, pp. 183-7; cfr. Ahmad Ibn Abī al-Yaʿqūbi, Ta⁠ʾrīkh, vol. II, 277-8. 79 The Chronicle of Zuqnīn, 234-5 and 256. 18 M. Campopiano / Studia Islamica 107 (2012) 1-37 favourites. Usually the only land tax they were compelled to pay was the tithe, as is noted above. This process started to expand under the governor ʿAbd Allāh Ibn Amīr Ibn Ḳurayẓ (650-56), who granted land to many of his relatives and mawālī.80 Land grants were also a very common way of inte- grating military wages.81 The destiny of the large landlords who owed their position to land grants was connected with their political fortunes and mis- fortunes: dynastical wars and expropriations could threaten their position. The difffusion of large land grants (ḳatīʿa) led to much disappointment among the smaller aristocracy of the dahāqīn and Arab landlords. The con- trast between the crown and its favourites who were granted big estates on the one hand, and the local aristocracy and Arab tribesmen on the other, led to the revolt against al-Hadjdjādj in 701. This uprising led to the battle of Dayr al-Djamādjim, culminating in victory for al-Hadjdjādj. This governor also contributed, as has been mentioned before, to the ruin of many dahāqīn by refusing to repair their dykes.82 Here there is further evidence of the paradox of the ʿUmayyad government of Iraq. The land grants usually helped, as I will show in the next chapter, in the reclamation of ruined land, but they created a large aristocracy, tied to the destiny of the dynasty, who largely profijited from the ruin of smaller landlords. 3. Water Management and Land Reclamation In addition to the efffects of inequality on the distribution of the tax burden and of the birth of a new, politically supported landholding elite, an analy- sis of the evolution of water management is also crucial to understanding the decline of Iraqi agriculture. Political instability and the continuous wars that upset the Middle East in the 7th century seriously undermined the highly developed water management system of the Late Sasanian period, especially after Heraclius’ invasion of Iraq (627). An example of this is the Great Swamp in Southern Iraq which was created in 628 when the Tigris expanded out of its bed, and attempts made to restore it by the Umayyad governor Khālid al-Ḳasrī were unsuccessful.83 Archaeological 80 Morony, “Landholding and Social Change”, 211-3. 81 H. Kennedy, The Armies of the Caliphs. Military and Society in the Early Islamic State (London-New York, 2001), 81-5. 82 Morony, “Landholding”, 162; see al-Balādhurī, Kitāb futūḥ al-buldān, 272-3. 83 Ashtor, A Social and Economic History, 61.  M. Campopiano / Studia Islamica 107 (2012) 1-37 19 surveys show the abandonment of a large area of the Euphrates’ lower central floodplain.84 There appears to be some discontinuity in the occupa- tion of the central floodplain of the Euphrates between the Sasanian irriga- tion and settlement patterns and Islamic ones (from the second half of 7th century).85 Ḳudāma speaks difffusely of floods and the breaking of dykes during the Late Sasanian period and the Arab conquest of Iraq.86 Islamic authors, such as al-Māwardī, make allusions to a decline in the tilled sur- face area after the conquest.87 This should be linked to the fact that Sasanian irrigation was highly cen- tralised, requiring a highly stable central government to work and to pre- vent the problems connected with high levels of irrigation (such as salinisation).88 In the late Sasanian period, direct investment by the State had been crucial for agricultural growth: written sources report that Hus- raw I (531-579) invested large amounts of money in building forts, bridges, canals, Ḳanāts (the famous underground canals) and other irrigation works and even in rebuilding villages.89 The most interesting case was that of the Ḳāṭūl al-Kisrawī, a huge canal (about 100 km long) drawn from the Tigris that solved the chronic problems of water shortage in some parts of the Diyālā region. Islamic sources, such as Yāḳūt ar-Rūmī (1179-1229) and al-Ḳazwīnī (1203?-1283), attribute the building of the canal to Husraw. At the same time, the king established the three administrative districts of the Upper, Middle and Lower Nahrawān, where the Ḳāṭūl al-Kisrawī played a central role in the canal system. As this illustrates, the growth of irrigated agriculture and the development of a central administration were clearly linked to each other: the fall of this administration would seriously have damaged the system.90 As I mentioned in the previous chapter, one of the most important strat- egies adopted by the Caliphate to bring untilled land under cultivation was to grant extensive land ownership to wealthy and powerful members of the 84 R.McC. Adams, Heartland of Cities. Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates (Chicago and London, 1981), 205. 85 Adams, Heartland, 206-11. 86 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj (ed. M.J. De Goeje), 240-1. 87 Al-Māwardī, Al-aḥkām as-sulṭā niyya, 218-9. 88 Adams, Heartland, 213. 89 Adams, Land, 71; F. Rahimi-Laridjani, Die Entwicklung der Bewässerungs- landwirtschaft im Iran bis in sasanidisch-frühislamische Zeit (Wiesbaden, 1988), 457. 90 Adams, Land, p. 76. 20 M. Campopiano / Studia Islamica 107 (2012) 1-37 elite. These land grants usually enabled the landlords to recover large areas which had fallen out of use because they had become swamps or lacked irrigation. There were disputes among the jurists concerning the necessity of granting the land as a condition for reclaiming it, but to a large extent this was a theoretical discussion: land grants and reclamation went together. In most cases only wealthy people were able to undertake this work and this led to an extension of the large estates.91 Owners of privately reclaimed land usually only paid the tithe (but there were debates among the jurists concerning this point) or even only half of it due to the extensive artifijicial works of irrigation or drainage that had to be carried out.92 Many of the land grants given by caliphs and governors consisted of dead land from the area of Basra, which was considered in the sources to be a tithe-paying area. Basra was an important area for sugar cane plantations.93 The largest estates were often used for market-oriented monocultures, often, from the end of the 7th century, due to the introduction of Zandj, African slaves.94 These ḳatīʿa, the Arabic word for these land grants, could yield very high revenues. ʿAbd Allāh Ibn Darrādj, for example, was able to reclaim swamp land which yielded fijive or maybe 15,000,000 dirhams (silver coins).95 These grants could be enormous, as large as 8000 djarīb (1 djarīb = 1592 m2).96 Arabic sources offfer some concrete evidence of how members of the elite were awarded these grants. An interesting example is provided by Al-Balādhurī, the famous historian of the Arab conquest, who describes the reclamation project of Maslama, the son of the Umayyad Caliph ʿAbd al-Malik, in southern Iraq during the reign of al-Walīd (705-715).97 Maslama 91 Ashtor, A Social and Economic History, p. 46. 