CHAPTER 1 THE ORTGINS OF GOVERNMENT How did medieval Muslims think that humans had come to live under government? Differently put, how did they explain the origin of the state? The short answer is that they did not normally see government as having developed at all, but rather as having existed from the start. It is worth examining this answer in greater detail, however, for it brings out some of the most basic assumptions behind their political thought. It is to such fundamental concepts and ideas that this chapter is devoted. Terminology The word 'state' in modern parlance refers sometimes to a set of governmental institutions which constitute the supreme political authority within a given territory (as when we grumble about the state and wish that, it would wither away) and sometimes to a society endowed with such institutions, that is a politically organized society or polity (as in the expression 'nation state'). In the question of how the state originated, the emphasis is on the agency, but the two meanings arc closely related. Medieval Muslims had no word for states in either sense, however. They saw themselves as governed by persons rather than institutions and would speak of a ruler, such as a caliph (khalifa) or king (rnalik), where we speak of the state in the first sense of the word; and they would identify the society of which the ruler was in charge as a nation (umma) or a religious community (milla), where we speak of states in the second sense of the word. (The term kbiläfa, caliphate, only referred to the caliph's office, not to his polity, though modern scholars freely use the word in both senses.} The Muslims did pick up from their Greek forebears the habit of describing a [ 3 J 4 1 Origins of Government politically organized society as a city (Greek polls). They did not know that Greek cities had once been states themselves and that this was how the habit had originated. They simply continued it by using the word madina in the same way that their counterparts in Europe would use civitas, as a term for polities of any kind, in a close approximation to the modern word 'state' in the second sense. But madina in the sense of polity was a fairly arcane usage in the Muslim world, confined to the philosophers and the few who read them. To everyone else, it just meant a city in the plain sense of the word. The concept of the state as an impersonal institution emerged in Europe from the sixteenth century onwards and eventually passed to the Middle East. In the nineteenth century the Muslims gave it the Arabic name of dawla (Persian dowlat, Turkish devlet), and this is now the standard word in the Middle East for a state in both senses of the word. In pre-modern usage, however, dawla meant a turn of fortune (and of the stars in their spheres) and thus the era in which a particular dynasty held sway rather than the governmental institutions or the polity of which it was in charge.1 But though the pre-modern Muslims lacked the concept of the state, they certainly had governmental institutions which conform roughly (if rarely precisely) to the modern definition of states in the first sense of the word and which held sway in units that we would identify as states in the second sense. How then did they explain their origin? Adam and Eve When medieval Muslims pondered the question why government exists, they formulated their answer in functional terms: rulers performed such and such roles for which there was a need thanks to the nature of human beings (see below ch. 17). They rarely addressed the historical question how rulers had developed or when they had first appeared, but it is clear from their creation myth that they did not share the medieval Western view of government as a secondary development of human history rooted in the Fall. They tacitly assumed government had existed even before the creation of mankind. The relevant part of their creation myth may be summarized as follows. When God had created heaven, earth, the angels, and the jinn (i.e. spirits). He created IblTs (the future devil), who was the first to receive power {mulk).1 God made him ruler and governor of lower heaven and earth, as well as keeper of Paradise, or, according to another version, He made him judge among the jinn, who were the first inhabitants of earth and who had kings, prophets, religious For abbreviations, sec the bibliography. 1. Cf. Lewis, Political Language, chs 2-3. 2. Thus 'lab., i, 78.10. Origins of Government [ 5 faith, long life, and blessings in abundance. The jinn grew wicked and caused corruption on earth, whereupon IblTs sent an army against them and defeated them, which made him haughty; others say that IblTs was a captive taken by an army of angels sent against them by God and that he became haughty because he grew up among the angels as a result of his capture; or, according to another version, IblTs was so successful a judge among the jinn that he grew haughty and started fighting them. In any case, God knew that IblTs was growing haughty and created Adam to bring out his true colours. IblTs duly refused to bow down to Adam, whereupon he was cast into the lowest Hell. God then created Eve, but she was subverted by IblTs in the form of a snake and both she and Adam ate of the forbidden fruit, whereupon they were expelled from Paradise. Eve was punished with menstruation, pregnancy, childbirth, and stupidity; Adam accidentally brushed his head against heaven when he fell to earth (they do not simply walk out of Paradise in the Muslim version), and so he became bald; both suffered the indignity of having to defecate: Adam wept when he smelt the stench. But above all, they lost their freedom from work: they and their descendants now had to do all the "irksome ploughing, planting, irrigating, reaping, threshing, milling, kneading, spinning, weaving and washing" which they had been spared in Paradise.3 This was the crucial way in which the Fall affected the human condition. There was no forfeiture of immortality. Humans did not become more sinful than they had been from the start either, and human history did not turn into a story of Paradise lost and regained. In fact, many scholars denied that the Paradise from which Adam and Eve were expelled was identical with that in which God's righteous servants would eventually find themselves.4 Their fall notwithstanding, Adam and Eve continued to live a sub-Paradisical existence. Adam was God's deputy [khalifa) on earth (cf. Q. 2:30), where he had been given power and authority {mulk wa-sultan). Both he and his son Seth were also prophets through whom God revealed His law. When Adam died, he passed the leadership (riyasa) to Seth, and thereafter each leader passed his deathbed instructions (wasiyya) to his successor along with "the political governance and management of the subjects under his control" (siyasat al-tnulk wa-tadbtr man tahta yadayhi min ra'iyyatihi).s They lived a life of religious purity and piety, spending their time in worship of God without any impure thoughts or feelings of envy, hatred, or greed.6 3. R1S, ii, 229 = Goodman, Case, 73; Tab., i, 103, T29. For the antecedents of this idea, see below, note 10. 