CHAPTKR 17 THE NATURE OF GOVERNMENT As seen in Chapter i, early Muslims tacitly assumed humans to have originated in a politically organized society based on revealed law, and to have recreated such a society whenever God sent them a messenger with a new law. In the ninth century they began to enquire into their own presuppositions. \X'hy do humans live social lives? Must their societies be based on religious law brought bv a prophet or might man-made law and morality suffice? Could one manage without a monarch? Must government be monarchic, or indeed autocratic, or could alternative forms of political organization be envisaged? Their answers mostly, though not always, endorsed the assumptions with which they had started out, but they did so with a plethora of explicit argumentation which often raised new questions and which continued beyond the period considered in this book. The debate was dominated by philosophers, mutakallims, and Shi'ite thinkers rather than by religious scholars from the Sunni camp, but the latter accepted many of the ideas it produced. Since the arguments are scattered in works of the most diverse genres and the study of them is still in its infancy, what follows is merely a preliminary survey. Why do humans live in societies? Most educated persons in the Near Hast before the rise of Islam knew the answer to this question, ultimately from Aristotle and other Greek philosophers. "Because of the arts and sciences and the useful things to which they lead, we have mutual need of one another, and because we need one another we come together in one place in large numbers," as Nemesius of Emesa (c. 400) explained in a widely read treatise on human nature (which was eventually [*59] 2ŔO ] Nature of Government translated into Arabic too); "to this human assemblage ... we give the nam city (polis), for man is a naturally sociable animal [polilikon won). N0 s person is in all ways self-sufficient."1 Most educated persons in the Near Ea gave the same answer after they had become Muslims. It is first encountered ' al-Jahiz (d. 2.55/869), the first Muslim author known to have considered the question: it is in the nature of humans that they need one another and mvj cooperate to survive, he says, adding "God, exalted is He, has not created anyone who is able to fulfil his need on his own."2 The Muslims could soon read the same in Arabic translations of Greek works, which they rapidly took to citing.3 Humans must come together to cooperate, especially as they also need to associate in order to live virtuous lives, al-'Amin (d. 381/992) declared paraphrasing Aristotle; exceptions would be in the nature of gods or beasts'1 "For this reason it is said that man is sociable by nature {al-insan madam bi'l-tab'),'' Raghib al-Isfahani (c. 400/1010) observed.- The same argument was used in varying versions, now with and now without the Aristotelian tag that man is a social animal {hayawan insi'madam), by the philosophers al-RazT (d. prob. 313/92.5),6 al-Farabi (d. 339/950),7 Ibn STna (d. 428/1037),8 and Ibn Rushd (d. 595/1198),* the secretary Qudama b. Ja'far (d. prob. 337/948),10 the polymath al-BTrunl (d. 440/1048),11 the religious scholar Fakhr al-DTn al-Raz] (d. 606/1209),12 the Shi'ite Mu'tazilite Ibn Abl '1-HadTd (d. prob. 655/1257),« the Ismaili philosopher Nasir al-DTn TusT (d. 672/1274;14 and, aftet our period, by authors as diverse as Ibn Taymiyya (d. 728/I32.8)15 and Ibn Khaldun (d. 1. Nemesius, §52 = 243. 2. Hayawan, i, 42ft. 3. Plessner, Bryson, Arabic text, 1461.; Thermsrius, Risala, zftf./qo; cited in Ibn Abl '1-Rabl', Suluk, 781. Cf. also Plato, Republic, 309b!, and the Aristotelian passages in the next note. 4. cAmirT, Sa'ada, t^o (cf. Aristotle, Politics, 1253a, 1278b). For the question whether 'AmirT is really the author of this work, see the bibliography s.v: 5. Dhari'a, 374. Adam had to perform a thousand tasks to eat bread, as Nasir al-DTn TusT says [Ethics, 189). 6. Razi, al-Tibb al-ruhani, 1056 = 88f. (ch. 17). 7. FarabT, MS, eh. 15, Jij cf. Tahsil, §16 = §18. 8. Ibn STna, SI, x, 441 = 99. 9. Ibn Rushd, Commentary, 22 = 113. 10. Qudama, Kharaj, 432!'.; cf. also MawardI, Adah al-dunya, 135!. Tashil al-nazar, 97. 11. Jawahir, 6i. = 6. His deathdate is placed larer by some. 12. RazT, aPMabahith al-mashriqiyya, ii, 523; cf. also Lambton, State and Society, 13MI Fouchecour, Moralia, 426, on his Jam? al-ulum, ch. 56. 13. IAH, Sharh, xvii, 491. 14. Ethics, i89f. 15. Al-Hisba fi 'l-islam, 4 = 20. Nature of Government [ 261 !oS/i40^^'16 ^ was a'so a P°Pu'ar argument in medieval Europe. It does not ppress a modern observer, for occupational specialization comes late in the development of civilization: simple societies have little or no division of labour, except by age and sex. Producing a loaf of bread is indeed beyond the ability 0f a single person wholly on his own, as the literature often says, but it is not |,oyond the capacity of a household, and there the division of labour may stop. £ven Ibn Khaldun, who had a strong sense of the differences between simple (tribal) and complex (urban) societies in North Aftica, overlooked this objection! however, and the traditional explanation had the merit of stressing the co0perative nature of society. Raghib al-Isfahani interspersed his account of the division of labour with Prophetic dicta on the solidarity of the believers. Why must human societies have laiv? Though the need for cooperation meant that humans had to live in societies, their nature was such that that they could not do so without some kind of restraint. God had imprinted a desire for good things on all humans, and indeed animals, an epistle credited to al-Jahiz says; He has implanted a desire for self-preservation in people, as al-Maturidl (d. 333/944) put it.17 Humans were competitive, brutish, swayed by strong desires, and avaricious too: unlike animals, they were for ever hoarding things they did not need. Left to their own devices they would ruthlessly pursue their own interests and diverse passions, engage in constant rivalry and strife without affection or altruism, ignorant of their true interests in this world and the next, and thus bring about their own ruin.18 "There is nothing as social by nature and as anti-social by corruption as the human race," as St Augustine had put it.19 This was the Muslim view entirely, except that they did not usually credit the anti-social streak to later corruption. But whether humans had been created deficient or were corrupted at some later stage (by Adam's fall,20 Cain's murder of Abel,21 or the like), 16. Muqaddima, 46t = 89fr. 17. 'Al-Ma'äsh wa'l-ma'äď in his Rasä'il, i, io2f.; MäturldT, Tawbid, 177. 18. In addition to the references in the previous note, see the ZaydT view in Jähiz, 'Maqälat al-Zaydiyya wa'1-Räfida' (also known as 'Bayän madhähib al-shTV) in his Rasä'il, iv, 3i8f.; al-Qäsim b. Ibrahim in Madelung, Qäsim, 14; Abrahámov, 'Käsim', 85; Abu Flätim al-Räzi, AHäm, rrof.; Juwaynl, Gbiyäth, §19; Ghazäll, Iqtisäd, 236; TurtushT, Siräj, 4lf. 19. In Markus, Saeculum, 95. 20. R/5, iv, 166; cf. ii, 21; iv, 18. 21. Thus ßlrünl, who also debits the envy and rivalry to the mixture of contradictory elements in human nature {Jawähir, 6i, 24 — 6, z6). 2(5z ] Nature of Government practically all medieval Muslims envisaged life in what Westerners state of nature as nasty, brutish, and short.22 The idea of human life in its unmodified form as lacking in morality c operation, and social cohesion had long roots in the ancient world. It had how ever once been balanced by a countervailing belief to the effect that the origin state of mankind was a golden age of freedom and innocence which had lasted until avarice caused coercion and inequality to appear. This view, promoted above all by the Stoics, fused with the biblical story of the Fall and went on to a long career in European political thought, where it placed a question mark over the necessity of states.21 In the hugely popular Roman de la Rose (c. ii70) for example, Jean de Meun tells of how people had once lived in mutual amity, without agriculture, work, private property, rulers, marriage, or other restraints on free love until a host of vices, including covetousncss, appeared' this was when human nature turned so nasty that a ruler had to be elected for the maintenance of order.2* But though a strikingly similar concept of a golden age was current in India, it is not clear that there was an Iranian version as well,25 and in its Greek form, the myth seems to have lost prominence in the eastern Mediterranean before the rise of Islam. Al-BIrunl, it is true, knew it from the astronomical poem of Aratus of Soli (d. 240 bc) and compared it with the Indian version of the myth; but he wrote as a scholar, not as somebody to whom the myth was alive.26 As live tradition it may be reflected in the Brethren of Purity, who tell a story of how humans lived carefree lives in mutual love like a single family until they started hoarding; but this story is meant to illustrate the divine world from which we have fallen (thereby becoming human), not a primitive stage of human history that we have lost, that is, its import is Neo-platonist or Gnostic, not socio-political.27 Elsewhere, too, there are suggestions 22. The parallel with Hobbes's Leviathan is drawn in Pines, 'La loi naturelle et la societě', 167; implicitly, also in Al-Azmeh, Muslim Kingship, There is no Arabic terra for 'the state of nature'. It is evoked with statements like 'if people were left on their own' (i.e. without divine intervention), or 'If God left them alone with their natures'. 23. For all this, see the references given in Crone, 'Ninth-Century Muslim Anarchists', 6-8. 24. Cf. Milan, 'The Golden Age and the Political Theory of Jean de Meun'; George, 'Jean de Meung and the Myth of the Golden Age'; also Cohn, Pursuit of the Millenium, T95f. 25. Cf. Crone, 'Zoroastrian Communism', 459. 26. BTrum, Hind. T92.r2 = 383 (ch. 43); cf. Sachau's comments ad. loc., and Lovejoy and Boas, Primitivism and Related Ideas in Antiquity, 34ft". (Aratus), 433ft. (Indians). 27. RIS, iv, 37f.; cf. ii, 326 = Goodman, Case, 157, where the animals pick out human avarice and hoarding of superfluous things as one reason why humans need religious laws; Abú Harim al-RázI, A'lam, 188, where the contrast between animals and humans in this Nature of Government 263 that the Brethren of Purity and other early Ismailis thought in terms of an aboriginal state of human innocence, especially when they speak of law and government as punishment, 28 but they never explicitly state that mankind had once been free of their many vices, and they do not focus on this question in what survives of their debate regarding the existence or otherwise of law in Adam's era.2' If the concept was there, it certainly was not prominent. In short, the Muslim view of aboriginal life was overwhelmingly that of Hobbes. without any admixture of positive views of the state of nature. Unlike Hobbes and other contract theorists, however, the Muslims usually saw the state of nature as having come to an end thanks to divine intervention rather than human action: God in His mercy sent a Prophet with a law, to found a polity. Differently put, the social contract was with God, not with a human being. Without God's law, there could be no civilization, indeed humans would not survive at all, as ShI'ttcs above all were prone to claiming.30 Why must the law be God-given? Hobbes took it for granted that humans can devise moral and legal codes on their own. How then did the Muslims explain their conviction that God had to send a prophet with a law in order for social order to appear? There were two answers to that question. No natural law The first explanation was that no human had the right to impose obligations on other people. Nobody was authorized to tell anyone else what to do or what to believe, be he a ruler, parent, husband, schoolmaster, or even a prophet: God was the only source of legal/moral obligations; before revelation, humans had lived in a state of fundamental non-obligation (bara'a asliyya)?1 Nobody had respect is one reason why humans must have imams. Compare the role of Az, covctousness, in Mazdak's vision of the abolition of private property and pair-bonding (Crone, 'Kavad's Heresy,' 28; add LirdawsT's presentation in Davis, Lpic and Sedition, 30). I 28. Cf. above, 200, 20T; below, 356, on the animal fable. ft 29. Cf. Madelung, 'Imamat', 102-4, ro6-8, where al-Nasaf! docs broach the question, [ but the argument does not make sense. 30. The reason why God had made humans dependent on prophets instead of implanting a religious instinct in them (contrary to Abu Bakr al-RazT's claims) is that self-sufficient humans would fight one another to death; "if it were not for religion and the laws of the Prophets .. . people would perish" (Abu Harim, A'lam, 190). Similarly Jahiz, 'Ma'ash wa'l- [ ma'ad', ro3t. 31. Thus GhazalT in Laoust, Politique de Gazali, i53f. (citing the Mustasfa). 264 \ Nature of Government Nature of Government the right to inflict punishments on anyone else either; only God could do so.a For only God could determine what was right and wrong or, as medieval Muslims preferred to put it, what was permitted and forbidden. Without revelation, humans would not have any morality or law at all. The expression 'before revelation' (qabla wurud al-shai*) with which people examined the nature of obligations did not refer to a historical stage. Kotj strictly speaking, does the Western concept of 'the state of nature', though there has been a strong tendency to envisage it as such. Both expressions stand for thought experiments in which human nature is imagined in the raw, stripped of divine guidance in the one case, of civilization in the other. The Muslim thought experiment focused on the moral status of human acts whereas the Greek experiment focused on the moral nature of humans themselves, but the issue was the same: how far was morality natural, how far conventional? Like other Ash'arites, al-Ghazall subscribed to the view that it was entirely conventional, in the sense of established by God: no human act had any moral value in itself; all acts were good or bad only because God had defined them so to us. It followed that humans could not have an inner moral compass, or any "law written in their hearts" (Rom. 2:15), enabling them to live moral lives on the basis of their own unaided reason. All morality took the form of positive law enacted by God.33 Humans might still be able to devise rules of their own (the possibility is not discussed in the context of this debate, which was not directly concerned with socio-political organization); but such rules would not be moral, nor would they lead to otherworldly salvation. According to this view of things, in short, it was only by divine intervention that humans could escape from their amoral state of nature. "For when the religion of God was taken away, they lost also the knowledge of good and evil," as the church father Lactantius (d. c. 32.0) says in his account of how the golden age (here envisaged as an age of perfect monotheism) came to an end. "Thus community-living perished among men, and the compact of human society was broken."34 This was how the causal connection between God, morality/law and human society was envisaged in Islam as well. Human abilities insufficient The second explanation was a modified version of the first, reshaped as J direct answer to the question why societies had to be based on laws brought by prophets. It did not claim that human actions have no moral value in them- 31. 