92 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj, 37 (English translation), p. 23 (Arabic text). 93 es-Samarraie, Agriculture in Iraq, 94. 94 A.A. Duri, Arabische Wirtschaftgeschichte, aus dem Arabischen übersetzt von J. Jacobi (Zürich-München, 1979. First Arabic edition Beirut, 1969), 90-2. Slaves were also common on crown land; Morony, “Landholding and Social Change”, 210 and 218-9. On the Zandj: A. Popovic, La révolte des esclaves en Iraq au IIIe/IXe siècle, (Paris, 1976). 95 Ashtor, A Social and Economic History, 61. 96 Morony, “Landholding”, p. 150, pp. 157-9; Idem, “Landholding and Social Change”. 97 al-Balâdhurī, Kitāb futūḥ al-buldān, 294. This event has been broadly discussed by Hugh Kennedy: “Elite Incomes in the Early Islamic State”, in The Byzantine and Early Islamic Near East, VI. Elites Old and New in the Byzantine and Early Islamic Near East, edited by J. Haldon and L.I. Conrad (Princeton, 2004), 13-28, 21-2.  M. Campopiano / Studia Islamica 107 (2012) 1-37 21 proposed the investment of money in a reclamation project, the cost of which had been estimated to be about 3,000,000 dirhams. Maslama asked that the land that remained under water was given to him as ḳatīʿa. The reclamation project was successful, farmers were brought in to work this land, and many other people entrusted their farms (diyyāʿ ) to Maslama for ‘protection’. The reclamation process came to an end around the middle of the 8th century.98 Problems with salinisation arose and it was difffijicult to fijind drinkable water in the area of Basra. Al-Muḳaddasī writes at the end of the 10th century that water in Basra was scarce and even had to be imported by ship.99 Since the revival of dead land was a requisite to acquiring property rights on it, with an emphasis on irrigational works, it is also possible that this law contributed to the over-irrigation of some areas, causing increasing salinisation. The nature of the soil in Iraq is very fragile, requiring both irri- gation and drainage to keep land productivity high.100 Water management was extremely important, since the delicate ecological balance that allowed high soil productivity could be seriously threatened by irresponsible land administration. Summer heat could dry out the land, but the floods of the Tigris and the Euphrates could be also a serious menace to the environ- ment, turning the tillable surface into swamp.101 Soil salinisation could reduce land productivity, and over-irrigation or flooding could increase the salinity level. This delicate balance of drainage and irrigation work was dif- fijicult to achieve for a class of landlords whose property rights on land were established by land grants (ultimately linked to political positions) and by their capability to revive dead land, which was mainly linked to the pres- ence of irrigation works. The arbitrary nature of these grants would have prevented any serious planning of the irrigation works, since the land grants were based on political rather then economic choices. A change in the central government entailed a change in the landholding elite, as can be seen to have happened under Abbasid policies in Southern Iraq, where   98 Morony, Iraq, 106; Christensen, Decline, 88-91.   99 Al-Muḳaddasī, Kitāb aḥsan at-taḳāsīm, 129. 100 C.E. Larsen, “The Mesopotamian Delta Region: a Reconsideration of Lees and Falcon”, Journal of the American Oriental Society 95/1, 1975, 43-57; E. Wirth, Agrargeographie des Irak (Hamburg, 1962), 97-8; P. Buringh, Soils and Soil Condi- tions in Iraq (Baghdad, 1960), 255-63; J.H.G. Lebon “The New Irrigation Era in Iraq”, Economic Geography 31/1, 1955, 47-59, 47-50. 101  Lebon, “The New Irrigation Era”, 47-50. 22 M. Campopiano / Studia Islamica 107 (2012) 1-37 the land was granted to people close to the dynasty.102 Climate, of course, also had a negative influence on the process of soil erosion and aridifijica- tion in the Near East, but the decline in the area of the tilled surface in particular fijinds its roots in a more inefffijicient way of dealing with an envi- ronment that, in the past, had been successfully tamed by other societies.103 Poor water management was one of the causes of the process of land abandonment that can be observed in Iraq in this period. Between the 7th and 10th centuries, there was a growing phenomenon of abandonment of the land by peasants; indeed, under al-Hadjdjādj force was used in order to bring them back to the land.104 Many villages and fijields were abandoned, leading to a decline in the size of the cultivated area. The abandonment of land was related to fijiscal pressures, as shown by the Chronicle of Zuqnīn, which describes the dramatic conditions of peasants trying to escape the greed of tax collectors.105 But this abandonment of the land was also deeply interrelated with problems of water management. Abū Yusūf, one of the acutest observers of the situation in Iraq at the end of the 8th century, wrote that in the time of ʿUmar Ibn Khaṭṭāb the greater part of the land was culti- vated (which suggests that in his time the situation was rather diffferent) and he added: “there are lands which were not cultivated for a period of 100 years, more or less, and could not be developed and irrigated without expense beyond the means of the peasants”.106 Abū Yusūf put a lot of emphasis on the revival of the dead lands and the importance of irrigation for this purpose. He also mentioned cases of uncultivated areas where ancient watercourses and canals had been found, which could be cleared to supply the untilled land.107 He argued that “if the main canals and water- courses, from the Tigris and Euphrates, need clearing and cleaning the 102 Morony, “Landholding and Social Change”, 215-7. 103 Ashtor, A Social and Economic History, 51; Lamb, Climate, History and the Modern World (London-New York, 1995), 184-5; A.S. Issar and M. Zohar, Climate Change-Environment and Civilization in the Middle East (Berlin-Heidelberg-New York, 2004), 214-5. 104 Forand, “The Status”, 28. On al-Hadjdjādj’s taxation policies see also: M. Cosgel, R. Ahmed and Th. Miceli, Law, State Power, and Taxation in Islamic His- tory, University of Connecticut. Department of Economics Working Paper Series, Working Paper January 2007, revised July 2008, 20. 