4. Cf. MawardT, A'lam al-nubuwwa, 781,; Ibn Babawayh, Ytiqad, 130 = 81 (ch. 29;, 5. Tab., 1, 165.1. 6. Thus YT, i, 5 (closely following the account in the Syriac Cave of Treasures). Origins of Government Origins of Government But when Cain killed Abel, he left the sub-Paradisical mountain on which ! Adam had settled to live somewhere else, where his descendants became 'despots and Pharaohs' (jababira wa-faraHna), that is, godless tyrants.' They invented musical instruments and took to entertaining themselves with music, wine-drinking, and sexual promiscuity. This caused an ever-growing number of Seth's descendants to leave the sacred mountain in order to join the fun. Enoch, alias Idris, and his son Methuselah both fought holy war against the Cainites and enslaved some of them, but to no avail: hardly any of Seth's descendants were left on the mountain by Noah's time, so God sent the Flood and wiped out the entire sinful lot. After the Flood, Noah's sons dispersed to become the ancestors of mankind as we know it. The fundamental assumption behind these accounts is that all the power in the universe and all the physical and moral laws by which it is regulated reflect the same ultimate reality, God. God rules in the most literal sense of the word, appointing rulers, governors, judges, and deputies and ordering armies to be sent against insubordinate subjects. Divine government has always been and always will be, and it must necessarily manifest itself as government on earth. Adam represents the fullness of God's power on earth, and both the Sethians and the Cainites are envisaged as living in politically organized societies, as are the jinn who preceded them. Contrary to what medieval Christians said, coercive government did not develop among humans as a result of the Fall. All God's created beings were subject to His government, directly or through intermediaries, whether they sinned or not, and divine government was certainly coercive. Of course God would not need to use violence against His subjects if they would obey, but all have a tendency to rebel, for reasons which the myth leaves unexplained; there is nothing special about humans in this respect: God sent armies against unruly jinn long before humans had been created. Disobedience, ma'siya, is the Muslim word for what the Christians call sin, and the archetypal act of disobedience is IblTs' refusal to bow down to Adam, not Adam and Eve's eating of the forbidden fruit, which only plays a limited role in the Muslim explanation of the human condition and none at all in the Muslim account of the origin of states. Right and wrong government (imam a and mulkj Government was an inescapable feature of the universe. But not all government was right, and sin certainly played a role in its corruption. The key event here is Cain's murder of Abel. Having killed his brother, Cam left Adam's community to found one of his own, in which his descendants came to be ruled by godless tyrants. As we have seen, the creation myth calls them jababira wa-farcfina, despots and Pharaohs. Another word for such rulers was muliik (sing, malik), kings. Either way, they were rulers who seized power for their own aggrandizement rather than the execution of God's will, turning God's slaves (that is, human beings) into slaves of their own and using their power for the satisfaction of private interests rather than the fulfilment of collective needs. To call a man a king was not necessarily to denounce him. A malik was simply somebody who lorded it over others, especially one who did so sitting on a throne and wearing a crown. You could describe a ruler as a king in neutral or flattering terms, but in the first centuries of Islam you could only do so as long as you were speaking in a secular vein. As far as religious language was concerned, the only being to whom you could legitimately apply the awesome titles of king (malik) and despot [jabbar) was God. One could not question the overweening power of ultimate reality, but it was both presumptuous and rebellious for humans to claim such power for themselves, and those who did so merely branded themselves as kings in the sense of impious tyrants. Yet claim it they often did: the despots of the Cainites were wiped out in the Flood, but they reappeared in Pharaonic Egypt, to be followed by Greek, Roman, Byzantine, Persian, Indian, and many other kings. The vast majority of humans had been, and continued to be, governed by wrongful rulers of this kind. There was also another form of corruption, though it does not appear in the creation myth. After the Flood some people stopped having government altogether, as the Muslims knew from their own history: Islam had originated in a stateless society. The Muslims took no pride in this aspect of their past, at least not when they were religious scholars,8 for in their view the absence of government in pre-Islamic Arabia reflected the failure of the pagan Arabs to acknowledge God. As pagans, the Arabs had lived in jahiliyya, ignorance and barbarism, not in a state of aboriginal freedom and equality such as that which the Greeks and their Western epigones were apt to impute to tribal peoples. Without recognition of God's sole government there could be no proper relations among people, only tyranny or anarchy, with all the bloodshed, arbitrariness, and immorality that both implied. Adam embodied the alternative to tyranny and anarchy alike. His leadership had been imdma, religious and political leadership in accordance with God's will. As khalifat allah fi 'l-ard, God's deputy on earth, he had dispensed God's law among his offspring as they multiplied. This was what government 7. Tab., i, 167.14. 