'Abd al-Jabbar, Mugbni, xx/z, 152.13- 33. Cf. Reinhart, Before Revelation, 70ft'., et passim. 34. Lactantius, Divine Institutes, v, 5. 265 pelves: certain things were indeed intrinsically good or bad. But humans did 1 not know what was best for them or how to achieve it, according to this argument. Only God could supply the wisdom, the authority, and the sanctions required for a community based on true morality. As regards wisdom, people were not intelligent enough to know what was test for them in this world, let alone the next. They needed a superior intelligence to tell them. It was thanks to divine intervention that they had acquired their knowledge of right and wrong, and also of many other things: how would they have worked out the difference between edible and poisonous substances, for example, without perishing in the attempt? How could they have lenovvn about agriculture? Prophets were not just founders of polities, but also culture heroes. I As regards authority, people were too similar to submit to one another. I They were always competing, for ever thinking that they had a better right to power and wealth than anyone else. They needed a superior authority to defer to. It was thanks to divine intervention that some were raised above others, boosted with divine authority, as everyone knew from the story of the rise of Islam: it was by divine intervention that government had been created for the egalitarian tribesmen of Muhammad's Arabia.35 I Finally, as regards sanctions, humans needed a strong incentive to submit, and again, the solution lay with God, who had instituted otherworldly rewards and punishments."' "We know that people cannot defend themselves against their own natures or act contrary to their desires except by a strong deterrent, the long-term threat of eternal punishment over and on top of punishment in ! the here and now," al-Jahiz has the Shakes say.37 God had instituted Paradise and Hell as the carrot and the stick (al-targhib wa'l- al-tarhtb), and without them society would go to rack and ruin, he (or somebody mistaken for him) says in another epistle.-'8 Ibn Sina agreed, with implicit reference to the phases of Muhammad's career: the lawgiver prophet should start by telling the masses that they had a maker who had prepared eternal rewards and punishments for them; this would make them obey. Next the lawgiver should institute acts of worship that would constantly remind them of this; and finally he should regulate the social and political aspects of his polity.39 Religion diverted people I 35. Cf. BTrunT, below, 285. »36.Cf. Jahiz, 'Maqalat al-Zaydiyya', 3TStf. and 'al-Ma'ash wa'l-ma'ad', ioiff.; MaturidT, Taivhid. 177ft; MawardI, A'lam al-nubuwwti, 49; Ibn Sina, 57, x, 441 = 91; Fakhr ■ il-Dln al-RazT, Muhassal, 176. I 37- 'Maqalat al-Zaydiyya', 320. 38. Jahiz, 'al-Ma'ash wa'l-ma'ad', 104. 39- SI, x, 44iff = looff. z66 Nature of Government from their desires and turned their hearts away from their selfish wishes "untli it comes to dominate their innermost soul and exercise restraint on their con science, supervising their soul in its inner solitude, and giving it sincere advi-in its misfortunes", al-Mawardl said in a good description of internalization * In short, religion enabled people to suppress selfish inclinations incompatiy with communal existence, as Dürkheim was to say centuries later in the West But whereas religion has formed part of all human societies, prophets have not, and medieval Muslims were well aware of this fact. How then had soci eties without prophets come to live by what the societies in question consid ered to be right and wrong? One might have expected the answer that they too owed them to belief in the supernatural, for people obviously did not need prophets in order to believe in gods, cosmic order, or an afterlife. But this would have implied that any religion, even paganism, could supply social and political organization, however misguided it might be. The philosophers did in fact hold this to be the case, and al-Blmni seems to have thought so, too.41 But it was not wise to air this opinion openly; and besides, many Muslims, like many Christians, had trouble seeing paganism as a religion at all. When they declared all kingdoms to be based on religion, the reference was to monotheist religion.42 Insofar as pagans were perceived to have laws, they were assumed to have devised them of their own, by conscious legislation rather than imperceptible development. On the rare occasions on which the problem of non-prophetic laws was openly confronted, the alternative to revelation was human reason, whether in the form of common sense or philosophy, not religion of other types. Human and divine wisdom compared Thus al-Jahiz tells a story in which a commander of the Umayyad period meets with the Turkish kbaqan to compare their respective laws, to find that they agree on some things and disagree on others. "You are a people who trace your laws to what reason permits and what seems to be a good idea," the commander concludes, "but we think that we are not fit to manage the servants (of 40. Adah al-dunya, 136. 41.lt is implied by al-FarabT's theories {above, 173) and explicitly stated by the philosopher in Judah Halevi, Radd, 5 = 34 (I, §1), while BlrunT had no trouble seeing the religious basis of socio-political organization in India in his Hind (esp. chs 9-10); and he dispassionately notes the similarity between Muslim, Zoroastrian, and pagan Tibetan and Turkish ways of invoking religion to single out the ruler in his jawahir (24f. = 2.6, cf. below, note 128]. 4z. Thus KM (M), 67; cf. Abu Ya'qub al-SijtsranT, above, 13, note zo. Nature of Government 267 God)." Only God knew the true nature of things that was hidden to people.43 \1-Jahiz' intention here is not to idealize the Turks as people who followed reason of their own accord, without the need for punishment, for the Turkish laws discussed include the penalties for theft, murder, and cutting off people's ears and noses. Nor is he trying the opposite, to depict them as amoral, for as a Uu'tazilite, he took it for granted that reason could provide moral guidance, and his commander speaks with respect of the Turkish laws. Al-Jahiz' position is simply that revealed law is better, not because otherworldly sanctions had a good effect on social order (which is not discussed) or even because revelation provided for otherworldly salvation (the discussion is only about this world), but rather because revealed law was rooted in supernatural wisdom: its institutions were intrinsically better in his view. But he evidently did not assume revelation to be the only option. Humans could devise their own rules on the basis of common sense. Similarly, philosophically inclined Iranians held their ancestors to have instituted kingship by deducing its necessity from the anti-social nature of human beings and enthroning Gayomard (the first, or almost first, man on earth): government here owes its existence to a social contract of the Hobbesian type, and the result is a virtuous polity based on philosophy.44 If people would follow rational laws, they would not need prophets to bring them law, as Aristotle's father was supposed to have said.45 We do not know whether Ibn al-Rawandf, al-SarakhsI, al-RazI and others who rejected prophethood considered the socio-political implications of their own views; but if they did, they must similarly have held that society could be based on human reason, or indeed that it was so based. Nonetheless, the assumption that only a prophet could bring a law suitable for the organization of a society was rarely challenged in our period. People only saw the cases which confirmed it. Thus it was noted that several African peoples lacked both religion and socio-political organization, whereas the Nubians and the Abyssians had something in the nature of both, which fitted the theory in that they were Christians thanks to their former proximity to Byzantium.46 It was not noted as a problem that India and China had complex societies of the most sophisticated kind, which did not fit the theory in that no 43. 'Manaqib al-turk' in his Rasifil, i, Sof.; tr. Walker, 692; cited in Pines, 'La loi [ naturelle et la societe', i85n. 44. MM, ii, io6f (i, §551). I 45. Pines, 'La loi naturalle et la societe', 184m, citing cÄmirT, Scfäda, 178.8. 46. Cf. Istakhrl, Masälik, 4, penult, (drawn to my attention by Adam Silverstein). 268 Nature of Government prophets had been sent to either them or their neighbours.47 One has to the way to Ibn Taymiyya for a straightforward statement that polities di< have to be based on religion,48 and all the way to Ibn Khaldun for a refutaa of the view that they did have to be thus based. Most people had acquired " crnment without receiving either prophets or books, according to Ibn Khald" which showed that laws suppressing selfish inclinations could be devised b' reason and imposed by force. Such rational governance [siyasa caqliyya) COu|j be cither of the philosophical type which had been practised by the ancient Iranians (i.e. under Gayomard), but which was not otherwise encountered in history, or of the selfish type practised by all rulers nowadays whether the were Muslims or not. God had however made manmade law unnecessary f0 the Muslims by means of that which the Prophet had instituted and the caliphs had upheld, and this was preferable because a divine law served both this world and the next.49 The type actually practised was simply a perversion. In short Ibn Khaldun agreed with al-Jahiz: revealed law was not indispensable, but it was certainly better. Why does the law necessitate rulers? Granted that human society owed its existence to the division of labour and that it had to be regulated by a revealed law, why did there have to be rulers; Could one not live by the law alone? One would have expected this to be a much debated question, given that the Muslims traced their spiritual ancestry to stateless Arabia. Ancient observers had commented on the political freedom of the northern Arabs, now in a negative and now in a positive vein; the Arabs had boasted about it in their poetry themselves; and the specialists in Arabian antiquities who collected their poetry along with their stories about tribal wars (nyyam al-arab) were clearly impressed by it too: before the rise of Islam the northern Arabs (Mudar, Nizar), and above all the Prophet's tribe Quraysh, had been laqah, free people who did not obey any kings or pay any taxes, they tell us with pride.j0 But as seen already, neither the pre-Islamic Arabs nor the tribesmen who continued to inhabit the peninsula after the conquests were regarded as a model of inspiration or imitation for Muslims as far as political organization was concerned.51 The scholars who did field work in the desert 47. Cf. the striking example of Sijistäm above, 13, note zo. 48. Al-Hisba ft 'l-isläm, 4 — 2.0: people without divine books or religion obey their kings in matters rhey think will serve their worldly interests. 49. Mnqaddima, 48, zizf., Z4if. = i, 93, jS^f., 448f.; cf. above, note 44, and ch. 1,13t 50. Cf. the attestations in "Wörterbuch, s.v; 'laqäh'. 51. Cf. above, 7, 68. Nature of Government 269 re philologists eager to record the language of the bedouin, assumed to be j,e purest version of Arabic, not utopianists seeking inspiration in tribal ofganization. Pious people without an interest in pre-Islamic lore assumed the -Islamic Arabs to have lived in a state of ignorant barbarism (jakiliyya), fighting each other and generally exemplifying the anti-social side of human beings, except insofar as the QurashI guardians of God's house had preserved joine monotheism among them; and the later bedouin were effectively living in the Jahihyya too, as people who crossed paths with them knew all too well. For political models one looked to Medina, where the Prophet had worked.52 This is not to say, of course, that the tribal tradition contributed nothing to Islamic political thought. On the contrary, as seen already, it dominated the thinking of the first two centuries and is discernible behind most libertarian and communitarian thinking in classical Islam. But it owes its overriding importance to the facts that it was islamized in Medina and that the conquerors were tribesmen whose values went into the foundations of Islamic culture, not to a memory of tribal organization as a model of admiration and imitation in its own right. A fourteenth-century work does preserve an undatable argument in favour of doing without rulers in which the bedouin are invoked as an example, but this is very unusual." Since the Muslims did not have a notion of an aboriginal state of freedom and innocence, they were not inclined to credit members of simple societies with the preservation of virtues they had lost, after the fashion of the Greeks, whose fascination with Scythians and other tribal peoples (Arabs included) did not reach them; and the many tribal peoples they encountered in the course of their conquests did not strike them as any better than their own. Ibn Khaldun did admire the Turks for their preservation of the martial values once possessed by the Arabs, but the Turks in question were Muslims serving as soldiers in the Middle East, not tribesmen back in their pagan arcadia. The Persian tradition did say that the earth had been devoid of kings, whether Persian or other, every now and again from the death of Gayomard onwards, but it did not say so in an anarchist vein.