105 The Chronicle of Zuqnīn, e.g. 225, 237 and 257. 106 Abū Yūsuf, Kitāb al-kharādj (Transl. Ben Shemesh), 100; Abū Yūsuf, Kitāb al-kharādj, 28 (Arabic Text). 107 E.g. Abū Yūsuf, Kitāb al-kharādj (transl. Ben Shemesh), 106; Abū Yūsuf, Kitāb al-kharādj, 62-3 (Arabic text).  M. Campopiano / Studia Islamica 107 (2012) 1-37 23 expenses should be borne by the Treasury and by those directly benefijiting from such watercourses”. He also stated that “the expense of the upkeep of the walls on the river banks, to prevent flooding, of the dams and the water locks on the Tigris and Euphrates and similar great rivers, must be borne by the Treasury alone, because it is in the public interest that they should be kept in order as any mal function will cause damage to agriculture and decrease the income from taxation”.108 Ḳudāma afffijirmed that the most important Muslim jurists, including Abū Ḥanīfa (ca. 699-767) and Malik (708/716?-796), also asserted that the supervision of irrigation and the dam- ming of the overflow of the great rivers had to be sustained by the govern- ment at the expense of the Treasury.109 It is possible that the relatively high tax revenue reported under the reign of Hārūn ar-Rashīd should be seen in connection with the great attention paid to a good central administration and proper water management system under the fijirst Abbasids. Abū ʿUbayd Allāh, the fijirst author of a book on kharādj, a general treatise about taxa- tion, already appointed under al-Manṣūr to the retinue of his heir al-Mahdī, was made vizier under the government of the latter (probably appointed in 775, died 786-7).110 Abū Yūsuf, whose Kitāb al-kharādj pays much attention to the problems of water management, was appointed kāḍī of Baghdad at the demand of Hārūn ar-Rashīd (however it is not known whether this appointment was made by al-Mahdī, al-Hādī or Hārūn ar-Rashīd) and he was often consulted by the Caliph of The Arabian Nights.111 It seems that several irrigation works were accomplished under the fijirst member of the ‘Blessed Dynasty’, initially particularly in the areas of Basra and Wāsit, with attention then shifting to the areas of Baghdad, Samarra and the central reaches of the Tigris river.112 A key case in understanding the history of water management in Ancient and Medieval Mesopotamia, the Diyālā plain, has been analysed by the archaeologist Robert McC. Adams. This area was crossed by the Diyālā River and the Ḳāṭūl al-Kisrawī canal. A network of smaller canals allowed the Diyālā region to have very high land productivity and a very large tilled 108 Abū Yūsuf, Kitāb al-kharādj, 106; Abū Yūsuf, Kitāb al-kharādj, 63 (Arabic text). See: Verrier, Introduction, 69. 109 Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj, 62; see also ʿAbdul Jabbar, “Agricul- tural and Irrigational Labourers”, 235-52. 110 S. Moscati, “Abu ʿUbayd Allah”, in EI (2), vol. I, 157-8, D. Sourdel, Le viziral ‘abbaside de 749 à 936 (132 à 324 de l’Hégire), vol. I (Damascus, 1997, 94-103. 111  J. Schacht, “Abu Yusuf ” in EI (2), vol. I, 164-5. 112  Lapidus, “Arab Settlement”, 187-9. 24 M. Campopiano / Studia Islamica 107 (2012) 1-37 surface area. The tilled surface area shrank from 8000 square kilometres at the end of the Sasanian era to just 6000 square kilometres by the middle of the 9th century.113 Two cities, fijive smaller urban centres, three large towns and ninety villages or boroughs were abandoned soon after the end of the Samarran period.114 Adams hypothesises that under the Sasanian Empire water management was more efffijicient and highly centralised.115 The evi- dence supports his views, as I have shown before: State investment and centralised administration were crucial in the last phase of Sasanian agri- cultural development. Other archaeological surveys conducted in the Cen- tral Euphrates floodplain show, according to Adams, a decline in land use since the 7th century, which almost became abandonment in the 9th and 10th centuries.116 Of course there are various methodological risks in using, as Adams does, the absence of diagnostic sherds to prove the absence of occupation at a site, and one of these problems can be the need to use pot- tery to establish a relative chronology of settlement.117 However, we should not forget that this process of abandonment seems to take place over sev- eral centuries, therefore making the possibility of an error in chronology or in settlement occupation smaller. Furthermore, there is less and less evi- dence of irrigational works taking place in the written sources from the 9th century onwards.118 113 Adams, Land, 102. 114 Adams, Land, 98. 115 Adams, Land, 70-1; about Sasanian water management, see also R.McC. Adams, “Intensifijied Large-Scale Irrigation as an Aspect of Imperial Policy: Strate- gies of Statecraft on the Late Sasanian Mesopotamian Plain”, in Agricultural Strate- gies, ed. by J. Marcus and C. Stanis (Los Angeles, 2006), 17-37. 116 Adams, Heartland, pp. 218-23; for land abandonment, see also R.McC. Adams and H. Nissen, The Uruk Countryside. The Natural Setting of Urban Societies (Chi- cago and London, 1972), 63-5. 117 M.G. Morony, “Land Use and Settlement Patterns in Late Sasanian and Early Islamic Iraq”, in The Byzantine and early Islamic Near East II Land Use and Settle- ment Patterns, ed. by G.R.D. King and Averil Cameron (Princeton, 1994), 221-29, 222-6. 118 Lapidus, “Arab Settlement”, p. 189. Christensen has argued that the deci- sive factor in the decline was the spread of epidemics in the Middle East and their demographic consequences. But demography does not directly explain the decline in water management. Whether population decline plays a role in water management or not depends on the social and political organisation of the water management system. The key issue is to explain how water management was organised. I remind the reader that plague epidemics had already occurred on the  M. Campopiano / Studia Islamica 107 (2012) 1-37 25 4. Reform in Tax Assessment, the Difffusion of Tax-Farming and Grants of Tax Revenue (End of the 8th Century to the 10th Century) As we have seen, the land was taxed in Iraq according to the system called ʿalā ’l-misāḥa. According to al-Balādhurī, at the end of the reign of al-Manṣūr (754-775) the people of the Sawād asked for a change in tax assessment. Under al-Mahdī (ruled 775-785), a new taxation system, based on a share of the crops, was enforced. As we know, this system was called muḳāsama: Yahyā Ibn ’Adam said: “concerning the muḳāsama in the Sawād; the people (nās) asked for this from the ruler in the last period of the Caliph al-Manṣūr’s reign”.119 Under this form of