8. For the attitude of specialists in tribal lore, see below, z6Si. 8 ] Origins of Government should be and what it had remained among the Sethians down to the Flood. Thereafter right government had existed only sporadically, but it had been restored to the world by Muhammad, the prophet who founded the Muslim polity, and it had been maintained (according to most Muslims) by his immediate successors, the first caliphs in Medina. It was their time which constituted the golden age. As always, proper government was coercive. Like Adam's successors, Muhammad and the first caliphs had used institutionalized violence: they imposed penalties, suppressed revolts, and conducted campaigns against infidels. But they always did so in accordance with God's law; their coercive power was wielded only against evil-doers. This was the essence of good government, and it was such government, not a pre-political stage of alleged freedom and equality, that the most Muslims hankered for and hoped to restore. The law (shar1, sharTay As the ruler of the universe, God issued laws. Adam had received a set of them; so had later prophets, most recently Muhammad, whose version was final. Living in accordance with God's law was the essence of religion. In the early centuries it was practically all there was to religion. There soon came to be so much more to it that some would have liked to jettison the law altogether, much as the early Christians had done; but it remained the heart of Islam in all its forms.9 The word shar* was often used to mean revealed religion in general. What medieval Muslims regarded as law included much that modern students have trouble recognizing as such. A traditional handbook of Islamic law will start with the 'ibadat, 'acts of service/worship',10 that is to say the five daily prayers, the month of fasting, the annual alms tax, and the pilgrimage to Mecca once (or more) in a lifetime for those who were capable of making it. Dietary law (rules about permissible and prohibited food and drink) also formed part of the 'ibadat, though its positioning in the lawbooks was unstable. None of this is law to a modern Westerner. Next, the manuals moved on to the mu'amalat, 'mutual dealings', meaning people's relations with one another. Here they regulated marriage, divorce, 9. Cf. below, ch. 15. 10. In the ancient Near East it was by labour services that humans did the will of the gods, who had created them as their slaves in order to save themselves the trouble of procuring their own food and housing. (From the dawn of history in Mesopotamia, it was unremitting hard work that people saw as the distinguishing feature of human, or at least civilized, life.) This is the ultimate root of the Christian concept of worship as 'liturgy' (from leitourgia, public service or works} and the Muslims concept of it as 'ibaddt. Origins of Government [ 9 inheritance, slavery and manumission, commerce, torts, crimes, war, taxation, and more besides, all of which the modern Westerner instantly recognizes as law on the grounds that rules of this type are enforced by the authorities. But the 'ibadat might also be enforced by the authorities (attending Friday prayer or fasting in Ramadan was not a matter of choice). Conversely numerous rules counted as law even though they lacked this feature, for the law extended into areas such as filial piety, the proprieties of clothing, behaviour at funerals, how to greet non-Muslims, and other matters that a modern Westerner would treat as purely moral, or as mere etiquette (and which in fact were not usually covered in the legal manuals, but rather in separate works). What distinguished a law from other rules was not that it could be enforced by the authorities, but rathet that it defined the moral status of an act in the eyes of God. The key question to which the law provided answers was how far doing something would assist or impede the journey to salvation, not whether it was allowed or forbidden in the here and now. Assessing the moral status of human acts was the work of the jurists (fuqaha*). They classified human acts as either forbidden or permitted, and, within the latter category, as disapproved, indifferent, commendable or obligatory, trying to work out God's view of them on the basis of the Qur'an and statements by the Prophet, plus some subsidiary sources. It was permitted to repudiate a wife, provided that the rules were followed, but it was not commendable; it was commendable to free a Muslim slave, but it was not obligatory; it was normally indifferent whether one wore this type of clothing or that, provided that the rules of modesty were satisfied and no silk was worn by men, and so on. The jurists did take an interest in how far the moral assessments should be backed by coercive power, but it was to God and His Prophet as represented in the here and now by the jurists that the law owed its authority, not to the rulers. Prophethood (nubuwwa) God revealed His laws to mankind by means of prophets. Adam, the first man on earth, was not just an imam but also a prophet, and the same was true of Muhammad. Prophets play a key role in Islamic political thought. A prophet (nabl) was a human being through whom God communicated with mankind, or more commonly with some subdivision of it (usually a people). He was not primarily someone who could predict the future, though he might be able to do that too; rather, he was a transmitter of God's wishes.11 rr. Predictive abilities came ro be regarded as one of the ways in which the line of transmission was authenticated. io ] Origins of Government Such mouthpieces were required because humans no longer knew God directly; as they had done in Paradise, and as the angels still did. Cut off from God, humans would corrupt His religion and government until nothing remained of either, but in His mercy God would respond by sending them a prophet to inform or remind them of the true way. Every time God selected a person for prophethood (nubuwwa), a window onto the unseen was opened up and a glimpse of ultimate reality was transmitted to the earth. After Muhammad's death the window was shut and so it would remain until the end of times, but it had been opened many times before him, maybe as many as iz4,ooo times.12 Most of the 124,000 prophets before Muhammad were merely sent to warn particular communities against their evil ways, but some brought a new version of God's law, and thus a new religion. The latter, numbering 315 at the most,13 were prophets of the type called messengers (sing, rasiil). Messengers would found polities, for a law requires government tor its realization. I his is why they were of great political importance. Some 1,400 years ago God in His mercy sent a messenger to the Arabs. He chose Muhammad, a trader who was born in Mecca in c. 570 and who began to preach when he was about forty in response to periodic revelations brought by Gabriel. (Unlike Moses he did not speak to God directly, and he also differed from Moses in receiving his revelation in instalments rather than all in one.) Some people converted, but most Meccans reacted with sneers, ridicule, and eventually persecution of Muhammad and his followers, many of whom were people in a weak position, such as slaves and freedmen who had no kinsmen to defend them. When things became intolerable in Mecca, Muhammad and his followers emigrated to Medina. His emigration (hijra) took place in 622., which later came to serve as the starting point of the Muslim calendar. When he arrived in his abode of emigration (dar al-hijra), he set about forming a community {umma) there with himself as leader, and took to consolidating his position with