-4 Had one asked how people coped in the periods without kings, the answer would probably have been amruhum shurd, that is, that they settled their affairs by-consultation, a Qur'anic expression which came to stand for anarchy in the sense of chaotic conditions.55 In short, there was no question mark over the need for rulers in the Islamic tradition, the Kharijite and Mu'tazilite anarchists notwithstanding. The 5z. Cf. below, 3i8f. 53.IJ1, Mawäqif, viii, 347. 54. Haniza, Itfrlkh, 14t. (G, 10}. 55. E.g. GhazälT, Fcicfa'ib, 106.ult. 270 Nature of Government normal answer ro the question whether one could live without rulers w as that one could not, and this seemed so obvious that many argued directly from th anti-social nature of humans to the need for rulers, without first explains how social life generates a need for law. For those who wanted the full 3^ ment, however, the explanation was that bringing together the diverse natures and ambitions of humans in a single society created a need for rules, which i their turn created a need for somebody to uphold the rules. There had to be ruler to apply the law, to judge, and to maintain order with the sword, which induced fear and deterred the wrongdoer, for people would not obey the law of their own accord: just as somebody had to teach them the law, so somebody had to reward and punish them for their obedience or disobedience to it, in this world as well as the next.* One needed both religion and government for an orderly society, then-without religious injunctions to obey, the ruler's authority would be weak; and without a ruler to enforce the laws of the religion, people would abandon them; this was why ArdashTr said that religion and government were twins.'"" The ruler on his own was better placed to enforce the law than either religion or reason on their own because selfish desires tend to overwhelm moral intentions, al-Mawardf noted; "God restrains (people) more through the sultan than through the Qur'an," as the Prophet had said.58 Without a king, sultan, or imam, people became disorganized, power passed to whoever was in a position to take it, chaos prevailed, trust disappeared, and the community disintegrated, just as flocks perish without a shepherd.3' "Civil strife results when there is no imam to take charge of people.'"50 Moreover, polirical leadership was natural, as was clear from the subordination of the body to the soul, of women to men, of slaves to the free, and of children to adults, al-'Amir! said, crediting his views to Aristotle.61 Al-Jahiz also found it natural: even animals have leaders that they follow.62 By contrast, the Ismaili Abu Ya'qub al-SijistanT held political leadership to be unique to humans because he equated it with 56. E.g. Qudäma, Kharäj, 436; <ÄmirI, Sacäda, 179, 185.z, i86f., citing Plato, Anstode, and Ps.-Aristotle's Fi siyksat at-mudun, §3.1; GhazälT, Iqúsäd, 2361.; ShafirastänT, Nihäya, 490 - 155; Abrahámov, 'al-Käsim ibn IbrähTm', 86. 57. cA/jĺ/ Ardashir, §4 (p. 53); cited in countless works, e.g. Qudäma, Kharäj, 436; Ibn Wahb, Burhän, 401. 58. MäwardT, Adab al-dunyä, 137.6. The saying (not always attributed to the Prophet) is also adduced in Ibn (Abd Rabbih,f Iqd, i, 7.6; Qudäma, Kharäj, 440.13; Juwayní, Ghiyälb, §19; GhazälT, Qistäs, 90.-5; NaysäburT, Imätna, 86, and no doubt elsewhere. 59. Jähiz, 'al-Nisä" in his Rasä'tl, lii, 149-51. 60. Ibn Hanbal in Abfl Yalä, Abkäm, 19. 61. Sa'äda, 187í. 62. 'Al-Nisä3', 150; cf. Dio Chrysostom, 'Peri basilcias' (third oration), §50. Nature of Government [ 271 r£Jigious leadership, which animals did not have: unlike humans, they were f equal because they all had the same instinctive knowledge of the basic things that animals needed to know.63 How do we know rulers to be prescribed by the law? All these rational considerations apart, one could of course settle the question bv an appeal to authority: the law made it obligatory for Muslims to have an imam, as everyone except for the Najdiyya and the Mufazilitc anarchists agreed. But appeals to authority merely took you back to reason, for how was the imamate known to be a legal duty? I Many Mu'tazilites said that it was known from reason {'aql), meaning from considerations of the kind just reviewed, or from both reason and supra-rational authority [sbaf, sanf, also translated 'revelation').64 But according to the Traditionalists and the classical Sunn is, and many Mu'tazilites too, the obligation rested exclusively on supra-rational authority.61 By this they did not usually mean that the obligation was grounded in the Qur'an or Hadith (though some found evidence for it there as well), but rather that it rested on ijmaf, the consensus of the community, starting with the agreement of the Companions to have the institution.66 This may sound like a retreat from reason, and so it was, in the sense that there conies a point where particular ways of doing things can no longer be explained in terms of universal rationality: secularists will then shrug their shoulders and say that this is how we happen to do things, while believers will point to their books or sacred persons and say that this is how God happens to have instructed us. Revelation typically works to justify the particular, as Ismailis, philosophers, Sufis, and others who distinguished between organized religion and the universal truth above it all had occasion to note. Reason could demonstrate that humans needed government of some kind or another; it could not demonstrate that they needed it in the particular form of the imamate: it was only on the basis of supra-rational authority that the specific form of government enjoined on the Muslims was known to be 63.it/jfcar, 174; Abu Hätim, A'läm, 185. Cf. also below, 336t 64. Cf. IAH, ii, 308, on Jähiz, the Baghdadis, and Abu 'l-Husayn (reason); Madelung, Qäsim, 143; IjT, Mawäqif, viii, 345 (reason and revelation). 65. Thus 'Abd al-Jabbär, Mugbni, xx/t, 17fr., 41, and the Basrans in IAH, ii, 308. It is identified as the position of most Mu'tazilites in RäzT, Muhassal, 176.9 and Arba'Tn, 426. 66. Cf. Juwaynl, Ghiyäth, §§17-18, and again §66 (in the context of election vs. designation), cf. Nagel, Festung, 298f.; Shahrastäni, Nihäya, 47SÜ. =i5of. In MäwardT, Ankam, 4 (5) — 5, those who ground the obligation in revelation adduce Q. 4:59 and a quietist tradition. 272 ] Nature of Government indispensable for human welfare, as Ibn Khaldün said, summarizing earlie arguments.''7 On these arguments, humans could see government to be in their interes God could see further and added instructions that they could not h- ' Mature of Government [*73 but ave worked out on their own. This line of reasoning rested on the assumption that there was a basic congruence between human needs and the ultimate nature of reality. The Greek philosophers, firm believers in that congruence, had called it providence. But what basis could there possibly be for such wishful thinking? On purely rational grounds we might well infer that God wished to destroy His creatures, as al-Juwaynl said; there was no way in which we could second-guess His views. Since He allowed the world to be without prophets at times it would have been reasonable to infer that He also permitted people to go without imams at times, but He did not. All this went to show that it was only on the basis of revelation that we knew the imamate to form part of God's law."1 In Abu Ya'la's formulation it had to be on the basis of revelation that one knew it, for one could not know whether anything was obligatory, indifferent permitted, or forbidden on the basis of reason at all.69 Al-JuwaynT's outlook was in line with the worldview of ancient Mesopotamia, which confronted the moral arbitrariness of the universe with extraordinary openness.70 But this was much too austere for the RafidTs. The Imamis accepted the providential nature of ultimate reality and claimed the imamate to be indispensable in terms of reason and revelation alike. This was also the Ismaili position, except that they put it in even more extravagant terms: the imamate was obligatory by nature, reason, considerations of governance (siydsa), revelation and custom, in every religion and community,71 Why must government be monarchic? Granted that we must have rulers, why could there only be one ruler at a timer Could one not have several, be it in the form of joint rulers forming a council, 67. Ibn Khaldün, Muqaddtma, 212. = i, 389?.; cf. above, note 49; MäwardT, Adah a\-dunyä, 138 and Ahkätn, 3JE./5 = 3, where reason does not even demonstrate the need for government, only for fairness and justice in mutual dealings. S8. JuwaynT, Ghiyäth, §§2.0-5; cf- Nagel, Festung, 2.97. 69. Abkam, 19. Cf. Reinhart, Before Revelation, 24 (where the same position is reported for his MuUamad), 33ff., on his ambivalent 'Udda. 70. Cf. Reaulieu, 'Theodicy, Theology, Philosophy: Mesopotamia'. It was not only in Juwaynl that it lived on; cf. Cook on the bleak conception of the relationship between man and God in Islam [Muhammad, 83;. 71. NaysaburT, Imäma, 28f. 0r semi-autonomous rulers forming a federation, or even wholly independent ulers dividing the Islamic world between them? Some Vlu'ta/ilites considered the first two options and endorsed them, as has been seen.72 But the vast majority of scholars rejected all three options without much attempt to distinguish between them. There was nothing wrong with the idea of several rulers from a rational point of view, :Abd al-Jabbar said, but authoritative instruction [sam') s against it.73 Some adduced the Qur'anic statement that heaven and earth would go to ruin if there were several deities (2.1:2.2.): it stood to reason that a plurality of human rulers would have a similarly dire effect.74 Others marshaled Hadlth: the Prophet was on record as having said, "when allegiance is given to two commanders, kill the second";75 and when the Prophet died, the Companions rejected the Ansar's proposal that they and the Muhajirun should have a leader each. 6 "Two amirs: the people have perished!" as Ibn 'Abbas reputedly exclaimed on hearing of the appointment of two leaders by the rebellious Medinese in 63'682,f.7' One leader would be able to disobey the other;"8 there would be rivalry and strife between them, even having several viziers was dangerous: too many cooks spoil the broth or, as the Arabic version of the proverb went, too many sailors caused the ship to sink.79 For all that, there were some who disagreed. Several imams Al-Asamm's idea of a federation was never fielded again, but a plurality of caliphs, each fully autonomous in his own sphere, was occasionally deemed acceptable. In 929 the amir of al-Andalus, cAbd al-Rahman III, declared himself to the caliph. Absurd though it must have sounded to most Muslims in the east, many Ash'antes were ready to accept him as such, presumably as an antidote to the Fatimids, who also claimed the caliphate. They proposed that there could be several imams if their domains were separated by a barrier obstructing \ 72.. Cf. above, 6H. I 73,'Abd al-Jabbar, Mughnl, xx/i, 243. 74. Thus Qudama, Kbaraj, 437; also cited in Abu TFawaris, Imama, ch. 8. 75. Ibn Hazm, Fast, iv, 88.12; cf. Kashif, Siyar, ii, 266. 76. Adduced in Jahiz, 'Jawabat', in his Rasa'il, iv, 29of.; 'Abd al-Jabbar, Mughm, xx/i, 244. 77. Khalifa, la'rikh, 290.5. 78. 'Abd al-Jabbar, Mughm, xx/i, 244.8. 79. Jahiz, 'Jawabat', 204f.; 'Nisa>\ iii, 149, 151 (nos. ic, tt); Ps.-Tha'alibT, Tuhfat al-wuzara', 53. ^74 . Nature of Government cooperation between them, such as for example the sea.80 If there could be m prophets in the same community, as there was in the time of Moses and Aaron or even three, as hinted in Q. 36:14, a fortiori there could be two or more imams, they said.81 This view unsurprisingly proved popular with Spanish and Maghrib! scholars,82 though the Spanish Ibn Hazm affirmed the classical posi tion,83 and it gained sufficient currency for the Fatimid missionary Abu '1 Fawaris (d. 411/1021) to find it necessary to explain why it was not allowed to have two or more imams when distances required it.84 Most Sunnis rejected it too.85 Of those who accepted it, some were outraged when the Karratniyya accepted cAlT and Mu'awiya as equally legitimate imams in their separate domains. 'All had been imam in accordance with the sunna and Mu'awiya imam 111 violation of it {'aid khilaf al-sunna), the Karramiyya said, and the followers of each had been obliged to obey. Al-Baghdadf marvels at an alleged duty to obey ft khilaf al-sunna, though the Sunnis operated with just such a duty themselves: even al-Ma'mun had to be obeyed, as Ibn Hanbal had said But al-Ma'mun had just been a quasi-caliph whereas cAlI was a real imam someone whose acts embodied God's law so that everyone had to follow him.86 There could perhaps be two quasi-caliphs at the same time, but real caliphs could not coexist, be it with each other or with quasi-caliphs.8' A plurality of kings and sultans was a different matter. This was what the Muslims had come to have in actual fact, but it was the leadership of the religious community, the church, that preoccupied them. The issue was whether more than one leader of this community (more than one pope, in the terminology of medieval Europe) could be acknowledged under exceptional circumstances in which some believers were isolated from the rest - the question 80. Baghdad!, Usui, 274 and Farq, 341; Ash'ari and Isfara'ml m JuwaynT, Ghiyath, §257; Juwayn! himself favoured this view in his Irshad, 425, but not in his Ghtyath, itfii (summarized in Hallaq, 'Caliphs, Jurists and the Saljiiqs,' 35), 81. Simnani, Kaivda, i, §§s;6f. (IsfanrinT and some Shafi'ites); for Juwaym's view that one cannot argue from prophets to imams, see above, note 68. 82. Ibn Khaldun, Muqaddima, ed. Quatremere, i, 348 = i, 393. The section is missing in the Beirut edition. 83. Muhalla, ix, 360 (§1771). 84. Abu TFawaris, Imama, ch. 8. His reply was "one God, one imam". 85. Cf. Sanhoury, Califat, 120ft.; cf. also BaqillanT, Tamhid, 180; cAbd al-Jabbar, MitghnJ, xx/i, 244^; Surinam, Rawda, i, 58, and the discussion in Mawardi, Adah al-dunyd, ijSf. 86. For this distinction, see above, 139. 