tax assessment, instead of a fijixed amount of money or crops, the taxpayers had to pay a share of the crop: one half of the crop for land irrigated by flooding, a third for land irrigated by water- wheel and one quarter for land irrigated by animal-turned wheels, accord- ing to al-Māwardī.120 According to Ḳudāma Ibn Djaʿfar, Abū ʿUbayd Allāh, al-Mahdī’s vizier,121 suggested that people who irrigated their land with buckets should pay with one third of their crops, while those who irrigated their land with waterwheels should pay with one quarter, and in general advised that the government should take into account all the expenses affforded by the taxpayer when calculating the fijinal assessment.122 Abū Yūsuf suggested that the people of the Sawād (ahl as-Sawād) should pay 40% of their output in the case of naturally irrigated lands and 30% for artifijicially irrigated lands producing wheat and barley.123 Other sources give other rates, with a minimum rate of 10%.124 Al-Māwardī writes that tax rates amounted to one half of the crop for land irrigated by flooding, a third for land irrigated by waterwheel and one quarter for land irrigated by ani- mal-turned wheels.125 Al-Māwardī’s rates seem to have been closer to real- ity, especially as Abū Yūsuf ’s rates are suggested as optimal ones and were Mesopotamian plain in the 6th century, a period of expansion for irrigated agricul- ture; Christensen, Decline, 88-99. 119 Al-Balādhurī, Kitāb futūḥ al-buldān, 271. 120 Al-Māwardī, Al-aḥkām as-sulṭāniyya, 221. 121  S. Moscati, “Abu ʿUbayd Allah” in EI (2), I, 157-8. 122 Ḳudama Ibn Djaʿfar, Kitāb al-kharādj, 40-1 (English translation), 119 (Arabic text). 123 123. Abū Yūsuf, Kitāb al-kharādj (English translation), 101; Abū Yūsuf, Kitāb al-kharādj (Arabic text), 28. 124 C.E. Bosworth, “Muqāsama” in EI (2), vol. VII, 506-7. 125 Al-Māwardī, Al-aḥkām as-sulṭāniyya, 221. 26 M. Campopiano / Studia Islamica 107 (2012) 1-37 not considered by the jurist to have been the rates actually assessed. Aṭ-Ṭabarī writes that the al-Ma⁠ʾmūn reduced the taxation rate to 40%. It is doubtful that this change should be considered to have been a permanent reform, but it is evident that it reveals that 50% was already a common rate at the beginning of the 9th century.126 According to Ibn aṭ-Ṭiḳṭaḳā (wrote in 1302), the new system was intro- duced thanks to Abū ʿUbayd Allāh.127 It is interesting to note this bottom- top-bottom movement. The initiative for the tax reform came from the ‘people’ (nās in the Arabic text of Balādhurī), but was further promoted by a minister. Why does there seem to be this special agreement between the two levels? Who are the ‘people’ who asked for the reform and why? It should fijirst be noted that we are told of an abrupt fall in prices around the year 772 in upper Mesopotamia: the price of wheat was one dirham for fijive muḍḍ, which is equal to 0.027 dinars for 100 kg of wheat. From the Chronicle of Zuqnīn, we can estimate an average price of around 0.112 dinars for 100 kg of wheat just a few years before in 766-767,128 and 0.088 dinars for 100 kg in 768-769.129 Ashtor argues that this fall in prices occurred because of a lack of currency.130 With such a fall in prices, it would have been more difffijicult to pay taxes in cash. I have made a rough calculation of how much the taxpayer would have been expected to hand into the State with such prices. If we accept the reported tax rate of 4 dirhams per djarīb of wheat, it implies that about 1,229 kg of wheat per djarīb (1,592 m2) had to be paid in tax. According to Adams, land productivity per hectare in the Diyālā region of Iraq would have amounted to around 1,132 kg of wheat per hectare.131 Of course these are very rough estimates, but we can defijinitely assume that 126 Abū Djaʿfar Muḥammad Ibn Djarīr aṭ-Ṭabarī, Ta⁠ʾrīkh ar-Rusul wa⁠’l-mulūk. Annales quos scripsit Abû Djafar Mohammed Ibn Djarîr At-Tabarî, ed. M.J. De Goeje, 15 vol. (Leiden, 1879-1901), III series, p. 1039; cfr.; M. Nejatullah Siddiqi and S.M. Ghazanfar, “Early Medieval Islamic Economic Thought. Abu Yousuf ’s (731-798 ad) Economics of Public Finance”, in Medieval Islamic Economic Thought. Filling the Great Gap in European Economics, ed. by S.M. Ghazanfar (New York, 2003), 209-27, 213-4. 127 Muḥammad Ibn ʿAlī Ibn at-Ṭiḳṭaḳa, Al-Fakhri, translated by C.E.J. Whitting (London, 1947), 176-177; Arabic text: Al-fakhrī fī ’l-ādāb as-sulṭaniyya wa ’d-duwwal al-Islāmiyya (Ḥalab, 1997), 179-80. 128 The Chronicle of Zuqnīn, 186 and 216. 129 The Chronicle of Zuqnīn, 245. 130 E. Ashtor, Histoire des prix et des salaires dans l’Orient médiéval (Paris, 1969), 42. 131  Adams, Land, 17 and 101-2.  M. Campopiano / Studia Islamica 107 (2012) 1-37 27 the situation was quite difffijicult for the taxpayers. Nevertheless, it would perhaps have been possible to solve the emergency using temporary mea- sures. Therefore there should also be other reasons behind the reform that took place in the Sawād. We should not forget that the muḳāsama was specifijically used to collect tax on grain, the staple food of the Mesopotamian populations since the Sumerian Era. Al-Māwardī, for example, explained that on palms, vines and trees the tax had to be collected under the ʿalā ’l-misāḥa system.132 Shi- mizu has shown how one of the reasons for the adoption of the muḳāsama system in as-Sawād was an attempt by the State to control the grain market in the area of southern Iraq, where the most important cities of the Empire needed to be supplied. With ʿalā ’l-misāḥa, the merchants had a lot of opportunity to speculate. Under ʿalā ’l-misāḥa the taxpayers had to sell the grain at a bargain price in order to pay their taxes: the merchants could therefore lead the dance.133 With the new system, it was probably easier for the government to ensure that the growing urban population was fed, in an area that became, with the rise of the Abbasids, the political centre of the Empire. However, we should not forget that under the ʿalā ’l-misāḥa system a portion of the taxes could nevertheless be collected in kind. Muḳāsama did not necessarily mean tax collection in kind: the prevalent implication of the use of the word muḳāsama was the collection of the tax burden as a share of the crop (muḳāsama comes from the Arabic root ḳsm, to divide, to share). Basically, what muḳāsama offfered was the possibility of risk sharing: in the case of a bad harvest the taxpayers would be not oppressed, in the case of a good harvest the State could increase its reve- nues. With muḳāsama it would have been possible to share the crops or the money gained by selling them at market, apparently preventing contention between the authorities and the taxpayers.134 We should also remember that Balādhurī wrote that the people of the Sawād asked for the reform: I believe that the difffusion of the new tax assessment was also related to the necessity of establishing good relation- ships between the Muslim landlords/taxpayers, who were purchasing or had purchased land in Iraq, as I have shown in the third chapter of this 132 Al-Māwardī, Al-aḥkam as-sulṭaniyya, 221. 