caravan-raiding, military expeditions, and battles with the Meccans, whose city he conquered in 630. He died in Medina in 632, whereupon his followers began the conquest of the Middle Kast. This, in a nutshell, is the story of how the Muslim community was founded. Origins of Government [it the meek and who was persecuted for his faith. But they usually react with bewilderment to the second half, in which he goes off to found a polity in Medina instead of suffering martyrdom. Since modern Christians (and ex-Christians) typically think of religion as something transcending politics and other mundane pursuits, Muhammad comes across to them as having abused religion to make a success of himself. This is why Western scholars have in all seriousness debated the absurd question whether he was sincere or not.14 Medieval Muslims did not generally see religion as above politics and other worldly affairs, but on the contrary as a prescription for their regulation. They granted that it could be abused, of course, and held numerous 'false prophets' to be guilty on that score. But they took the abuse to lie in the falsehood of the claims advanced by such prophets, not in the worldly use to which the claims were put, except in the sense that they resented the benefit that such prophets derived from them; for religion was actually meant to put things right for people in this world no less than the next, and it stood to reason that the bearer of the true religion had to acquire political power in order to bring this about. The more power you have, the more good you can do. Some prophets were assisted by worldly rulers, but others, such as Moses, David, Solomon, and Muhammad, acquired worldly power themselves.1' In Mecca, Muhammad had been constrained by the pagan power of Quraysh, but in Medina he had gained the power to execute God's law and to embark on warfare.16 God had allowed Muhammad to unite prophethood and kingship (in the flattering sense of great political power) so that he could accomplish his mission, see to the execution of the law, and overcome the infidels, as al-Tha'alib! (d. 429/1038) put it.17 It is not that Muhammad was desirous of this world, the tenth-century Ismaili philosophers known as the Brethren of Sincerity explained, but God wanted Muhammad's community to have religion and this world together; and when the Jews and Christians found this hatd to understand, God sent down the story of David and Solomon, in whom kingship and prophethood had similarly been united without their prophethood being degraded thereby, so that the Muslims could argue their case with reference to them.18 Religion and politics Westerners do not normally have any problems with the first part of this story, which follows a familiar model: Muhammad was a prophet who preached to 12. Tab., i, 152; Ibn Babawayh, l'tiqad, 137 = 92 (ch. 35); cf. Wensinck, 'Muhammed und die Propheten', 169ft. 13. Wensinck, 'Muhammed und die Propheten1, 171. 14. The question was, however, put to them by their sources: it was Muslim freethinkers who first dismissed Muhammad as a trickster who abused religion for worldly ends (cf. below, 172L). 15. JuwaynT, Ghiyath, §267. 16. MawardT, A'lam al-nubuwwa, 316ft. 17. Tha'alibT, Ghurar, 4 [where Adam, Joseph, David, Solomon, and Alexander are said similarly to have combined prophethood and kingship). 18. R/S, iii, 496. By the thirteenth century people wondered why Muhammad had not been more of a king: "If kingship and sovereignty have so many advantages, and are means Origins of Government Origins of Government All this is religious polemics, of course, here concerned with the moral status of power in the service of the truth. In terms of historical explanation -why and how things happen, whatever one's moral evaluation of them — the fact that Muhammad was a political prophet is clearly related to the stateless environment in which he was active. The Buddha and Jesus had both lived in societies in which political authority already existed so that it was possible, prudent, or positively liberating to leave politics alone. Both preached messages which were largely about how to transcend politics along with everything else in this world: the Buddha is said to have been a prince who renounced his kingdom; Jesus was a carpenter who declared his kingdom to be not of this world. But Muhammad was active among warring tribes and had to take political and military action if he was to accomplish his mission. The religion could not survive without communal embodiment, and the community could not survive without defence. Hence it had to have political organization. To put this point at greater length, in tribal Arabia all free males protected themselves and their dependents in cooperation with their kinsmen by threatening to avenge any injury inflicted on them: there were no other ways of insuring oneself against murder, assault, robbery, theft, and the like. It is true that some Arabs had been incorporated in the Byzantine and Sasanid empires, in which there were armies and police, and that imperial subsidies had enabled others to develop petty kingdoms of their own, but most Arabs lived under conditions of self-help. Chiefs should not be envisaged as petty kings. Their role was to keep their tribes together by engaging in dispute settlement, helping the needy, and presiding over discussions of public issues in which all or most male adult tribesmen would participate and in which the chief would formulate the consensus as he saw it emerge. Chiefs might or might not be 1 military leaders too, but they could neither coerce nor protect their fellow- I tribesmen after the fashion of kings. Every tribesman defended himself and his dependents. The sources say that when Muhammad and his followers adopted their new religion, they severely tested the loyalty of their fellow-tribesmen and (in the case of slaves and freedmcn) their masters and protectors, so that it was clear that the latter's cooperation would eventually be withdrawn. That would have left the Muslims in the position of outlaws, and as such they probably would nor have survived for long. But Muhammad hoped that another tribe could be persuaded to adopt them, and those of Yathnb (the later Medina) did eventually agree to afford them protection, apparently because statelessness in their case manifested itself in the form of endless feuds: they wanted peace and hoped that a man of God could provide it. They did not offer him protection as a host to a guest, but rather joined Muhammad's followers to form a new community in Medina in which all members defended one another as if they were kinsmen. It could be said that Muhammad created a new tribe, a super-tribe of believers; but it was led by a prophet with powers unknown to tribal chiefs. "Whenever you disagree about something, the matter should be referred to