87. Baghdad!, Usui, 274^; Shahrastanl, Mild, 85; Ibn Hazm, Fast, iv, 88, where the same . view is said to been held by Abu TSabbah al-Samarqandl (on whom, see van Ess, TG, u, 562f.). For Ibn Hanbal, see above, 137. Nature of Government 1-75 -hich the Zaydis and Ibadls had already given a positive answer, and which some Sunnis now answered in the affirmative too.88 The fact that secular rulers had divided the Muslim world between them was also problematic, as has been seen, but it did not affect the religious unity of believers as long as the upstart rukrs were willing to cast themselves as servants of the caliph. There was no question of recognizing them as ultimate rulers of the believers' souls. One could perhaps elevate one of them to the position of caliph, along the lines suggested by al-Juwayni. One could also think away the caliph altogether. This would leave a plurality of de facto sovereigns in the political sense, but they still would not be sovereigns of the particular fragment of the religious community that happened to be in their charge. The religious community would simply be acephalous. Perceptions changed when the caliph disappeared in actual fact, but the secular kingdoms (regna in the terminology of medieval Europe) never succeeded in breaking up the religious community so as to turn each kingdom into a sovereign church of its own. Conciliar government So much for independent rulers. The conciliar model also found occasional adherents after al-Asamm. Thus al-Farabi acknowledged that it was difficult to find a man endowed with all the characteristics desired in a virtuous ruler. If two men possessed the characteristics between them, they should rule together; and if the characteristics were dispersed in many men, then they should jointly form the government. He took that to be what the Greeks had meant by 'aris-ftocracy' (riyasat al-akhyar), as has been seen.89 65/33.7 ~ 35* 2-9° 1 Functions of Government from its association with government altogether. It was customary, according to al-Juwaynl, for the imam to supervise events which served as an external emblem (shi'ar) of Islam, such as the Friday prayer and the pilgrimage, and he ought indeed to pay attention to anything involving large numbers, but in law people were free to organize the rituals themselves.115 According to the Hanaf; Ibn al-Simnanl, all duties that fell on the believers as individuals, such as ritual prayer and alms, continued whether there was an imam or not. Whether he included the Friday prayer under this heading, not just the daily five, he does not say.1' Jama't scholars who deemed the validity of the Friday prayer to depend on the presence of the ruler, in person or through his governor, held his presence to have the requisite effect whethct he was morally upright or not;18 but Shi'ites would not pray in the Friday mosques of illegitimate rulers,19 and the Imamis ceased to have a public Friday service when their twelfth imam went into hiding.20 This briefly changed under the ShI'ite Buyids, when al-Mufld (d. 413/1022), followed by his pupil al-Tusi (d. 45of/To66f), postulated that the imam had delegated his authority to the jurists: they were permitted to conduct the Friday and the Festival service on his behalf, provided that they could do so without getting into trouble.21 But later scholars such as Ibn IdrTs al-Hilli (d. 598/1202) disagreed.22 There was no Friday prayer in early Safavid Iran; to conduct it was to identify oneself as a Sunni. In al-Juwaynl's opinion, no physical act of worship required validation by the imam.24 That the pilgrimage remained valid regardless of the moral status of the ruler is explicitly affirmed by the HanbalT Ibn Batta and the Zaydl l6.Juwaynr, Ghiyath, %%zS$t., 553; cf. Shirazi in Calder, 'Friday Prayer and Juristic Theory', 41. iy.SimnanT, Rawda, i, §114; cf. §115. 18. Ibn Batta, Profession de foi, 67 = 127; cf. Lewis, Political Language, 101. 19. Already under rhe Umayyads wc are told that the adherents of 'All would stay away from Friday service and other public prayers (Tab. ii, 234): such absence counts as a sign of rafd in a Prophetic tradition cited in SN, ch. 41, §14. 20. There were no Friday prayers in Qumm in Buyid times until Rukn al-Dawla forced the Qummls to rebuild and use the Friday mosque (MuqaddasT, Ahsan, 395;. Kulini J. 329.'94of.) has the imams make provisions for performance of the Friday prayer by oneself, or prayers in lieu of it (Newman, Formative Period of Twelver Shi'ism, 168,170). zi.Mufld, Muqni'a, 811; TusT, Nihciya, 302, cf. 107. For Mufid on rhe law during rhe ghayba, see also Arjomand, 'The Consolation of Theology', 562!. 22. Ibn IdrTs, Sarazir, i, 302!!. 23. EMC. Iran., s.v. 'jumca.' 24. JuwaynT, Ghiyath, §§289!". (cf. above, note t6). Functions of Government [ 291 Maj>"tf>23 anc' everyone else seems to have taken it for granted. As regards fasting in Ramadan, only the Ismailis held it to be suspended in the absence of the Ijparn, presumably because they relied on him to announce its beginning and end, calculated by astronomers.26 Other Muslims simply relied on the appearance of the new moon, which did not require expert knowledge. (This did not prevent the Malik! al-WansharlsT from adducing the fast of Ramadan among t[,e ritual obligations that Muslims would be unable to fulfil if they stayed on jn al-Andalus after the fall of Granada in 1492: without imams and their deputies, the sighting could not be accomplished, he declared, urging them to leave.)"' \ Finally, obligatory alms (zakab, sadaqa) remained payable to any ruler whatever his moral status according to some Sunnis, but here there were dissenting voices, above all (though not only) among the Shafi'ites: obligatory-alms on gold, silver, and easily hidden things kept at home could or should always be paid directly to the recipients, listed in Q. 9:60, the dissenters said, rather than to officials (who would violate the privacy of the home); and the same was true of obligatory alms in general when the collectors were unjust, or even when they were not, according to some.28 In the absence of a ruler of any kind, everything to do with public money would have to be managed by the scholars, according to JuwaynT, who does not explicitly mention alms. According to Ibn al-Simnanl, zakab would continue to be payable because it was a duty which fell on the believers as individuals, whereas the imposition of jkya would stop because the duty did not fall on them, but rather on the imam.29 The IbadTs and some Imamis also held that people could disburse the alms directly to the recipients, at least if there was no legitimate imam or one could not rely on the money reaching him.39 What happened if one gave one's alms to a wrongful ruler, voluntarily or under duress? The IbadTs and some Imamis said that one would have to pay them again, assuming that there was someone to pay them to. But there were also Imamis who said that it was lawful to pay them to rulers such as the Umayyads, and that they were not in any case to be I *5-Profession de foi, 67 = 128, cited in Lewis, Political Language, 101; Zayd (atrnb.), wUajmu', 236, no. 853. I 26. Above, note 11: cf. Walker, al-Kirmani, 35. 27. WansharTsi, Mi'yar, ii, ij8f. i 28. Ibn Batta, Profession de foi, 67 = 128; cited in Lewis, Political Language, jot; Aghnides, Theories of Finance, zy6tf.; MawardT, Ahkam, 209/121 = 135. There were even some who held that one could kill unjust collectors, cf. below, nore 93. I 29, JuwaynT, Ghiyath, §560; SimnanI, Rawda, §§TT4f 30. Cf. Bisyani, Mukhtasar, 93: Newman, Formative Period of Twelver Shi'ism, i66f. 2.0)2. Functions of Government paid twice.31 in the imam The Zaydis and Ismailis held that (the true) zakah was suspended s absence,32 and so apparently did some Imamis, on the grounds that ordinary believers lacked the knowledge to hand it directly to the rightful recipients, and that thtee of the categories of recipients listed in the Qur'an presupposed political organization: the Imami al-Mufld and his pupil al TusT brushed aside these objections, arguing that zakah was payable to the. jurists, who did have the requisite knowledge, and that the three problematic categories were simply suspended.33 Execution of the law (tanfidh al-ahkamj Executing the law was the essence of the imam's shaft functions. It was to implement the moral and legal rules [sunna, hukm, hudud, fara'id, huqiiq) brought by the prophets that God had raised up caliphs, as al-Walid II said in 744.34 All lists of the imam's functions mention this duty, and all the functions listed separately can be seen as subdivisions of it. In the early days the caliph would execute the law in person by adjudicating in person, and whether he did so or not, it was generally agreed that only the imam (or a delegate of his) could appoint judges.35 It followed that if there was no imam, people would have "no judgements {ahkamy, as the Umayyad poet Jarlr put it.3S One would have make do with sulk, private agreement or settlement out of court.3' The Imamis had no courts capable of enforcing their decisions even back in the days when their imams were present, and their traditions sternly warn the believers not to use the courts of the opponents, telling them to submit their dispute to a traditionist or jurist and accept his judgement of their own accord/8 When al-Mufld, 31. Muhammad h. MahbQb to the North Africans in Käshif, Siyar, 11, 2301. (where the wrongful officials are IbädT, not Sunni); Tüsl, Nibäya, 185; Newman, Formative Period of 'Fwelver SbiHsm, 174, T77. 32. Cf. Qäsim b. Ibrahim in Strothmann, Staatsrecht, 5m Abrahámov, 'al-Käsinť, 86; not even a muhtasib imam could collect them (Madelung in EF, s.v. 'imäma'); Nu'män, Díŕa'im, i, 263í. 33. Mufid, Muqni'a, 252; Tüsl, Nibäya, 185; Calder, 'Zakät in Imämí Shí!í Jurisprudence', 469. 34. Tab., ii, 1759fr.; Lr. Crone and Hinds, God's Caliph, 121ft. 35. Ibn al-Muqaffa', Sahiba, icnf. (P, §17); Sanhoury, Califat, i68f£; Tüsl, Khiláf, v, 343.6. Cf. above, 238ft, on Ghazali. 36. Above, note 8. 37. Cf. ťAbd al-Jabbär, Mugkm, xx/i, 53t 38. KulTnT, Käß, vii, 410-12; Madelung, Authority in Twelver Shiism', 166, citing Ibn Bäbawayh; Newman, Formative Period of Twelver ShT'ism, r8of. Functions of Government followed by al-TusI, proposed the theory of delegation to the jurists (tafuMd), f they affirmed that it was up to the jurists to take over the functions of qadis.39 aJ-Tusi added that a person appointed by the wrongful regime should try to use his position to apply Imami law; under duress he might even apply the law of the opponents, but only as long as it did not cause him to take lives: tuqiyya (dissimulation under duress) could not legitimate unlawful killing.40 On the Sunni side al-Mawardl agreed that there could be no execution of the law (and also no collection of taxes) in a community without an imam, whereas an imam raised up by rebels would in his opinion validate its execution in the rebel community: he did not deem the (mostly Sunni) subjects of the Fatimid caliphs to be living in sin, one infers, though he does not mention any examples.41 Here as elsewhere, the idea is that the public domain is created by God, the only power capable of overruling private interest, and that God has to be represented in the here and now by a single person, a deputy. But al-JuwaynT argued along the same lines as al-Mufld and al-TusT that the scholars could take over the execution of the law in the absence of a such a deputy: they could marry off women without marriage guardians and administer the property of orphans, for example.42 Al-Ghazall, his star pupil, disagreed, as has been seen: all transactions dependent on judges would be invalid in his view, or alternatively validated by overriding necessity [darura) alone, for the sultans to whom most judges owed their appointments had no moral right to appoint them unless they were authorized to do so by the caliph. Later Sunnis solved the problem by tracing the authority of the sultan to the community, thus reversing the original relationship between the community and its head, and/or by seeing the sultans as imams themselves.43 Execution of the hudud The modern state is often held to be an agency for the maintenance of internal order and the conduct of external defence distinguished by its monopoly on the right to use violence, where violence means force of the type required for physical damage, imprisonment or death: people may still slap their 39-MufTd, Muqni'a, 811.5; Jusi, Nihdya, 301. The view was affirmed again in the Safavid period, cf Calder, 'Judicial Authority in Imam! ShT'I Jurisprudence,' 105. 40. Tusl, Nihdya, 302. 41. MawardT, Abkdm, 95/59 — 65; cf. Mikhail, Politics and Revelation, 23 (wrongly having Maward! speak of the religious duties in general); Abou El Fadl, 'Islamic Law of Rebellion', ii, 205!. I 42. Juwaynl, Ghiyath, §557—8. 43. Above, 238ft 294 Functions of Government children (if only just), but in mosr countries they may not bear arms (except b special licence), and nowhere may they physically injure, detain ot kill oth people, except in self-defence (narrowly defined). If the state were to lose th' monopoly, it would not simply be sharing a function with, or ceding a function to, private citizens; rather it would cease to be a modern state. Contrary t what one might have expected, given the level of violence they had to tolerate some medieval Muslim jurists operated with similar notions, or so at least al-Juwaynl. Private individuals (ahad al-nas) were not allowed to unsheath weapons against each other or their rulers, or only in self-defence and th» res cue of others, he said, adding that this did not apply if there was no govern ment. (Using arms against infidels was also another matter.)44 Many jurists disagreed, as will be seen. There was, however, complete unanimity that the imam had a monopoly on the right to use force in one key area. The area in question was that of the hudiid, penalties prescribed in the Shar'Ta which resulted in physical damage (by lashing), mutilation (by amputation) or death (by stoning or decapitation).45 That only the imam could apply or authorize the application of these penalties to free persons was affirmed by all Muslims with rare exceptions, such as the ninth-century anarchists, to whom they were a major problem.46 Many jurists went further by crediting the imam with a monopoly on all punishments, including discretionary flogging (ta'zir), torture, imprisonment, and banishment, which are not prescribed in the Shar'Ia;47 but only of the hudiid can it be emphatically said that if he ceded them to others, he would cease to be the imam, or alternatively the penalties would cease to be hudiid. They were penalties for fundamental transgressions of the moral code which held society together and had to be inflicted by a representative of God, that is to say a representative of collective interests, because there would not otherwise be anything to distinguish them from private vengeance. If everyone had the right to kill or maim other people, there would be no legal order to uphold: power would lose its moral 44. Ghiyath, %%i6y 479b, 4856, 554 and irshad, 370.1 (Cook, Commanding Right, 546). 