133 Shimizu, “Les fijinances”, 20-2. 134 M. Campopiano, “Land Tax ʿalā l-misāḥa and muqāsama. Legal Theory and the Balance of Social Forces in Early Medieval Iraq (6th-8th Centuries ce), Journal of the Economic and Social History of the Orient, 54, 2011, 239-269. 28 M. Campopiano / Studia Islamica 107 (2012) 1-37 paper, and the new dynasty. This attitude was in marked contrast with the more severe fijiscal policies of the Umayyads, who had to rely more heavily on cadastres to fijirmly establish fijiscal duties on the basis of the ʿalā ’l-misāḥa tax assessment (based on the size of the tilled surface area).135 In particular, I refer to the land survey of the Sawād conducted in 105 ah (723-4 ad) by the governor ʿUmar Ibn Hubayra at the request of Yazīd II. This land survey is reported to have caused hardship and discontent among taxpayers.136 The payment of a share of the crops instead of a fijixed amount would have meant that the landlords were better able to affford the payment of taxes. This system, however, perfectly fijits into a situation in which invest- ment in land administration was declining, and the landlords were living offf their land as parasitic rentiers. If a share of the crops has to be paid instead of a fijixed amount of money or cash, the State and taxpayer can share the risk, but this system also stimulates less investment in agricul- ture. Under the ʿalā ’l-misāḥa system, the taxpayers had to pay four dirhams for each djarīb of land planted with wheat, but they could keep the remain- ing crops for themselves: to produce one more pound of wheat meant to have one more pound for themselves. Under the muḳāsama assessment, each pound of wheat had to be shared with the State. Another downside to muḳāsama was that it would have been more dif- fijicult to calculate the State budget. A similar case can be found in the his- tory of the Sasanian Empire. At fijirst, the Persians adopted a system similar to muḳāsama: a portion of the harvest was taken as tax. The harvest varied from year to year, making it difffijicult for the government to estimate its rev- enue. In the case of a shortage, the government had recourse to extra imposts, causing riots and discontent. The reform of the started tax assess- ment, already seen by Kawād (488-531), was fijinally fulfijilled by Husraw I.137 135 A similar process has been observed by Elton E. Daniel for Khurasan “It is more likely that Umayyad attempts to survey land, draw up cadastres, and enforce the land tax led to dissent than their alleged impiety”; The Political and Social His- tory of Khurasan under Abbasid Rule 747-820 (Minneapolis and Chicago, 1979), 196. 136 Wadād al-Qāḍī, “Population Census and Land Surveys under the Umayyads (41-132/661-750)”, Der Islam 83/2, 2006, 341-416, 365. 137 M. Grignaschi, “La riforma tributaria di Hosro I e il feudalesimo sassanide”, in Accademia Nazionale dei Lincei. Problemi attuali di scienza e cultura. Atti del con- vegno internazionale sule tema La Persia nel medioevo (Roma, 31 marzo-5 aprile 1970) (Rome, 1971), 87-138; R.N. Frye, The History of Ancient Iran (München, 1984), 324-5; Z. Rubin, “The Reforms of Khusro Anûshirwân”, in The Byzantine and Early Islamic Near East. III: States, Resources and Armies, ed. by A. Cameron (Princeton, 1995), 227-97; U. Mårtensson, “Discourse and Historical Analysis: the case of al-Tabarî’s  M. Campopiano / Studia Islamica 107 (2012) 1-37 29 A fijixed amount, payable either in cash or in kind, was collected. This sys- tem probably made it easier to estimate the amount of revenue due to the State and would have provided a steadier source of income.138 As we know, this system was the predecessor of the Islamic ʿalā ’l-misāḥa.139 However, the possible weakening efffects did not go unnoticed by Mus- lim scholars. It is noted, for example, that aṭ-Ṭabarī. (839-923) favourably portrayed the tax reform enforced by Husraw, since cadastral assessments, a land tax based on fijixed rates and a centralisation of tax revenue would have strengthened the State.140 Since the taxation was not based on fijixed cadastral assessments, but required a calculation of the State’s share, the role of intermediaries became more important in order to deal with the difffijiculties of centralisation. The difffijiculty of estimating tax revenue was one of the reasons the Abbasid Empire had to have recourse to more tax- farming and to diffferent forms of tax revenue, which fijinally culminated in the creation of the iḳṭāʿ system under the Buwayhid. Tax-farming and iḳṭāʿ either assured a more stable income for the State, or saved the government from paying a salary to offfijicers and soldiers and hiring agents to poll taxes around the Empire. The role of the tax-farmers (ḍāmin) was extremely important: in the year 280 ah, for example, the totality of the Sawād income depended on a contract with Aḥmad Ibn Muḥammad aṭ-Ṭāʾī.141 Tax-farming was possible at diffferent levels, within the central or local government, or within smaller districts. Before the contract was signed, the sum to be handed over to the State, the general amount of the revenue and all the extra charges, was fijixed on the basis of the previous incomes from that area. The Treasury could also order their offfijicers, at the demand of the tax-farmer, to allocate some funds for important works in the farmed area History of the Messengers and the Kings”, Journal of Islamic Studies 16/3, 2005, 287-331; A. Gariboldi, Il regno di Xusraw dall’anima immortale. Riforme economiche e rivolte sociali nell’Iran Sasanide del VI secolo (Milano, 2007), 178-97. 138 M. Khan Fateh, “Taxation in Persia. A Synopsis from the Early Times to the Conquest of Mongols”, Bulletin of the School of Oriental Studies 4/4, 1928, 723-43, p. 730; A. Christensen, L’Iran sous les Sassanides, second edition (Copenhague, 1944), 124; Løkkegaard, Islamic Taxation, 110; Daniel, The Political and Social His- tory, 195. 139 Løkkegaard, Islamic Taxation, 119. 140 Mårtensson, “Discourse and Historical Analysis”, 287-331. 141  Shimizu, “Les fijinances”, 11-12, for the contract with Aḥmad Ibn Muḥammad aṭ-Ṭa⁠ʾi: The Historical Remains of Hilâl al-Sâbi. First Part of his Kitab al-Wuzara (Gotha Ms. 1756) and Fragment of his History (389-393) A.H. (B.M.Ms, add. 19360), ed. by H.F. Amedroz (Leyden, 1904), 10-1. 