God and Muhammad," as Muhammad laid down in the document generally known as the Constitution of Medina.19 Muhammad was the ultimate decision-maker. His community was a politically organized society, if only in a minimal sense. By claiming divine authority he had created an embryonic state. Thanks to the environment in which it originated, Islam was thus embodied in a political organization almost from the start: the umma was a congregation and a state rolled together. Christians originated with dual membership. As believers they belonged to the church and were administered by the clergy; as citizens they belonged to the Roman empire and were ruled by-Caesar. Islam originated without this bifurcation. As believers and as citizens they were members of the umma and ruled by the Prophet, thereafter by his successors. Thanks to Muhammad's career, Muslims came to think of prophets as the paradigmatic founders of states. Far from being assumed normally to transcend political organization, messengers of God were assumed normally to create it. The well-known fact that pagans also had polities was not normally perceived as a problem in this connection. The tenth-century Ismaili philosopher Abu Ya'qub al-SijistanT observes that kingdoms are to be found all over the world whereas prophethood has only flourished in a small part of it, but he nonetheless insists that kingship arises by usurpation of leadership established by a prophet. This was indeed how kingship arose in the Islamic world, but it is hard to sec how it could account for the kingdoms of the Indians, Chinese, Turks, Slavs, Africans and other peoples, to whom no prophets had been sent according to his own explicit statement.20 It took the fourteenth-century Ibn Khaldun to point out, in polemics against the philosophers, that in purely historical terms it was kings rather than prophets who were the paradigmatic founders of states: most of the world's inhabitants had rulers even though they had not received any prophets; and contrary to what the philosophers said, for drawing near and gaining closeness to God, then why was the kingship of this world not given to the Prophet in the same perfection that it was given to Solomon?" (Daya, God's Bondsmen, 406). 19. Ibn Hisham, Sira, i, 503.10; tr. Wensinck, Muhammad and the Jews of Medina, 56. 20. Itbbäl, 172, 174. J4 . Origins of Government authority could be established without religious law, for people in power cou and did devise injunctions of their own.21 In short, the philosophers had been wrong to generalize on the basis of the Muslim experience. But Ibn Khaldun's ability to see the formation of his own polity as an exception to the norm was highly unusual. Restating the question We are now in a position to restate the question with which we began. How did medieval Muslims imagine the state to have originated? Clearly, the question is misformulated. As medieval Muslims saw it, government was the inseparable companion of monotheism, and since humans had originated in a monotheist polity, the problem was not how they had come to live in states but rather why government had so often been corrupted thereafter, or disappeared altogether. The answer was that human disobedience repeatedly caused things to go wrong so that God had to send messengers to set things right again. But why-had God implanted this propensity for disobedience in human beings, or indeed all his creatures? And why did He send messengers to some people and not to others? Such questions seem to have been regarded as beyond human undemanding. The problematic fact that some people had government even though they had not received prophets was normally left as a loose end. The unusual nature of the paradigm But lbn Khaldun was right: most polities in history have indeed been founded by men who accumulated power from below rather than by prophets who received it from above. Consequently, most polities in history have also been characterized by a distinction between the political and religious spheres rather than by their fusion. There was no fusion of the religious and the political spheres in the complex societies of the Middle East that the Muslims were to conquer, nor had there been as far back as people could remember. It is true that Hellenistic kings and, following them, the Byzantine and the Persian emperors were credited with power over all things material and spiritual, but the fusion was limited to the ruler; it did not obtain in the society beneath him. Thus the subjects of Hellenistic kings were not expected to worship the same gods or to follow the same laws, except in the area pertaining to the shared government; and though both the Byzantines and the Persians did expect Origins of Government J-5 ejj ous uniformity, they had to cope with an ecclesiastical organization distinct from their governmental agency beneath them. In Sasanid Iran, the two encies were held to be twins; in the Byzantine case, they were unrclarcd partners Either way, they were separate, despite the monarch's fullness of power. What is so striking about early Muslim society is precisely that it started out without such a separation. The monarch's fullness of power here reflected a complete fusion of the religious and the political all the way through: there was no religious community separate from the politically organized society, and no ecclesiastical hierarchy separate from rhe political agency. One has to go all the way back to the ancient Near East to find a situation comparable to that of early Islam. The Sumerians may have started their history with temple communities ruled by priests alone. But that was thousands of years ago by Muhammad's time, nobody in the Middle East remembered it, and modern scholars usually deny it.22 The fusion (as opposed to blurring) of political and religious communities has not in fact been common in the history of complex societies at all, be it in the Middle East or elsewhere. Complex societies are usually much too differentiated in social, economic, intellectual, and cultural terms to tolerate the incorporation of all their interests in a single structure; and their rulers are usually much too coercive and rapacious for their subjects to have any desire to entrust them with the ultimate meaning of their lives. The Muslims were soon to find this out for themselves. The simplicity of Muhammad's all-purpose community matched the undifferentiated nature of the tribal society in which he was active. Like everyone else in rhe Middle East, the Arabs were affected by the Hellenistic concept of kings as endowed with a fullness of power. ('Hellenistic' here is a shorthand for a mixture of ancient Near Eastern, Persian, and Greek ideas.) It shows in their conception of the caliph.23 But there is only one real precedent for their all-purpose community in Middle Eastern history, and that is the federation of Israelites that Moses took out of Egvpt for the conquest of Palestine (which he, like Muhammad, did not live to see). Moses was a prophet and statesman like Muhammad, and he is also the paradigmatic prophet in the QuPan. Wittingly or unwittingly, Muhammad was a new Moses. Like him, he united scattered tribes in the name of God and led them on to conquest (though for one reason or another the conquest continued far beyond Palestine this time round). Moses was an inspiration to many people, but his admirers were not usually able to imitate him in any literal way since they lived under 2.1. Muqaddima, 48, 212. = i, 93, 390; cf. Nagel, Staat und Glaubensgemeinschaft, ii, 100; below, 268. 