45. The penalties are classically identified as those for unlawful sexual relarions [zina), false accusations of such relarions (qadhf), highway robbery (qat' al-tariq), theft (sariqa) and wine-drinking; but many others count as hudiid in medieval works, including those for apostasy, blasphemy, and homicide; and there are times when hadd seems to mean capital punishment. 46. Cf. above, 67b 'Abd al-Jabbar knew of Medinese jurists who would allow all individuals ro perform them on behalf of an imam unable to do it himself (Mughm, xx/2, 155). For the question whether slave owners could perform them on their slaves, see Johansen, 'Mise en scene du vol,' 46; Naysaburl, Imama, 67. 47. Cf. Cook, Commanding Right, 268, note 103 (Murtada), 342, note 21 (HalimI); Naysaburl, Imama, 67.5. Functions of Government 2-95 urp0sc, anarchy would prevail;48 and one would be back in the proverbial man 'azW tew ('whoever has power takes the spoils'; which had prevailed in the ijhiliyya. The Imamis added a consideration likely to appeal to modern opponents of capital punishment: one needed infallibity to impose such penalties.49 gut with or without infallibility, the authority behind them had to be public. Consequently, the imam's monopoly on the hudiid is frequently affirmed in the literature, whether Sunni,50 Mu'tazilite,51 ShT'ite,-'2 or IbachV3 Since the hudud could not be applied by unauthorized people without ceasing to be hudiid, they could not be executed in the imam's absence. This was the prob-jem that the ninth-century anarchists had struggled with. When the twelfth imam went into hiding, the Imamis duly declared the hudiid to be suspended, pending his return.54 Xot all held the suspension to be total, however. A tradition in al-Kulmi allows self-help in the case of unlawful sexual relations, as long as the penalty was carried out in secret." Several jurists of the Buyid period tried the alternative method of ruling that if an Imami official employed by a wrongful ruler was in a position to apply the hudiid in accordance with Imami law, then he was authorized (or even obliged) to do so, for he would in fact be acting on behalf of the true imam and should think of himself as doing so.56 This ruling reflected the fact that the wrongful rulers at the time were ShTcites; living under Sunnis, thirteenth-century jurists tended to disagree with it.57 Al-Mufkl and al-TusI held that the jurists were allowed to apply the hudiid even without holding office, by delegation from the imams, when they were able to do so, as within their own households: they were authorized 48. Cf. Naysaburl, Imama, 67, explaining what is implicit elsewhere. 49. Cf. l'usi, Mahsut, vii, 41.6", cited in Sachedina, just Ruler, ioof. Ir was also !Abd al-Jabbar's understanding of the Imami position [Mughm, xx/i, 74). 50. BaghdadT, Vsiil, 272.7. As he observes, some held that slave owners could impose I them on their slaves (cf. above, note 46). 51. 'Abd al-Jabbar, Mughm, xx/'i, 41, 74, noting that if there is no imam, the hudiid must be suspended [la budda nun suqut al-hudiid); Mankdlm in Cook, Commanding Right, 215. 52. Below (Imami); Strothmann, Staatsrecht, $n, cf. 97 (ZaydT); Naysaburl, Imama, 67 (Isrnaili: not even slave owners). 53. Bisyani and Muhammad b. Mahbub in Kashif, Siyar, ii, 197!., 239. 54. Murtada in Sachedina, 'Treatise on the Occultation', 114, and other Imamis in Madelung, 'Treatise', nore 25; similarly cAbd al-Jabbar, Mughm, xx/'i, 74f., on the Shi'itcs. 55. Newman, Formative Period of Twelver Shi'ism, 177L 56. Murtada in Madelung, 'Treatise', 23 = 16, with other Imamis in note 25; Mufld, Muqni'a, 810; TusT, Wihdya, 301, 302. 57. Thus Ibn Idrls and Muhaqqiq al-Hilll; cf. Calder, 'Legitimacy and Accommodation', 96b; Madelung, 'Treatise,' note 25. 2j6 Functions of Government children, and slaves.58 Functions of Government to apply the penalties to their own wives, children, and slaves.58 Here tlie; attempt to save public authority ended up as an endorsement of self-help if only for jurists, and perhaps only in their capacity as domestic tyrants. It was in any case a far cry from the infallible public authotity that had once bee required for the task, and it was too much for Ibn Idris al-HillI: it was only t his slaves that a man could apply the badiid in his view.59 According to al-Mufid, people who took it upon themselves to kill bias phemers and other apostates 'out of anger on God's behalf, would be free to do so, apparently whether the imam was present or not.1' His Mu'tazilite contemporaries agreed, though only if there was no imam, ot so 'Abd al-Jabhar insists.61 The people in question should probably be envisaged as applying verdicts formulated by the jurists tather than acting on their own accord; in other words, the reference is to laymen acting as the jurists' henchmen. This was a role which laymen were often encouraged to take, by jama'i and ShT'ite jurists alike (and which became infamous in the West when Khomeini used it to deal with Rushdie). "Anyone who meets him and kills him is acting on my order" as an early 'Abbasid judge declared with reference to an alleged crypto-Manichean (zindiq).61 The occasions on which scholars acted as rabble-rousers against theologians and philosophers are legion. As long as the decision was reached in accordance with the law (as formulated by them), they did not mind appealing directly to the community for execution. It was how outlawing functioned in medieval Europe too. Authors connected with the government, however, were well aware of the overriding importance of reserving the infliction of physical punishment to the state and its officials. A model letter of appointment instructs a military commander "not to apply a hadd or a rule concerned with retaliation, whether involving loss of life or limb, without asking the Commander of the Faithful for his opinion and awaiting his reply".63 A military commander should always execute major physical punishments himself, or at the most delegate it to a close associate, according to a military treatise.64 Anyone who beheaded, 58. Mufld, Muqni'a, 810 (the imams qad fawwadit al-nazar fihi ild fuqaha' shi'atibim ma'a Y-i'mkarc); TusT, Nihaya, 30of., clearly on the basis of Mufld. 59. Ibn IdrTs, Sard'ir, ii, 24. 60. Mufld, Muqnfa, 743, with reference to blasphemers {man sabba rami allab an ahaa min al-a'imma). They are deemed to be outlaws on the ground that they are apostates, so the rule presumably applied to other apostates as well. 61. 'Abd al-Jabbar, Mughni, xx/z, 156.1, 8. 62. WakF, hi, 265.4; cf. van Uss, TG, iii, 34, on Sa'Td b. 'Abd al-Rahman al-Jumah! and his victim, Qirar. 63. Thus Qudama, Khardj, 45. 64. HarthamI, Mukhtasar, 17. [ 2.97 mutilated, castrated or otherwise punished anyone else without the king's permission, "even his own servant or slave", should be punished according to Nizam al-Mulk, "so that others may take warning and know their places".65 \Vhat then would happen if there were no imam? Al-JuwaynT avoids the issue, but Ibn al-SimnanT concedes that the hudud would have to be abandoned.66 Jihad Holy war was one of several types of warfare regulated by the jurists. Al-jvlawardi called the other types 'wars of public welfare' (hurub al-tnasalih), by which he meant the suppression of apostates {murtaddun), rebels (bughdt) and brigands (muhdribun) ,67 Only jihad will be treated in detail in this book. The main account will come in Chapter 11; what follows is concerned only with the relationship between the imam and holy war. If one could have a valid community while tempotarily deprived of an imam, one was also entitled to defend it by force of arms: defensive warfare remained legitimate whether there was an imam or not according to all (though not all counted such warfare as jihad, except in the case of emergencies).68 But the legitimacy of jihad in the sense of warfare for the spread of Islam was more problematic. The only reason why Muslims were entitled to invade the lands of other people to impose their own government on them was that they were doing God's will, and this was not self-evident if they were led by sinners. Companions are sometimes said to have had their names removed from the military roll after the death of TJmar or that of 'Uthman, or whenever they took right guidance to have come to an end,69 and the question how far it was lawful to participate in jihad under sinful rulers was horly debated. A negative answer implied that the activity was suspended, not because it had ceased to be obligatoty, but rather because circumstances made it impossible for the obligation to be discharged. I The abl al-sunna wa'l-jama'a took the view that jihad could and should be waged behind the ruler of the time whatever his moral status; the activity 65. SN, 98 = 76 (ch. 11, §4). > 66. Rawda, §irs. 67. MawardT, Ahkdm, ch. 5 (caption and introduction); cf. SarakhsT, Mabsut, x, 2, where the generic term for all types of licit war is siyar. On wars against rebels and brigands, see Abou F.l Fadl, 'Irregular Warfare and the Law of Rebellion', and 'Islamic Law of Rebellion'. 68. Jihad and border defence appear as different rubrics (nos. 5 and 6} in MawardT, Ahkdm, 23/16 = 16; similarly Abu Ya'la, Ahkdm, 27; Juwayni, Ghiyath, §§508-10. The idea that jihad was primarily defensive is of apologetic origin and did enter juristic writings in medieval times (though one does encounter it elsewhere, cf. below, 382.). 69. Crone, 'Qays and Yemen', 40, note 223. 29 8 Functions of Government Functions of Government 299 derived its validity from the law, which was validated in its turn by communal agreement, not by him, and whether he was sinful or upright was of no conse quence to anyone except himself.70 Nor would the duty be suspended in his absence, for it was imposed on the community, not on him. Holy war was a fard ktfdya, a collective obligation, and such obligations could not be suspended71 If those who normally fulfilled them stopped doing so, they would devolve to others and eventually become individual obligations (sing, fard cayn) on everyone until somebody fulfilled them again. A collective duty was not primarily a duty for the state. Rut given his role as the upholder of the law on the one hand and the vast resources at his disposal on the other, the ruler was naturally expected to take a leading role in the organization of holy war. He was in effect all Muslims in a single person, or the agent of their community (na'ib 'an jama'atihim), as cleventh-centurv jurists put it.72 But did the imam, when there was one, have an actual monopoly on the conduct of war, in the sense that only he could authorize the inception and termination of campaigns? The answer may once have been ves. All the other items on the standard lists of the imam's duties are activities that only he could perform or authorize others to undertake, and Ibn al-Muqaffa' (d. c. 757) explicitly says that only the imam was entitled to obedience in matters of "starting campaigns and marching back (al-ghazw iva'l-quful), collecting and distributing (booty) . . . and fighting the enemy and making ttuces with him".73 Sunnis often express themselves in similar terms.74 For all that, the Sunnis held that laymen were free to initiate campaigns on their own. Participation in jihad was highly meritorious for everyone, and civilians would often join the official campaigns as volunteers {mutatawwfa) or go to live on the frontier for extended periods, attaching themselves to fortified settlements of a private nature known as ribats. The jurists make it clear rhat such volunteers would often set off in small raiding parties, now from the regular army and now "from a town in Syria or elsewhere",75 to campaign in enemy territory on their own, without permission from the imam or his representative. 70. cAbd al-Razzaq, ivlmannaf, v, nos. 9610—13; Ibn Batta, Profession de foi, 67 — 127; Ibn 'Ukasha, and others, above, 136, 137; Malik in Talbi, P.mirat aghlabide, 4ijn. (also in Mortahedeh and Sayyid, 'Idea of Jihad', 26, who link it writh the debate over the obligatory nature of jihad); SarakhsT, Siyar, i, r6o, no. 161. Similarly Zayd b. cAli, Wlajmif, 236, no. 853: the dominance of the wicked does not invalidate holy war. 71. Cf. Juwayni, Ghiyath, §553. 72. Ibid., 5307; SarakhsT, Siyar, 1, 189. -i.Sahaba, 197! (P, §17). 74. Hamidullah, Conduct of State, §372 (Abu Yusuf, MawardT); Baghdad!, Usui, 272; cf. also the Mu'tazilite MankdTm (d. 425/1034) in Cook, Commanding Right, 215. 75. Abu HanTfa in Tabari, Ikhtildf, 79. ' ■fhey do not condemn the practice. They do have their reservations about it with reference to the safety of the participants, the maintenance of military discipline, or the problematic status of the booty (could they keep it, or was the imam entitled to his fifth?).76 The Malikls said that if the imam had prohibited fighting for the sake of general welfare, then nobody was allowed to fight unless attacked (in which case no permission was needed to fight back).77 But no school prohibited unauthorized campaigns outright. "We allow volunteers in holy war to penetrate the land of obdurate infidels on their own, though it is better that they should do so at the initiative of the imam," as al-JuwaynT observed.78 The question how the imam was supposed to enforce truces and otherwise manage relations with the enemy if the frontier population was out of control does not seem to attract attention, perhaps because unauthorized expeditions were usually too small to make much difference, when a religious scholar who campaigned with great success against the Turks in Tahirid Khurasan was denounced by envious people, the ruler accused him of "going out and gathering this army around you and disobeying the assistants of the government". Perhaps he had overstepped the limit by gathering so large an army, or maybe his detractors had accused him of rebellious intentions (he was released wdren the ruler was convinced of his loyalty).7' In any case, private warfare was lawful as long as it was jihad. The Kharijites held jihad to be highly meritorious whether they had an imam or not, but how they explained its validity in his absence is unknown.80 The ShT'ites, on the other hand, held that jihad in the sense of missionary warfare could only be waged under the imam's banner; nobody was to participate 76. Abu Zayd al-QayrawanT (d. 386/996) in Bredow, Heilige Krieg, Arabic text, i8ff. (drawn Co my attention by Christopher Melchert); Tabari, Ikhtildf, 78!; Qaffal, Hilya, vii, 1657: Abu HanTfa prohibited unauthorized expeditions unless they were at least ten men strong; al-Awza'I allowed the commander to punish or forgive men who went off on their own as he wished; the Shafi'ites disliked expeditions undertaken without the imam's permission because of the risk, but did not hold them to be forbidden whatever their strength; the MalikTs said that troops were not allowed to undertake unauthorized expeditions whereas people along the frontiers could do so if a good opportunity arose and it would take too long to await the imam's permission. 