30 M. Campopiano / Studia Islamica 107 (2012) 1-37 (such as irrigation). However, the tax-farmers were expected to take upon themselves tasks such as ground amelioration and hydraulic works, and their activity was controlled by inspectors.142 As the contract ran for a lim- ited period, the interest of the tax-farmer lay in the opportunity to mini- mise costs (including ground amelioration costs) in order to maximise profijits. Therefore he had more interest in ‘squeezing’ money from the tax- payers than in investing in water management. The tax-farmer was forced to hand over a fijixed sum to the State even if the general revenue was smaller than the sum fijixed in the contract.143 In the case of a decline in revenue, the tax-farmer, since he could not modify the sum to be handed over to the State, would try to increase the burden on the taxpayers. To offfer the neces- sary fijinancial guarantees to the State, the tax-farmers had to be chosen from the upper classes of ʿAbbasid Society. According to al-Muḳtadir’s vizier Ibn al-Furāt, three social groups could act as tax-farmers; merchants (tādjir), trustworthy offfijicers (ʿāmil wafī) and wealthy landlords (tān ghanī).144 The role of merchants, and in particular of grain merchants, is very relevant because it clearly links tax-farming and the grain trade: tax-farmers, such as Aḥmad Ibn Muḥammad aṭ-Ṭāʾī, often abused their position in order to speculate on grain prices.145 Other possible areas exempted from tax collection rights were īghār and muḳātaʿa, districts enjoying fijiscal immunity which paid only a fijixed sum to the State, and taswīgh, an area which was enjoying full tax exemption for one year.146 According to Cahen, from these concessions under the Buway- hids the military iḳtāʿ developed, based on the institution of iḳtāʿ al-istighlāl, iḳtāʿ of usufruct.147 The tu‘ma, a life-long grant of revenues to a province or a district, in use from the early Islamic period until the beginning of the 10th century, must also be mentioned.148 The iḳtāʿ al-istighlāl was a concession of the fijiscal rights of the State over land in place of a salary. This 142 Løkkegaard, Islamic Taxation, 96. 143 Shimizu, “Les fijinances”, 12-14. 144 The Historical Remains, 71. I write the Arabic words in their singular form as in the original text. 145 Aḥmad Ibn Muḥammad aṭ-Ṭa⁠ʾī’s speculations concurred with rising prices in Baghdad in 272 A.H.; aṭ-Ṭabarī, Ta⁠ʾrīkh, III series, 2110; Shimizu, “Les fijinances”, 16. 146 146. C. Cahen, “L’évolution de l’iqtaʿ du IXe au XIIIe siècle. Contribution a une histoire comparée des societés médiévales”, Annales. Economies, Societés, Civi- lizations 8, 1953, 25-52, 28-9; Løkkegaard, Islamic Taxation, 107-8. 147 Cahen, “L’évolution”, 30-2. 148 Kennedy, The Armies, 85.  M. Campopiano / Studia Islamica 107 (2012) 1-37 31 kind of grant became an increasingly common way to pay soldiers after 946, under Muʾizz ad-Dawla.149 Some iḳtāʿ also included tax-farming assign- ments and land grants. The grantees often acquired rights to the adminis- tration of these lands, initially connected to their right to collect taxes, and often developing into a more general administrative power.150 To give an idea of how important this phenomenon was, it must be noted that in 958, at the very end of the period under consideration, Muʿizz ad-Dawla granted the revenue of the provinces of Wāsit, Basra and Ahwāz.151 The grantees often extended their land estates in the areas which were given to them as iḳtāʿ by accepting the submission of farmers who, unable to affford the pay- ment of imposts, gave their fijields to the lords in exchange for protection.152 This phenomenon resulted in the difffusion of large latifundia. A Conclusion The competition of the diffferent ruling groups to gain access to land sur- plus seems therefore to have been the ultimate reason for the decline of Iraqi agriculture. These groups fought to receive land grants and to acquire fijiscal privileges. Even when they made relevant investments in agriculture, as was the case with the difffusion of market oriented monocultures in the Basra region, they were made to a large extent through the massive impor- tation of slave labour, a social time bomb, and at the expense of smaller landlords and cultivators. The difffusion of the iḳtāʿ system was the fijinal consequence of a process in which diffferent factions or groups of the ruling classes competed to acquire control over land and its surplus: this was. for them, a reward for their political afffijiliation, for their military service, another source of revenue to integrate into their political, bureaucratic or commercial activities, or a limited resource that they could speculate on. The possibility of acquiring and keeping possession of land surplus ulti- mately depended upon their political position- their need to make land 149 S. Tsugitaka, State and Rural Society in Medieval Islam. Sultans, Muqtaʿs and Fallahun (Leiden-New York-Köln, 1997), 20-1. 150 For a more detailed classifijication, see Cahen, “L’évolution”, 28-32; Tsugitaka, State, 1-4 and 18-38. 151 Ashtor, Histoire des prix, 71-2. 152 H. Ben Abdallah, De l’iqta’ étatique à l’iqta’ militaire. Transition économique et changements sociaux à Baghdad, 247-447 de l’Hégire/861-1055 ap. J.C. (Uppsala, 1986), 110-112. 32 M. Campopiano / Studia Islamica 107 (2012) 1-37 prosper and fructify was secondary. In all probability, a stronger influence from those social groups whose life and prosperity entirely depended on the land and on its returns would have halted the process of decline in Iraqi agriculture. Appendix Prices and Wages It is extremely difffijicult to establish trends for wages and prices in early medieval Iraq, fijirstly because of the lack of sources. For example, the larg- est set of prices we have, wheat prices, only begins at the end of the 8th century. Furthermore, the sources we do have (essentially literary sources, chronicles, juridical treatises etc.) usually inform us about excep- tionally low or high prices, making it more difffijicult to compare the difffer- ent data available. However, it is not impossible to establish trends. First of all, some price data are not considered in the sources as exceptionally high or low, but as general or average prices. Furthermore, it is interesting to compare the peaks and valleys of a sequence at diffferent periods in time. If, for example, exceptional prices tend to be higher at any one time, this fact could be eventually considered as an element in evaluating price trends over time: therefore we should only compare homogeneous entities (e.g. the ‘normal’ prices among them). Another problem is related to the fact that we have recourse to some data available for Upper Mesopotamia. Even though this region is close to Iraq (and in part belongs to what is now the Republic of Iraq) and was a grain supplier for the people of Sawād,153 we cannot assume that the level of prices was always the same. The Chronicle of Zuqnīn tells us that in 766-767 25/30 djarīb of wheat were sold for one dinar. Since one djarīb in the Chronicle of Zuqnīn was ten ḳafīz, we can estimate an average price of around 0.112 dinars for 100 kg of wheat.154 The chronicler recorded that the market price of wheat for the years 768-769 was one dinar for 30-40 djarīb,155 therefore it can be calcu- lated that 100 kg of wheat were sold for 0.088 dinars. The chronicler also 153 Ashtor, Histoire des prix, 42-45 and 453-7. 