22. Cf. Crone, 'The Tribe and the State', 58ft. 2,3. Cf. below, 34, 4of. i6 Origins of Government such utterly different conditions. This was where Muhammad differed. He could and did re-enact Moses' career. Thanks to the astonishing military success of the people he united, an antique model of prophetic state formation developed by a minor tribal people of the ancient Near East thus acquired paradigmatic status in the complex society of the medieval Middle East in which Islamic civilization took shape. It is this starting point which gives Muslim political thought so different a character from that of its counterpart in the West. The Christians, as noted already, started with the conviction that truth (cognitive and moral) and political power belonged to separate spheres. Ultimately, of course, both originated with God, but they had appeared at different points in history, they regulated different aspects of life, and though they had to be coordinated for the common good, they could not be fully identified. The Muslims started with the opposite conviction: truth and power appeared at the same times in history and regulated the same aspects of life, more precisely all of them. It was a conviction that the post-conquest developments were soon to make untenable, but which was nonetheless difficult to give up. The result was an intense debate along utterly new lines dictated by the unusual starting point. In medieval Europe, where religious authority and political power were embodied in different institutions, the disagreement over their relationship took the form of a protracted controversy over the relationship between church and state. But in the medieval Middle East, religious authority and political power were embodied in a single multi-purpose institution, Muhammad's umma. Here, then, the disagreement took the form of a protracted controversy over the nature and function of the leadership of the umma, that is the imamate. CHAPTER THE FIRST ClVll, WAR AND SECT FORMATION The reader is warned that there are a lot of names, dates, and Arabic terms in this chapter. The first four caliphs, the first civil war, and its aftermath form part of the elementary vocabulary without which one cannot even begin to understand what medieval Muslims said about government. What follows is an attempt to serve the requisite knowledge in as short and simple a manner as possible. The succession to the Prophet \X'e saw in the previous chapter that the leader of Muhammad's community (umma) was called the imam. The dictionaries define an imam as somebody to be imitated, whether head of state or not. A simple prayer leader was an imam: you stood behind him and did as he did in performing the ritual prayer. Other righteous leaders were imams too: one modelled oneself on what they said and did. Great scholars, for example, came to be known by that title. But the head of state was the supreme imam. His imitators were not merely a small group of people at prayer or a major school founded by a great scholar, but rather the entire community of believers, the entire umma. At some point his leadership was dubbed 'the great imamate' (al-imama al-kubral'upna) to distinguish it from leadership of other types.1 Unless otherwise specified, the imamate always means the great imamate in this book. I. The earliest attestations known to me are Abu Yacla, Ahkdm, 24.-10; BaqillanT, Tamhtd, 183.-4; Naysaburl, lmama, 60.12., 61.2. [i? ] M4S^VKOVA,-Xfv fakulra so™!„m. „. ER2FTA JoStov, Hi W2 00 B R Jv" o 2O r.8 First Civil War and Sect Formation The first imam was the Prophet himself.2 When he died in 632, the imamate passed to Abu Bakr (632-4), an early convert and emigrant from Mecca who was of the same tribe as the Prophet - the tribe known as Quraysh. We are told that he was elected at a public meeting at the initiative of another companion, TJmar; some sources describe his election as a kind of coup.3 When Muhammad died, we are told, his Medinese supporters, known as the Helpers (Ansar), wanted to dissolve their union with Quraysh by electing a ruler of their own: "we'll have our leader and Quraysh can have theirs," they said. But TJmar got wind of their plans and foiled them by gatecrashing their meeting and pushing through the election of Abu Bakr as leader of the undivided community in a sequence of events which came to be known as 'the Day of the Portico' (yawm al-saqlfa). Whatever the truth of all this, Abu Bakr's position as leader of the community did not rest on prophethood, so he had to identify it in other ways and he is said to have adopted the title of khalifa (caliph), meaning deputy or successor. One would assume this title to have stood for khalifat allah, 'deputy of God', the expression that God uses of Adam and David in the Qur'an (2:20; 38:26) and which is attested for the caliphs from the 640s or 650s onwards.4 But the religious scholars deny it. According to them, it stood for khalifat rasiil allah, 'successor of the messenger of God', a more modest title which no caliph is on record as having used until the early cAbbasid period (750 onwards), except for Abu Bakr, who is said by the scholars to have adopted it in this version. The scholars, who emerged in the course of the century after Abu Bakr's death, disliked the grander title. They also denied that the caliph of their own time should be accepted as a guide in religious matters. Presumably, their insistence that the more modest form of the title was the original one reflects their attempt to cut the caliphs down to size (which will figure prominently in what follows). It has to be said, however, that modern scholars usually accept the scholars' claim at face value.5 Abu Bakr was succeeded by the above-mentioned cUmar (634-44), wh° was also an early convert of Quraysh. This time there was no crisis. Abu Bakr had designated cUmar as his successor, and/or TJmar's standing was such that his succession was a foregone conclusion. He called himself amir al-mu'minin, 'Commander of the Faithful', which became a standard title of the imam along with khalifa (however completed). The sources also credit TJmar with the 1. Cf. Kaimayt, Hasimijjat, 1:46; BF, 454.9; Ibn al-Muqaffac in Kurd CAIT, Rasayil, 111.