77. Bredow, Heilige Krieg, Arabic rcxt, 20.t. 78. Ghiyath, §486. 79. DhahabT, Nubald', xi, 34 (drawn to my attention by Christopher Melchert). The scholar was Ahmad b. Harb (d. 234/849). Compare Ibn al-Athir, vi, 361, year 205, where an earlier volunteer is suspected of rebellious intentions because he has gathered a large army I on his own initiative in order To fight Kharijites. 80. Cf. Crone and Zimmermann, Epistle, 281. Salim tells people going out to holy war to appoint 'imams' for the duration. 3oo Functions of Government unless he was summoned by him or his representatives: thus the Imatiiis Ismailis, and ZaydTs alike.81 The activity was suspended in his absence. "When no imam is manifest and nobody appointed by the imam is present, it is no( allowed to fight holy war against the enemy, for jihad with imams of injustic or without an imam is an error for which the agent incurs sin," al-TusT said' there could be "no fighting to make them (namely the unbelievers) adopt Islam", only to protect the Muslims, and no sojourn in military settlements along the frontier unless the warfare was defensive, until the coming of the Mahdi.82 It would have entailed mixing with and taking orders from the opp0. nents. Al-Mufld did think that Imamis holding office on behalf of illegitimate rulers could, indeed must, wage jihad against infidels and sinners alike, and that the community had to assist them .83 But al-Tusi omitted holy war from the functions delegated by the imam to the jurists, be it with or without appointment from the wrongful regime. Later scholars also held missionary jihad to be suspended, though here as so often the coming of the Safavids caused them to rethink.84 Commanding right and forbidding wrong Islamic law obliged its adherents to intervene when they saw other believers engage in sinful behaviour and to persuade them to stop, or even to force them to do so if they could. This was called 'commanding right and forbidding wrong' {al-amr bi'l-ma'ruf wa'l-nahy can al-munkar), and it was often compared with holy war: like jihad, it was a call to Islam backed by force where necessary; fighting sinners and fighting infidels were much the same.85 Some saw government in its entirety as a type of 'commanding right and prohibiting wrong',86 presumably on the grounds that the ruler's function was in essence 81. For the Imamis, see the next note. For the Ismailis, see Nucman, Da'a'im, i, 264.-, Naysaburi, Imama, 66 (no jihad except under his banner, no expedition unless sent by him or his representative); cf. also Abu '1-Fawaris, Imama, 7 — 25 (85V). lor the Zaydis, sec Strothmann, Staatsrecht, 511., 97, cf. 9611. (the mubtasib imam can use the sword in the performance of al-amr bi'l-ma'ruf). Differently Zayd b. 'All (attrib.), Majmuf, 236, no. 853, but this manual played no role in classical ZaydT law. 82. TusT, Nihaya, 29of; cf. Kohlberg, Tmami ShT'T Doctrine of jihad', 79ft; Sachedina, Just Ruler, not There is no reason to think that defensive warfare was ever regarded as suspended. 83. Muqni'a, 810. 84. Kohlberg, Tmami ShT'I Doctrine of Jihad', 8if. 85. Cook, Commanding Right, 198, note 21 (Mas'udi), 34T (Haliml). %6.Al-sakana biya hadhd, as HalTmT says in Cook, Commanding Right, 342n.; cf Juwayni, Chiyalh, §113. Functions of Government 301 to apply c'le law> but Sunni lists of the tuler's functions do not normally mention it. The duty was usually (though not always) seen as collective, and the ruler for his part fulfilled it by appointing a mubtasib (censor and market inspector) who would patrol the streets with armed assistants to ensure that people obevetl the law in public, for example, by attending Friday prayer, fasting in Rjrnadan, abstaining from wine, and observing the rules regarding relations between the sexes. What people did in the privacy of their own homes was their own business as long as their actions did not affect others, directly or indirectly (e.g. by their behaviour coming to be known). But the existence of a public censor notwithstanding, it was meritorious, or even obligatory, for private citizens to take the duty of enforcing public morality upon themselves when they were able to do so. As in the case of jihad, the question arose as to how far they had to act under the imam's control. Since performers of the duty normally took spontaneous action at the sight of what they deemed to be wrong, rather than responding to fatwas issued by the learned, the jurists now had to consider the wisdom of allowing laymen to trespass on not only the imam's territory, but also their own. The earliest material is dominated by the problem of the sinful behaviour of rulers rather than that of fellow-citizens, and the key issue is the legitimacy of revolt: may one use the sword against the wrongdoing of caliphs and governors?87 The answers became increasingly negative with the passage of time, and by the ninth century only the Mu'tazilites, the Kharijites, and the ZaydTs said that one could.88 There remained the question how far one should attempt to rebuke them (given that one might risk one's life thereby): most scholars said that it was not obligatory, and not necessarily even meritorious, especially in public.89 How far could private individuals go in their action against the wrongdoing of their fellow-citizens, then? Some scholars would prefer them not to do anything at all. Thus al-HalTmT (d. 1012), a Shafi'ite jurist, held the performance of the duty to be so closely related to the infliction of punishments that it would be better for the ruler, in his case the Samanid amir, to take it over by appointing an upright and learned man in every town and village to execute it. If the ruler did nothing, the duty devolved onto the scholars, he said, though he allowed upright laymen to intervene when the law involved was simple.™ 87. Cf. Cook, Commanding Right, chs 2-3. 88. Cf. ibid., chs 9,10, 15. 89. Cf. Cook's own summary in his Forbidding Wrong, ch. 7. 9c. Cook, Commanding Right, 342f.; there is also a Prophetic tradition prohibiting lay performance (p. 43, note 56), and Ma'mun is said ro have done the same (p. 71). 302 Functions of Government Other scholars agreed that the common people should stick to simple they preferred them just to disapprove 'in their hearts', without taking a, action; and many of them placed the use of violence, or at least armed vi lence, in the ruler's hand, forbidding it to laymen or requiring them to obtain the ruler's permission.91 Even the Hanbalites came around to this view. ]n th early tenth century they were notorious for their street violence in Baghdad led by the rabble-rouser al-Barbaharl: in a famous incident in 935 they raided pri vate homes, poured away wine, broke musical instruments, beat up singing girls, thrashed men unable to convince them that the women or young hoys with them were lawful companions, and organized assaults on Shafi'ite schol ars, causing first the chief of police and next the caliph himself to intervene But thereafter their relationship with the government warmed, and by the time we reach 'Abd al-Qadir al-JTlT (d. 1166) and Ibn al-JawzT {d. 1200), the ruler's permission had come to be required for the use of violence in performance of the duty52 But there were some notable exceptions, especially in the Shafi'itc school " and above all in al-GhazaiT (d. nil), who authorized private individuals to take up arms in fulfilment of the duty, to collect armed helpers iff wan), and even to form troops (tajnid al-jtmud). He insisted that unjust rulers had to be endured, but he saw their subjects as direct executors of Islamic law, bypassing government officials. This was too extreme for most. His discussion of al-amr bi'l-ma'riif was enormously influential, but few later scholars shared his vision of every Muslim as his brother's keeper, equipped with permission to use force in the maintenance of the public morality that had once been, and in principle still was, a key concern of the state.94 Outside the Sunni ranks, the Zaydls, the Kharijites (insofar as they wrote about it) and the Mu'tazilites held laymen to be free to use armed violence when it was needed, as one would expect, though one Mu'tazilite would only-grant them this right in the absence of a legitimate imam (whose monopoly would otherwise be infringed).9' The Imamis were closer to the Sunni position. They were often accused of declaring al-amr bi'l-ma'ruf 10 be suspended in the imam's absence, but the accusations were false; in fact, they affirmed its 91. Cf. Cook's summary in his Forbidding Wrong, ch. 3, §§91-6. 92. Cook, Commanding Right, ují., 138, t40. 93. Thus Qaffal (d. 976) and Kiyá al-HarrásT (d. ino) held it lawful to kill offenders; the latter (inspired by the Mirtazilite al-Jassas) singled out collectors of illegal taxes as legitimate targets. Mawardi permitted the recruitment of (armed?) assistance in one work, though he forbade it in another (Cook, Commanding Right, 341, 344b, 347)- 94. Cook, Commanding Right, 431, 441, 456ff. 95. Cook, Forbidding Wrong, ch. 3, §§94, 96. Functions of Government 3°3 c0Iltinuing validity. Indeed, they sometimes went so far as to affirm the right of private individuals to use force 111 the course of its performance; but the majority including al-Mufld and al-TusI, ruled it out by requiring the permission of tne (absent) imam for killing, wounding, or for the use of violence of any I Wid."1 This is in line with the quietist nature of the sect, but they must have been motivated by a concern for internal order as well, for like al-Ghazali, they could have authorized laymen to use force against each other without thereby authorizing revolt against wrongful rulers. Empowering private individuals to use force on a sponraneous basis in everyday life was however to invite chaos, and particularly dangerous in a community living under alien authorities. Preservation of the religion (hifz al-dlnj [t was the imam's duty to maintain orthodoxy (right belief) and orthopraxy (right behaviour) among his subjects. The muhtasib played a key role in both. He secured observance of the law in public, as seen already. He also checked on beliefs propagated in public. "If an innovator appears or a holder of suspect views goes astray, the imam should explain and clarify the correct view to him, and make him undergo the penalties appropriate to him, so that the religion may be preserved from flaws and the community preserved from error."97 The muhtasib was to ensure that people performed their ritual duties. He was also authorized to test religious teachers and correct anyone engaging in false interpretation of the Qurian, the transmission of bad traditions, and the dissemination of doctrines contrary to the scholarly consensus. If the culprit repented, no further action was needed; if he persisted in his ways, the censor was ro pass him on to the ruler.98 It was the ruler's duty to examine him and either make him repent or execute him, according to jurists and mirror writers alike.99 There is plenty of attestation of the procedure in practice, sometimes with the judge taking the muhtasib's role. Most culprits were probably denounced by their neighbours or colleagues, or brought 10 the attention of the muhtasib or the qddi by an outraged mob. They included Mu'tazilites, dualists, materialists, philosophers, Sufis, ShT'itcs, would-be prophets, deviant Qur'an reciters, and blasphemers. Active hunts for heretics on the part of the government, 96. Cf. Cook, Commanding Right, 270, note 116, z66f(. I 97. Mawardi, Abkam, 23/15 - 16'. Similarly Abu Ya'la, Abkam, 27. 98. Mawardi, Abkam, 4o8ff., ^i^ti.iz^ii., 247ft = 263ft, 268ff.; Abu Ya'la, Abkam, li87ff., 292L 99. Shahrastani, Nihaya, 478; Ps.-Gha7.ilT, Kasiha, 59t., where the culprit may also be exiled. 3°4 Functions of Government 00 and there was little prying into peopl, though not unknown, were rare homes. Some authors recommend more active steps to secure orthodoxy: the rule should send out missionaries and Qur'an readers to every area under his con trol.101 But though the government would eliminate heretics, it does not ofte seem to have engaged in internal missionary activity Fiscal services The imam was expected to collect and distribute three taxes, the poll-tax (jizya), land-tax (kbaraj), and tithe I'ushr, often called sadaqa). If there was no imam, they could not be lawfully collected. Generally speaking, this was both the Sunni and (until the Safavid period) the Imami position, with the usual disagreement over who or what the imam was: one paid one's taxes to the ruler whether he was righteous or not, according to the Sunnis, but there could h* no such thing as an unrighteous imam for the Shl'ites.102 The imam was also entitled to a fifth of all moveable booty (khums), but he was not supposed to collect any other taxes. In practice he always did. Attempts to suppress uncanonical impositions (mukus, sing, maks) are a constant feature of Islamic history.103 Two of the above-mentioned taxes, the poll-tax and land-tax, were classified as fay', revenues from immobile booty. This had interesting implications for people's view of them. When the Arabs conquered the Middle East, we are told, they distributed moveable spoils {ghamma) among themselves as they went along and considered doing the same with land taken by force (all such land having passed into the ownership of the conquerors), but decided to leave it in the hands of the original occupants in return for taxes. The government administered these immobile spoils on behalf of the conquerors who were their real owners and passed them their income in the form of stipends. Who then were the owners when the first generation of conquerors died? The children of the actual participants in the conquest of the lands involved, all Arab soldiers in the garrison city administering them, or all free Muslims there? By the ninth century it was agreed that the fay3 was the common property of all Muslims regardless of whether they were descendants of the conquerors or of too. The best known example is al-Mahdi's hunt for 'dualists' (cf. Chokr, Zandaqa). roi.juwaynl, Ghiydth, §282; ShahrastanT, Aqddm, 478. Compare Halimi's vision of official enforcement of orthopraxy, above, note 90. 