154 The Chronicle of Zuqnīn, 186 and 216. Ashtor assumes 7 ḳafīz, but he was wrong in this case (Histoire des prix, 42). 155 The Chronicle of Zuqnīn, 245.  M. Campopiano / Studia Islamica 107 (2012) 1-37 33 noted that in the same year 30 ḳafīz of wheat were sold for one zūz (a silver coin, one dirham is what is meant), which would indicate a price of one dirham (around 1/12 of dinar) for 97.35 kg of wheat, and this more or less confijirms our previous calculation.156 Since it is not indicated that these prices are exceptional (perhaps just ‘moderate’), they represent a good indicator of wheat prices in the area. Reading al-Azdī, we can see that in Mosul in 791 1,200 raṭl of wheat were sold for 30 dirhams and the same quantity of barley for 20 dirhams, which equates to around 0.51 dinars per 100 kg of wheat and 0.34 dinars for the same amount of barley.157 In 815 100 kg of wheat flour in Mosul would have cost 0.51 dinars.158 It seems that the cost of wheat was by far the most important element in establishing flour prices.159 It is difffijicult to estimate to what extent the wheat price deter- mined the flour price. We know that at the end of the year 941, because of a sudden growth in grain prices, one kurr (2,925 kg) of wheat was sold for 210 dinars, and one makkūk (probably 5.625 kg) of flour for 6 dirhams, which equates to one kurr of flour being sold for around 312 dinars (estimating 1 dinar = 10 dirhams).160 In a period of scarcity the role of wheat in estab- lishing the flour price would have been overestimated, but we can never- theless reasonably assume that the grain represents around two-thirds of the flour price, giving an estimated price of around 0.34 dinars for 100 kg of wheat in Mosul in 815. At the beginning of the 10th century, Ḳudāma Ibn Djaʿfar wrote that the price of 2 kurr of mixed wheat and barley was 60 dinars.161 Since the price of wheat was almost double that of barley, and one kurr was 2,925 kg, the price of wheat at the beginning of the 10th century was approximately 1.367 dinars and the price of barley around 0.68 dinars. This fijigure also does not refer to an exceptional price. Therefore, at the beginning of the 10th cen- tury the nominal wheat price was nearly ten times higher than in middle of 156 The Chronicle of Zuqnīn, 223. 157 Yazīd Ibn Muḥammad al-Azdī, Ta⁠ʾrīkh al-Mawṣil (El Cairo, 1967), 276; Ahsan reads 40 dirham for the wheat but the source says 30; Ahsan, Social Life, 138. 158 al-Azdī, Ta⁠ʾrīkh al-Mawṣil, 340; Ahsan, Social Life, 138; see Hinz, Islamische Masse, 31. 159 Ashtor, Histoire des prix, 48; see also Id., “Prix et salaires à l’ époque mam- louke. Une étude sur l’ état économique de l’ Egypte et de la Syrie à la fijin du moyen âge”, Revue des études islamiques, 1949, 49-94, in particular 71. 160 See Ashtor, Histoire des prix, 40-41 (value of dinar) and 43-8 (prices). 161  Ḳudāma Ibn Djaʿfar, Kitāb al-kharādj (ed. M.J. De Goeje), 239 (Arabic text). 34 M. Campopiano / Studia Islamica 107 (2012) 1-37 the 8th century and the price of barley at least double by the end of the 8th century. Other sources give us exceptionally high or low prices related to particu- lar events such as wars, famines and, in the case of low prices, lack of cur- rency. The Chronicle of Zuqnīn tells us that the price of wheat rose, because of the famine, to one dinar for 7 or 8 ḳafīz in 743-4 (equal to one dinar for 22.715 or 25.96 kg); one dinar for one djarīb (34.25 kg) in 750-1 and 3.5 zūz for ḳafīz in 766-7 (0.29 dinars per 3.245 kg).162 We are told of an abrupt fall in prices around the year 772: the price of wheat was one dirham for fijive muḍḍ, which equals 0.027 dinars for 100 kg of wheat. Regarding exception- ally high prices, we must remember that aṭ-Ṭabarī writes that the price of wheat in the year 822 was around 40 dirhams for one ḳafīz. In Iraq, in the 10th century the ḳafīz equalled 48.75 kg of wheat (or half of this, since the same word was used for double the amount).163 Since one dinar was equiv- alent to about 20 dirhams, the price of 100 kg of wheat was about 4 dinars. We also know about exceptionally high prices due to speculation. Excep- tionally high prices were recorded for the year 874: 150 dinars for one kurr of wheat (around 5 dinars for 100 kg) and 120 for barley (around 4 dinars for 100 kg). Under al-Muʿtaḍid (892-902), the vizier Ibn al-Furāt sold 2 kurr of wheat and barley at 90 dinars, which was approximately 2 dinars for 100 kg of wheat and around half of this for the same quantity of barley. In 919, due to speculations connected to tax-farming, the vizier Ḥāmid Ibn al-ʿAbbās was able to acquire a large quantity of grain and the wheat and barley were sold at 55 dinars for one kurr muʿaddal.164 As the prices were too high, the government was forced into fijixing the wheat price at 50 dinars for one kurr muʿaddal, which is equal to 609.375 kg,165 meaning that the price had risen to around 9 dinars per 100 kg of wheat. Higher prices were also recorded in later periods: 120 dinars for a kurr of wheat in 935 (4 dinars per 100 kg), 130 dinars for a kurr of wheat in the summer of 941 (4.4 dinars per 100 kg), 210 dinars for a kurr of wheat in December of the same year (7.17 dinars for 100 kg) and 120 dinars for a kurr of barley, 316 or 317 for a kurr of wheat in 942 (about 10.8 dinars per 100 kg) and 120 for a kurr of barley and 25 dirhams (1.78 dinars) for one makkūk (probably 5.625 kg) in 946.166 162 The Chronicle of Zuqnīn, 167, 186 and 213. 163 Hinz, Islamische, 48-9. 164 Ashtor, Histoire des Prix, 46. 165 Hinz, Islamische Masse, 42. 166 Ashtor, Histoire des prix, 43.  