-5; Jahiz, 'Maqalat al-7aydiyya waTratida', Rasiil, iv, 321. 3. Cf. Madelung, Succession, ch. t. 4. the first caliph known to have used it is TJthman (644-56) cf. Crone and Hinds, God's Caliph, 6ff. 5. For all this, see Crone and Hinds, God's Caliph, ch. 2. e byname <> some ad ,-1 1:-:..., u First Civil War and Sect Formation 19 of al-Faruq, a loanword from Aramaic meaning 'redeemer'. There is dditional evidence that TJmar was once cast in a messianic role, but tradition has forgotten about it and it plays no role in later political thought.6 TJmar's byname is accordingly explained as an Arabic word meaning a person good at distinguishing right from wrong, and when the messiah reappears, he does so under the native Arabic label of al-mahdi: It was under TJmar that the conquest of Syria, F.gypt, Iraq and Iran was achieved. Much wealrh and numerous captives were brought to Medina. Among them was an Iranian slave who developed a personal grievance against TJmar. In 644 TJmar was stabbed by this slave and died after having instructed the main contenders for the succession, of whom there are usually said to have been six, to choose a caliph from among themselves. This method of election, known as shurd (consultation), was designed to prevent the contenders from fighting it out among themselves.8 Armed conflict was narrowly avoided when the members of the shurd agreed on TJthman (644-56), yet another Qurashl' and early convert. But it was under him that things began to go seriously wrong. Large numbers of Arab tribesmen had now settled in Syria, Egypt, and Iraq. The provinces had not so far played a role in high politics in Medina, but it was above all among the Arabs in the provinces that there was dissatisfaction with TJthman, and in 656 delegates from Egypt and Iraq went to Medina to complain of his methods of government. They ended up by killing him. This act split the community down the middle, with reverberations that continue to this day. The killing of TJthman was a deed of the same order as the execution of Chatles I in the English Civil War, or of Louis XVI in the French Revolution. Unlike the two European kings, TJthman was not subjected to a formal trial, and his killers did not proceed to abolish the monarchy either: instead they elevated Muhammad's cousin, 'All, to the throne. But the fact that disgruntled subjects should have claimed the right to take the life of their monarch was as shocking to the early Muslims as it was to the later Europeans.10 By what right had they acted? The answer took the form of an intense debate, not over the relative rights of rulers and subjects, but rather over the moral status of TJthman himself: if TJthman had been a rightly guided imam, the rebels had 6. Except perhaps in al-Mukhtär's revolt; cf. below, ch. 7. 7. Cf. Crone and Cook, Ha%arism, 5; Bashear, 'The Title "Färüq"'; EI-, s.v. 'mahdf; Donner, 'La question du messianisme'; below, ch. 7. 8. EI1, s.v.; Crone, 'Sbürä as an Elective Institution'. 9. This is the normal Arabic form, not 'Qurayshl', a medieval variant used by modern scholars on the assumption that the reader would find 'Qurashl' confusing. 10. For the Europeans, see Walzer, Regicide and Revolution. 20 ] First Civil War and Sect Formation been wrong to kill him; but if he had forfeited his imamate by his 'innovations {ahdath), as his misdeeds were usually called, then the rebels had been entitled or indeed obliged, to remove him by force, inasmuch as he had refused to abdicate. The vast majority of Muslims adopted what one might call a toyalist position: TJthman had remained a legitimate ruler till the end; it had been! wrong to kill him; he should now be avenged and a new caliph elected by consultation (shurd). Adherents of this position were known as 'UthmanTs or shi'at 'Uthman, 'Uthman's party The rebels adopted the opposite position: 'Uthman had violated the law in a manner incompatible with his status as imam and had been killed as a wrongdoer (zaliman), not as somebody wronged (mazluman)-'All was now the legitimate imam and everyone ought to obey him. Adherents of this position were known as 'AlawTs or shi'at 'Ali, 'All's party.11 'All's caliphate is coterminous with the first civil war (656-61). He began by leaving Medina for Iraq, where he established himself at Kufa and defeated a section of the "Uthmanls in the so-called Battle of the Camel in 657. The leaders of these 'Uthmanls were two distinguished Companions of the Prophet, Talha and al-Zubayr, both of whom fell in the battle, and Muhammad's widow 'A'isha, who was sent back to Medina. But 'All could not defeat the other section of 'Uthmanls, which was headed by Mu'awiya b. Abi Sufyan, a QurashT of the same Umayyad lineage as 'Uthman. Mu'awiya lacked the distinction of early conversion, but as a kinsman of 'Uthman's he was well placed to demand vengeance for him; and having long been governor of Syria, he had a good power base. In 658 he met 'All in the famous battle at SiffTn in northern Mesopotamia, which the Syrians claimed to have won,12 but which the Iraqis say they would have won if the Syrians had not cleverly called a halt to the fighting by hoisting Qurians on their lances, signalling that the dispute should be submitted to arbitration. The Iraqi tradition further claims that the call for arbitration split 'All's party into two: when he accepted the call for arbitration some remained loyal to him, but others left his camp, protesting that "judgement belongs to God alone" (la hukma ilia li 'llak). These dissenters came to be known as Kharijites. Whatever the truth of all this, the Kharijites proved to be 'All's downfall: he was assassinated by one of them in 661. His followers then paid allegiance to his son al-Hasan, but al-Hasan stepped down when Mu'awiya claimed the caliphate, and in 661, the so-called 'year of unity' {'am al-jama'a), Mu'awiya was generally recognized as caliph. He moved the capital to Syria and founded the Umayyad dynasty, which lasted down to 750. Muslims are today divided into Kharijites (less than 1 percent), Shfites (c. 10 percent), and SunnTs, an amalgamation of earlier groups {close to 90 n. For all this, sec El', s.v. "Uthmaniyya'. 