102. Mawardi, Ahkam, 55 ''34 = 36; SimnanT, Rawda, i, §115; Ibn Batta, Profession de foi, 67 = 128; Madclung, 'Shiire Discussions of the Legality of the Khardf. ro3. Cf. Aghnides, Theories of Finance, 376ft Functions of Government 305 later arrivals, and whether they were Arabs or non-Arabs, settled people or bedouin, male or female, slaves or free: all that counted was membership of the jvluslfm community. By then the debate was academic in the sense that only a [jnv number of Muslims were registered for payment of stipends, but it established the fundamental principle that public revenues should be spent in the interests of all, not of rulers or privileged groups.104 "There is no doubt that the treasury belongs to the Muslims while you spend much of it for yourself," as Harun al-Rashld's wife tells her husband in a story told by Nizam al-Mulk, "they are quite justified if they complain about you."103 When al-Ghazali tenounced his chair in Baghdad and gave away most of his wealth, he justified his retention of enough to maintain himself and his children on the grounds that the revenues from Iraq (which had been conquered by force) were earmarked for good works, since they were a pious foundation (tvaqf) for the Muslims; such money was absolutely clean, he claimed (many, especially Sufis, will have disagreed).106 Just as conquest had established a sense of common ownership of land in many early Greek communities, so it created a sense of common ownership of the proceeds of land in that of the Muslims. The imam was also expected to strike coins, but this function did not achieve shan't status, though it was an unquestioned monopoly of his and of well-known symbolic significance too: the very first thing a rebel claiming soveteign status would do was to strike coinage of his own.10' NON-SHARcI DUTlhi Internal security Most lists of the imam's sbaH functions mention that the ruler must defend the community against external enemies (whether or not they count it as jihad),™ and some add that he must suppress evildoers, but this is mostly treated as self-evident, given his responsibility for the execution of the law. One has to turn to the non-legal literature to find internal safety spelt out as a fundamental desideratum in its own right, often with reference to the roads. Ensuring the safety of the roads was one of the three things that the common people really wanted government for according to al-Mansur, and one of the 104. Madelung, 'Has the Hijra come to an End?', itff, 105. SN, T92 = T45 (ch. 40, §5). 106. Munqidh, 38 = 61 (Wart's translation is not entirely right here); cf. below, 348. 107. EI2, s.v. 'sikka'. to8. Cf. above, note 68. 3o6 Functions of Government two tasks that rulers were appointed for according to lbn al-JawzT; both si gled it out along with external defence."™ It is also one of the duties mentioned in practically every mirror for princes. Roads, bridges, inns, walls, mosques, and other infrastructure The construction of roads, as opposed to their safety, rarely figures among the services expected of rulers,110 but the building of bridges, mosques. madrasa% and fortified centres (ribats), is often mentioned. The imam had to budget for such activities under the rubric of general welfare according to al-Ghazali,111 \ proper king would build underground canals, bridges across great waters, inns along the highways, fortifications, and new towns, and would restore ruined villages and farms to cultivation, according to Nizam al-Mulk.11J Muslim kings and governors often did so, generating the question how far it was permissible to use their bridges and other amenities when they were oppressors.113 Headmen and other notables often did so as well.114 Insofar as there was a formal division of labour, the treasury was responsible for the upkeep of the water supply, mosques and defensive walls in a particular locality when funds were available, but the responsibility devolved to the local men of means when the treasury was empty, or so according to al-Mawardl. The local men would need the ruler's permission to destroy buildings of importance to all, such as the city walls or the Friday mosque, but ordinary mosques were private.115 109. Above, 158; lbn al-JawzT, lalbis, 129 = IV, 172. no. But "Greek rulers were always building level roads through difficult territory, filling hollows, cutting through high mounrains and banishing fear of them," the eleventh-century Mubashshir b. Farik said in an account in which the Greeks typify rhe good old days 'Rosenthal, Classical Heritage, 35); and Transoxanian nobles were famous for their upkeep (lbn Hawqal, Sur.it al-ard, ii, 466.2.0). in. Fada'ib, 11H (ch. 9, under the rubric of wartf). 112. SN, izf. = 10 (ch. 1, §3). Compare the budget in Tankb-i Sislan, 3if. = zif.; rhe Greek kings who in the good old days "were always constructing various kinds of bridges, erecting strong walls, building aqueducts and diverting rivers" (Rosenthal, Classical Heritage, 35); and MawardT, Tashfl al-nazar, 214, 279 (bridges, water supply, safe roads). it3.GhazalT, lbya*, ii, 125.21 (mosques, ribats, bridges, watering provisions). For the worries, see below, 348. j 14. Thus notables of Transoxania (lbn Hawqal, Surat al-ard, ii, 466.2.1), the headman in SN, 197 = 149 (ch. 40, §14;. 115. Ahkam, qiifji^f. = 266; discussed in Aghnides, Theories of Finance, 35off. Functions of Government Poor help, disability pensions, famine relief [ 307 I Help1"? ^ foor WaS 3 5^'"X' ^uty fulfilled partly by the government and j artly by private individuals. It was the ruler's duty to distribute the obligatory alms, paid once a year by all adult Muslims, to the poor and the indigent, I slaves buying their freedom, debtors, travellers, holy warriors, collectors of the alms, and 'those whose hearts have been conciliated' (i.e. enemies placated by financial largesse), these being the recipients mentioned in the Qur'an, 9:6; and he also had to disttibute the fifth (khums), which was collected from moveable spoils taken in warfare and similar windfalls (such as mines and treasure troves), and which the Qur'an, 8:41, assigns to God and His Messenger (represented by the ruler), kinsmen (understood as the Messenger's), orphans, indigent people, and travellers.116 But the ruler might, and frequently did, assign further funds to charity, for the imam had to help every needy Muslim he heard of. The same was true of private people: the wealthy were guilty of sin if they neglected the poor among them.117 In the early centuries there seems to have been a conception of a system of regular government support for the poor and the invalid. This conception was rooted, not in the institution of obligatory alms or the fifth, but rather in the stipend system of the Umayyad period. Registered soldiers disabled in service would receive an invalidity pension, and some sources claim that similar pro- Hsions were made for the disabled and poor in general in Umayyad times. Thus al-Walld I allegedly assigned payments, in cash or kind (arzaq), to the : poor, the blind and the crippled (or lepers), or he supplied the blind with guides, giving servants to cripples, and paying pensions to all disabled people with pensions so as to spare them the need to beg.118 Others say that it was al-Walid II who gave money and clothes to the blind and cripples, supplying the latter with servants.119 Still others say that it was cUmar II who did so.lal There were also some who held the 'Abbasid caliph al-Mahdl to have been the first to give stipends to the poor, the maimed (or leprous) and others,121 though he defends the absence of funds for such people in another story: a Byzantine 116. Cf. Aghnides, Theories of Finance, 207ft., 407ft'., 461ft. In lmami law the khums was an income tax payable to the imam, who was meant to distribute it to the Qur'anic rccipienrs. 117. Juwayni, Ghiyatb, S338f. 118. YT, ii, 348; Tab., ii, 1196.3; clqd, iv, 424.4 (in Syria only?). For rhe hospirals wirh which he is also credited by Ya'qubi and others, see Conrad, 'First Islamic Hospital?'. 119. Tab., ii, 1754.5 (in Syria only), 1799.14; lbn Hamdun, Tadhkira, i, no. 1102. 120. lab., ii, 1367.T2; lbn al-JawzT, Sirat 'Umar b. cAbd at-'Aziz, i54f. i2i.SadusI, Hadhf, 12 (adds foundlings); lbn al-Athlr, Kdmil, vi, 57, yr 162 (adds prisoners). 308 ] Functions of Government emissary who came to Baghdad saw cripples begging on a bridge on his w al-Rusafa and commented that the ruler (then al-Mansiir) ought to do so thing about them; a secretary said that the revenues did not suffice for that h al-MahdT hotly denied this; according to him, the reason why the caliph did nothing was that he did not want to monopolize merit making: his subjects too should have a chance to gain rewards by giving to beggars. Needless to say th Byzantine accepted his point.122 This story takes us away from the Uma'yyad stipend system, but all seem to envisage the funds as coming from the taxes in general, not just from the obligatory alms or the fifth, and those reflectingtne stipend system rest on the conviction that the fay' was communal property Precisely how much truth there is to them is debatable, but they do testify t0 \ concept of what one might call a welfare state. The idea persisted, too. "Give pensions from the treasury to the blind" Tahir b. al-Husayn, the governor of Khurasan, advised his son in zo6/8iif when the latter was appointed to the governorship of Raqqa. The son was also advised to provide assistance to noble houses fallen on hard times,123 to look after the poor and destitute in general, to be careful not to overlook victims of oppression unable to complain to him and wretches ignorant of how to claim their rights, and to assign pensions to victims of calamities and the orphans and widows they leave behind "in imitation of the Commander of the Faithful".124 Similar advice is given in an Imami and an Ismaili testarm Both al-FarabT and Ibn STna assigned funds to cripples in the ideal city126 The ruler was also expected to provide help on a one-off basis, especially in times of famine. Pseudo-Aristotle told Alexander to examine the state of the weak ones in his realm, to assist them with funds from the treasury when they were starving, and to store up grain for distribution in years of drought.12 The ruler should store up food supplies for a year in every region for emergencies 122. Zubayr h. Bakkar, Muwaffaqiyydt, 68f.; Jahshiyari, Wuzara', 133; Ibn Hamdun, Tadhkira, ix, 110. 637. 123. Cf. above, 158, note 58. 124. lab., hi, 1058.14, 1059.2, 5. At 1054.14 he could be taken to say "Grant all Mus imi a share and portion of your fay"', by emendation of the word niyyatika to fafika (thus tilt Addenda et Emendanda, followed by Bosworth in The History of al-Tabari, xxxii, 119 .though differently in Arabic Mirror', 36J; also in Ibn Kbaldun, Muqaddima, 340.-5 = ii, 148). But it is not entirely convincing: the fay' was not "your fay"', nor were the normal words for a share in it hazz or nasth. 125. IAH, xvii, gtf.; tr. Chittiek, Anthology, 77; Nu'man, Da'd'im, i, no. ifS. Salinger, A Muslim Mirror for Princes', 37. 126. FarabI, Fusul, §62/66; Ibn STna, SI, x, 447 = 104, condemning the suggestion that incurable people should be killed. 127. Sirr al-asrdr, 8if. Functions of Government 309 Id move grain to needy provinces, Ibn al-Daya advised.128 Pseudo-GhazalT [jesses that the ruler was obliged to help his subjects when they fell into poverty and distress, especially in times of drought, specifying that he had to supply food and financial help alike, and stop oppression, since they would otherwise leave and thereby deplete his treasury.129 (Peasants were not legally tied to the land.) Al-Juwaynl advised rich people to store up grain for a year and give tr,e resc to trLe needy, for it fell on them to look after poor people overlooked by the imam; if burying the dead was a collective duty, a fortiori the same was true of preserving the lives of the poor.130 In practice, rulers did sometimes help communities stricken by disaster with measures such as the remission of taxes, importation of food, and charitable payments. Ensuring that everyone had enough to eat was after all a good way of keeping the subjects orderly'- Devastating famines are nonetheless repotted from time to time, and generally speaking, private charity seems to have played a far greater iole than government measures in the alleviation of poverty and misfortune. Medical services Tahir's letter of advice to his son, which was published to immense acclaim, included the statement, "Set up establishments (ditran) for sick Muslims where they can find shelter, and appoint custodians who will be kind to them and physicians who will treat their illnesses. Comply with their desires as long as it does not lead to the treasury being squandered."132 The lepers, cripples, and blind to whom al-Walid I, TJmar II, al-Walid II, or al-Mahdl were reputedly the first to assign stipends are not said to have received medical treatment, except in the claim, probably false, that al-Walid I founded a hospital.133 But from the late ninth century onwards hospitals were all the rage, and the money for them seems often to have come from the treasury. By the reign of al-Mu'tadid (d. 289/902) one had been founded at public expense in Baghdad.134 By 304/9161'. there may have been five. In 306/918 the pagan court physician Sinan b. Thabit added another two in the name of the caliph al-Muqtadir and his mother; in 311/923 the vizier Ibn al-Furat founded one as well. So too did 128. Ibn al-Daya, 'llhud, 34.-9, 35.-8. 129. NM(G), 101. 130. Ghiyath, §338-42. 131. Cf. Sirr al-asrdr, 82.5, and the dialogue between a Byzantine and a Persian king cited in Kraemer, Philosophy, 19. I 132. Tab., iii, 1059. 133. Conrad, 'First Islamic Hospital?'. I 134. Cf. below, note 150. 