M. Campopiano / Studia Islamica 107 (2012) 1-37 35 The prices of grain have to be compared, of course, with wages. Notwith- standing the scarcity of primary sources, a consistent increase in nominal wages can be noticed from the 8th to the 10th century. However, real wages did not increase and even declined. In 762-766 an unskilled worker employed on the construction of Baghdad earned between 1 ¼ and 1 7/8 dirhams per month, equal to between 0.1 and 0.156 dinars. The price of wheat at this time was 0.112 dinars per 100 kg, so the monthly wage of an ordinary worker could be converted into 89-139 kg of wheat per month. However, it is also possible that the authorities distributed grain for free.167 In the second half of the 9th century a glass craftsman could earn 1.6 or 1.8 dinars per month.168 In the fijirst half of the 10th century, a young accoun- tant was hired for 2 dinars per month (but as the source explains that he was desperate, it can be supposed that his salary was quite low for his posi- tion). Two dinars per month, in the fijirst half of the 10th century, more or less corresponds to 146 kg of wheat.169 If it is considered, as the evidence suggests, that the unskilled worker also received a contribution in kind, the second wage (that of a clerk) seems to be comparable or even lower than the fijirst one. The largest set of wage data available is for military wages. An analysis of military wages is interesting because the soldiers had considerable bargain- ing power. Military expenditure represented the bulk of the budget of the Caliphs. It can be estimated that under the governor Ziyād Ibn Abī Sufyān (665-673) 86% of the revenue from the provinces of Basra and Kūfa (which more or less covers the whole of Iraq) was devoted to military expenditure.170 When al-Muʿtadid came to power in 892, 80% of expenditure was devoted to the army.171 The army was usually paid in mint coins, but they also received land grants (ḳaṭīʿa). At the time of the Arab conquest, it is said that the soldiers who took part in the early campaigns in Iraq were paid 3,000, 2,500, or 2,000 dirhams per year, but after the battles of Ḳādisiyya and Yarmūk and the collapse of the Persian empire their wages fell to 1,000 and then to 500, 300 or even 200 dirhams per year, which equates to a little less than 50, 30 or 20 dinars.172 It can be assumed that 30 dinars per year was a 167 Ashtor, Histoire des prix, 64. 168 Ashtor, Histoire des prix, 65. 169 Ashtor, Histoire des prix, 65. 170 Kennedy, The Armies, 71. 171  Kennedy, The Armies, 156. 172 Ashtor, Histoire des prix, 61. 36 M. Campopiano / Studia Islamica 107 (2012) 1-37 common soldier’s wage after the Arab conquest of Iraq. Under the Umay- yads a soldier’s wage was usually 8 dinars per month. One ‘month’ of pay did not correspond to a common ‘month’, in this context a period of 3 or 4 months could be meant. There were also troops who earned less anyway. At the beginning of their power the Abbasids reduced the wages. As-Safffāḥ, the fijirst of the Abbasids (d. 754), paid 6.5 dinars per month (80 dirhams), and this seems to have been an exceptionally high wage, due the conquest and the large booty of the Abbasid army.173 These lower wages were prob- ably due to a greater supply of soldiers, since many non-Arabs, for example the Khurasians, were enrolled in the army. The evidence shows that under al-Ma⁠ʾmūn, the infantrymen received one dinar (20 dirhams) per month and horsemen double this.174 In the year 200 ah Abū ’s-Sarāyā paid 1,000 dirhams to the horsemen and 500 to the infantry, probably per year, which equates to 50 and 25 dinars respectively.175 At the beginning of the 10th century, under al-Muḳtadir, the wage for mounted troops was 12.5 dinars per month until 917, when it was diminished by two-thirds.176 It can be seen that there were attempts to reduce wages, but eventually they increased again. In 929 a cavalier could be paid as much as 42 dinars per month, but this seems to be an exceptionally high wage.177 Therefore, if we want to calculate real wages in terms of wheat quantity, it must be realized that in the second half of the 8th century the real wage of a horseman with a nominal wage of 6.5 dinars would have been 5803.5 kg of wheat (using the wages reported for As-Safffāh, d. 754). At the beginning of the 10th century, the real wage of mounted troops (nominal wage of 12.5 dinars) would have been around 914 kg of wheat. It is even more interesting to calculate the real prices of wheat in terms of wages: 100 kg of wheat would have been more or less the 1.723% of a monthly wage of 173 This is probably the wage of a horseman: ʿIzz al-Dīn Muḥammad Ibn al-Atīr, Tārīkh al-kāmil, 12 vol. (El Cairo, 1873), vol. V, 322. 174 ʿIzz al-Dīn Muḥammad Ibn al-Atīr, Tārīkh al-kāmil, vol. VI, 228. 175 Aḥmad Ibn Muḥammad Miskawayh, Tadjārib al-umam, in Fragmenta his- toricorum Arabicorum: et quidem pars tertia operis Kitábo al-Oyun wa al-hadáïk fiji akhbári al-hakáïk, et pars sexta operis Tadjáribo al-Omami auctore Ibn Maskowaih (Leiden, 1869-1871), 423. 176 Aḥmad Ibn Muḥammad Miskawayh, Tadjārib al-umam (The eclipse of the ʿAbbasid caliphate; original chronicles of the fourth Islamic century), edited and translated by H.F. Amedroz and D.S. Margoliouth. 7 vol. (Oxford, 1920-21), vol. I, 43 (Arabic text). 177 Ashtor, Histoire des prix, 70-1.  M. Campopiano / Studia Islamica 107 (2012) 1-37 37 a horseman, and at the beginning of the 10th century it would have been 10.9% of the monthly wage of a cavalier. Recorded land tax revenue from Iraq (in dirham, Middle of the 7th c. to ca. 918-919) Period or year Tax revenue Caliph ʿUmar ibn al-Khattāb (634-644) 116,000,000 dirhams c. Caliph Muʿāwiya (661-680) 150,000,000 dirhams c. Caliph ʿUmar ibn ʿabd al-Azīz (717-720) 112,000,000 dirhams c. 787-8 123,740,000 dirhams 819 114,457,650 dirhams. This fijigure is overestimated as it is based on prices at the beginning of the 10th century. Part of the tax amount was collected in kind. Middle 9th century 113,097,760 dirhams This fijigure is overestimated as it is based on prices at the beginning of the 10th century. Part of the taxamount was collected in kind. 918-919 27,354,100 dirhams Recorded grains prices Year Wheat Prices Barley Prices 743-4 1 dinar for 22.715/25.96 kg 750-1 1 dinar for 34.25 kg 766-7 0.112 dinars for 100 kg 766-7 0.29 dinars for 100 kg 768-9 0.088 dinars for 100 kg ca 772 0.027 dinars for 100 kg 791 0.51 dinars for 100 kg 0.34 dinars for 100 kg 815 0.34 dinars for 100 kg 822 4 dinars for 100 kg 874 5 dinars for 100 kg 4 dinars for 100 kg 892-902 2 dinars for 100 kg 1 dinar for 100 kg Beginning 10th century 1,367 dinars for 100 kg 0.68 dinars for 100 kg 919 9 dinars for 100 kg 935 4 dinars for 100 kg 941 (summer) 4.4 dinar for 100 kg 941 (December) 7.17 dinars for 100 kg 4 dinars for 100 kg 942 10.8 dinars for 100 kg 4 dinars for 100 kg 946 1.78 dinars for 5.625 kg