31° 1 Functions of Government the Büyid rulers Mu'izz al-Dawla in 355/965 and cAdud al-Dawla in 37I/ o ■ There were hospitals in tenth-century Basra,136 Shlräz, and Raw th» 1 ■ *1 u'e tnhabi- Sovernor one after Functions of Government 311 early tants of tenth-century Nishapur asked for one, hut were told by their that it was more than the treasury could bear, or maybe he did build all."7 Hospitals had also been established in Mecca and Medina by th tenth century, and a tenth- (or eleventh-) century Zaydi imam composed manual for censors instructing them, in words similar to Tahir's, that the ho * pital (ddr al-mardd) should have a skilled physician and that the expense' should be charged to the treasury, if it could bear it.133 In the western Musi' $ world it seems to have taken longer for hospitals to appear. All the hospitals of the tenth and eleventh centuries were founded by mem bers of the ruling elite, and like the libraries and other public amenities t0 which Iraqis above all were treated in the tenth and eleventh centuries they must have been a function of the competition for power in the politically frag mented and highly unstable world of the time. Why the competitors should have chosen to advertise their status in terms of hospitals, as opposed to some Muslim version of bread and circuses or, as in the Seljuq period madrasas is another question. But hospitals did have the advantage of giving the philosophers (who often made a living as doctors) a much-needed populist face. There was more to the medical craze than hospitals. At a time of much sickness in Iraq, 'All b. cIsa, 'the good vizier' (d. 334/946), wrote to the pagan court physician Sinan b. Thabit suggesting that some doctors be assigned the task of treating the inmates of prisons, who were bound to be riddled with diseases "in view of their numbers and the harshness of their whereabouts". This done, 'All b. 'Isa despatched another note, saying that there must also be sick people who went without treatment in the Sawad, the countryside of southern Iraq, so would Sinan send a mobile medical team to tour this area? Sinan complied. The team soon encountered the problem that not all the inhabitants of the Iraqi countryside were Muslims. Sinan inquired whether Jews were to be treated too, adding that in his own hospital treatment was given to Muslims and Christians alike. The answer was that it was certainly fight to treat dbimmii, and cattle too, but Muslims came first, dhimmis second, and 135. Mez, Renaissance, 377; El', s.v. 'bimaristan'. 136. Tanflkhl, Nishwär, viii, no. 102 = no. 81. 137. BP, s.v. 'bimaristan*; MuqaddasT, 300, 430; Ibn Bäbawayh, cVyün, ii, 286, no. tm The governor was Hamawayh b. cAiT, 138. El', s.v. 'bimäristln'; Serjeant, 'ZaidT Manual', 30.16, identifying the imam as al-Näsir li'l-Haqq (d. Daylam, 304/917). nimals last.139 Whether these justly famous services continued beyond the enute of the good vizier is unknown. In 3t9/'93I> however, the death of a patient by medical misadventure caused I Jvluqtadir to instruct his censor to prevent doctors in Baghdad from practising unless they had been authorized by the same Sinan b. Thabit, who was instructed to examine them. No fewer than 860-odd doctors are said to have shown up for the test.140 A similar story is told of the caliph al-Mustadf (566-75' 1170-80) and his Christian court physician Ibn al-Tilmldh.141 In Fatimid Cairo the court physician Ibn Ridwan (d. 453/1061) would have liked more state control of doctors along these lines. "I shall tell you some stories about these doctors, their deceit and ignorance, so that you will be wary of them," he told his reader. "The government (al-stdtdn) might look into their affairs and prevent anyone from making a living by this profession unless he is skilled, and single out the best of them for the rest to emulate."142 This was also the view of the Aleppan doctor al-Shayzarl (d. 589/1193) and the peripatetic'Abd al-I.atlf al-Baghdadl (d. 62.9/1237), who both had dealings with Saladin. According to al-Shayzan, the ancient Greeks had set up a chief physician in every city for the examination of doctors and ongoing quality control; doctors were required to provide written accounts of the treatment they prescribed and to give copies to their patients' relatives so that if a patient died, the chief phvsician could decide whether the doctor was at fault and, if he was, order him to pay compensation.143 According to al-BaghdadT, good practices still prevailed in Constantinople, where people took proper care that only competent physicians were allowed to practise by holding medical examinations and imposing the Hippocratic oath on the candidates, or so at least until the Franks conquered the city (in T2.04;, he said.1"4 Back in the good old days, Gteek rulers had also supported the development of new drugs, and tested them before releasing them, according to the eleventh-century al-Mubashshir b. Fatik; public funds were always available for that.14-' But nowadays was a different story. In the absence of a chief physician it was the mubtasib who had to 139. Ibn al-Qiff!, llukamx1, 1931. 140. Ibid., i9if. \ 141. Ibn AbT 'Usaybi'a, 'Uyun, i, 26if. 142. Gamal and Dols, Medieval Islamic Medicine, 24 = 123. Goitein, Mediterranean Society, ii, 247, claims that there was a licensing system for doctors in Egypt, adducing Ibn Ridwan (in Ibn Abi Usaybi'a); perhaps it was cAbd al-Lapf al-Baghdadi he had in mind (cf. below, note T47). 143. Shayzarl, Kihdyat al-rutba, 97ff. 144. Stern, 'Collection of Treatises by 'Abd al-Latif', 60. 145. Rosenthal, Classical Heritage, 35. 312 Functions of Government test the competence of doctors, armed with books by Galen, Paul of Aei Hunayn b. Ishaq, and others, according to al-Shayzarl; he would administet Hippocratic oath to them.14'' No such procedures are mentioned by a] Baghdadl. According to him, one needed a certificate signed by a leading man (or men) of the profession to practise as a physician in Cairo, Damascus 0r Baghdad, but chaos prevailed in Aleppo, where he was writing.147 AH in a|] rulers seem to have found it more attractive to found hospitals than to maintain control of doctors or their drugs. Education, culture Children learnt to read and write on the basis of the Qur'an in elementary schools (sing, kuttab) run by a single teacher. Higher education in subjects recognized as valid by the religious scholars took place in private homes or in mosques, where one would join a circle (halqa) of pupils around a particular scholar; from the tenth and eleventh century onwards one could also study in madrasas, residential colleges attached to mosques which specialized in law. For more rarified subjects such as philosophy (including science) one studied in private homes or on one's own, though one could also be attached to a hospital for the study of medicine. Most teachers charged a fee, and living expenses also had to be met. Schoolteachers were usually hired and paid by parents, and neither the Umayyad nor the 'Abbasid caliphs seem to have concerned themselves with education at so elementary a level, but later rulers sometimes did.148 According to the Hanbalite Abu Ya'la (d. 1066), the imam was obliged to provide basic religious instruction for his subjects, women included, and to make public funding available to both teachers and pupils; this, he said, was more important than waging jihad."9 Even so, it was more commonly to higher education that subsidies went. Mosques, madrasas and hospitals were usually paid for by waqfs, charitable endowments established by the ruler and other members of the elite with funding from private sources or the treasury; though some were maintained directly by the treasury150 The 146. ShayzarT, Nihaya t al-rutba, gSff. 147.Stern, 'Collection of Treatises by 'Abd al-I.arif, 60; cf. Goitcin, Mediterranean Society, ii, 247, for the certificate. 148. Nur al-DTn supplied elementary education for orphans (Elisseeff, 'Document contcmporain,' 138; cf. Talmon-Heller, 'Religion in the Public Sphere', 54). 149. Abu Ya'la, K. al-amr bi'l-mcfruf, fol. 114b (I owe this reference, along with a xerox of the passage, to Michael Cook). 150. Cf. El', s.v. 'wakf. The caliph al-Mu'tadid paid the running expenses of the Sa'idi hospital directly out of the treasury, at 450 dinars a month (Busse, 'Hofbudget', 19), and the Functions of Government 313 -teat libraries in which the Muslim world abounded in the tenth and eleventh nturies were usually also financed by charitable endowments. They were public in the sense of open to all. Many served learning of a type that the religious scholars disliked (notably philosophy and various types of ShT'ism), but no teaching seems to have gone on in them. What rulers and other magnates did above all, however, was to patronize poets, scholars, litterateurs, and philosophers and other men of learning once their education was completed. No list of the ruler's shaft duties mentions this function directly, but it was implicit in 'preservation of the religion': as the army guarded the religion with their swords, so the scholars guarded it with their arguments and proofs, and the imam had to ensure that both were adequately funded, as al-GhazalT said.151 Besides, rulers needed religious scholars, jurists, poets, litterateurs, and, when they wete thus inclined, philosophers for their own advice, edification and entertainment; and they also liked to have some control of what was being said, not only in the sense that poets and chroniclers could have great propaganda value, but also in the sense that they needed to gauge, and where necessary to influence, intellectual trends in order to stay on top. As patrons of learned men, rulers had a greater say in the shaping of the cultural orientation of Islam than anyone quite realized, not only because the court conferred wealth and prestige on those who frequented it, but also because some branches of learning depended on patronage for their very survival. Religious scholars could always find some source of income or other, however lowly, because they served needs common to all Muslims; but poets might find it more difficult to make a living, and only the elite found at courts appreciated the services of philosophers and tnutakalltms (who reacted to the competition for scarce resources by being at each others' throats). Conclusion All in all an amazing number of services were supplied by the ruling elite at various times. It was, however, rare for all of them to be supplied at the same time, for the costs were high and political stability was low. Of the Buyid ruler 'Adud al-Dawla (d. 372/983) we are told that he swept Baghdad clear of thieves, restored order in the troubled deserts of Arabia and Kerman, provided a quick news-service, dug wells on the pilgrim routes, constructed cisterns, and built a wall around Medina, renovated the half-ruined Baghdad, built mosques and undated budget in larlkb-i Slstan, 32 = 22, assigns ic.ooo dirhams a year to hospitals. The upkeep of prisons is considerably more costly in both budgets. 151. Fada'ih, 117 (ch. 9, on warn1). 3T4 Functions of Government bazaars, repaired the bridges over the great canals, put railings on the bridg-over the Tigris and appointed guards to it, made the wealthy repair dilapidated weirs, redug canals that had silted up and built mills along them, repair^ dams and transferred people to the land thus reclaimed, established a rich] endowed bazaar in Fars and made arrangements for the cultivation of fruits there, and introduced the cultivation of indigo in Kerman. On top of that we are told, he loved learning and gave stipends to theologians, jurists, philolo gists, physicians, mathematicians, and engineers, as well as to preachers and muezzins, and built a huge library in Shlraz. He also founded a hospital in Baghdad, made provisions for the poor and the strangers who lived in mosques, and gave away much money in charity His governors engaged in similar deeds.152 Or so we are told. One certainly would not characterize this as minimal government, which was what rulers normally dispensed in the medieval Islamic world or indeed in pre-modern times in general.153 But in return he is said to have extorted money in every conceivable way, and there may have been less to the result than advertised. People would complain the vizier Ibn Sa'dan (373-5/983-5) was told, of interference with their lives and property, disturbance of their tranquillity, and confiscation of their homes and land; they would lament that money was counterfeit, taxes doubled, business slack, soldiers and police swaggering about, mosques in ruins, hospitals desolate, enemies rabid, and more besides, quite apart from all the things thev did not dare to say for fear of the whip.1"4 What people really wanted rulers to supply was more and better government at lower cost, without violence or oppression. It sounds familiar. 152. Summarized from Mez, Renaissance, 25-7. 153- Goitein, 'Minority Selfrule', ioz; Crone, j're-industrial Societies, 49L, 57. 154. TawhTdT, Imlcf, iii, 88, quoting his teacher (who was patronized by 'Adud al-Dawla), in Kraemcr, Philosophy, 252. CHAPTER 19 VISIONS OF FREEDOM In practice, government was more often than not both weak and oppressive: weak in the sense that it could not get much done, oppressive in the sense that [ rulers would freely sacrifice the lives and property of their subjects in order to ! stay in power and keep some semblance of order. It was normal for members of the elite, scholars included, to spend time in jail; most high-ranking governors and generals died violent deaths; and torture, assassination, poisoning, confiscation, and extortion were matters of routine. Yet the desire for freedom remained. Not that medieval Muslims used that term. They did speak of political oppression as enslavement,1 but they did not call the opposite freedom, for the choice as they saw it was not between slavery and freedom, but rather between slavery to other human beings and slavery to God. No humans had the right to impose obligations on other humans, whether they were rulers, masters, fathers or husbands, or for that matter prophets; only God could do so.2 To be governed in accordance with God's rules was to be protected from other people's arbitrary desires {bawa). In other words, it was to live as an autonomous person under the law, which is also how political freedom has traditionally been understood in the West. Living in accordance with God's rules fwas what most Muslims desired. In practice, however, this freedom only 1. Above, 45t, 52; for posc-Umayyad examples, sec Ps.-NashP, §83 (of non-Muslim regimes); (Amin, Flam, 175 (of the Persians); Matin-Guzman, "Umar ibn Hafsun', 194 (of al-Andalus). 2. ChazalT in Laoust, Politique de Gazdli, T$yf. (citing the Mustasfa). f 315 1