12.. See the references in Crone and Hinds, God's Caliph, 69, note 67; add Tab., ii, i^i. First Civil War and Sect Formation [ 21 cent) and the key issue that divides them is the legitimate leadership of the Muslim community after the Prophet's death: who was entitled to the lma-niate? This was endlessly discussed with reference to the participants in the first civil war, and eventually the caliphs before it too, so that by the time the sources available to us began to be compiled (roughly a century after the events) the historical events had been through too many polemical mills to be retrievable today; (In fact, all narratives relating to pre-Umayyad history in this book arc given without commitment to their historical truth unless the contrary is specified.) But why did the events get to be so disputed? How did the e to generate sects'! That is the question to ammad com succession to Muh which the rest of this chapter is devoted. The imamate and salvation A modern reader has trouble seeing how the civil war could generate sects, for it comes across as a purely political conflict: the protagonists were rivals for power; no disagreement over religious doctrine was involved, except perhaps in the opaque case of the Kharijites. For this reason it is commonly said that the parties or at least 'All's patty, only acquired a religious dimension at a later stage.1-But this is difficult to accept. More probably, we should correct our modern perspective. It is perfectly true that the civil war did not confront the participants with a choice between different articles of faith or concepts of spirituality, but this does not mean that it was purely political. On the contrary, it went to the heart of religion as understood by Muslims at the time in that it confronted them with a fatal uncertainty over the whereabouts of the path to salvation. To understand this, it helps to envisage the community (umma) as a caravan. The early Muslims saw life as a journey through a perilous desert in which one could all too easily go astray and perish. To survive, one needed to band together under the leadership of a guide (imam al-huda, hadi, mahdi) who knew the right paths, often called the paths of guidance (manahij or suhul al-huddlal-mshdial-rashad), that is the right things to do: the terms sunna (normative custom), sira (exemplary behaviour), and shari'a (Islamic law) arc all derived from roots to do with travelling and roads. The imam performed two tasks indispensable for the achievement of salvation. First and most fundamentally, he gave legal existence to the umma. Without him there was no caravan, only scattered travellers; they became a community by theit agreement to travel under him. Hence one could not be a 13. E.g. Sanhoury, Califat, 7411.; and, most recently, Halm, Shi'a Islam, 6, 16 (at first the ShT'ites were "merely a party in the struggle for power"; there "was no religious aspect to Shrism prior to 680"). 22 ] First Civil War and Sect Formation member of the Muslim community without declaring allegiance to its leader in the Prophet's days one converted to Islam by paying allegiance to him (with a handclasp), be it in person or via tribal delegates; thereafter one remained a member of the umma by paying allegiance to his successor, in person or via his governors. The Prophet is credited with the statement that "he who dies with-out an imam dies a pagan death."14 Nobody could achieve salvation without an imam (or at least agreement that there ought to be one), for there was no community without such a leader, or in other words there was no vehicle of salvation. Thus we are told that when the Prophet died, the Muslims hastened to elect a new imam because they did not want to spend a single day without being part of a community (jamd'd).15 For all that, both the Kharijites and the Shakes combined living without an imam with a strong sense of being the only saved, apparently by deeming agreement on the obligation to establish a true imamate, if and when it should prove possible, sufficient for the creation of a] community, a saving vehicle.16 The second task of the imam was to lead the way. I Ie did not simply cause the caravan to exist: he also guided it to its right destination. An imam was 'somebody to be imitated', as the dictionaries say: one went where he went and expected to prosper together with him in this world and the next. For a true imam was an imam al-budd, an imam of guidance who could be trusted to show his followers the right paths. He was compared to way-marks, lodestars, the sun, and the moon for his ability to show the direction in which one should travel.1' Without him one would not know where to go. He knew better than anyone else because he was the best person of his time: it was his superior merit that made people follow him. His guidance was seen as primarily legal, or in other words he declared what was right and wrong, for it was by living in accordance with God's law that people travelled to salvation. The coercion he might use to prevent people from straying from his caravan, or sowing dissension in it, was part of his guidance too, for anyone who strayed from the right path was lost and everyone would perish if the caravan broke up. His duty was to get people into the caravan and to make sure they stayed there without raising trouble or trying to go off on their own. This done, he could lead them along the right track until they reached their destination. Everyone who travelled with him would be saved, everyone else was lost. 14. E.g. Ibn Sa'd, v, 107 (R, 144) (cited by Ibn cUmar to make Ibn Muti* pay allegiance to Yazld I); Tbn Abi Shayba, xv, no. 18997, cf- ty°47; Ibn Hanbal, Musnad, iv, 96. 15. Tab., 1, 1824.16. 16. Cf. further below, 287. 17. Cf. Crone and Hinds, God's Caliph, 34ff, (Umayyad), 82 ('Abbasid; add Azdl, Mawsil 427, on al-Muctasim: aqama 'l-imamu manara 'l-huda). First Civil War and Sect Formation 23 A true imam was the opposite of an imam al-daldl, an imam of error, somebody who claimed to be a rightful leader, but who was actually illegitimate and wbo would thus take his caravan to Hell. If you paid allegiance to a false imam, you were doomed, for you would necessarily end up in the same pla the hereafter as the man whose caravan you had chosen to join. Anyone who joined the wrong caravan became an unbeliever (kafir), for there was only one community of believers. It travelled under the one and only imam of guidance representing the one and only God. All this was generally agreed. But who was the imam of guidance when TJthman was killed? How could one be sure that one was travelling to Paradise rather than to Hell? That was the problem raised by the first civil war. Fitna "The imam [TJthman] Ibn 'Affan has been killed and the Muslim cause has been lost. The paths of guidance have become dispersed," the poetess Layla Akhyaliyya complained: not knowing where to go when 'Uthman was killed, the believers had gone in different directions, inevitably meaning that some of them had gone astray.18 Where was one to go? "Which party shall I deem infidel and which believing?," as a young man asked himself with reference to the followers of 'All and Mu'awiya.19 If 'Uthman had remained an imam of guidance till the end, he had been unlawfully killed and his position was in abeyance until he had been avenged and a new caliph elected; 'All was in that case an imam of error leading a party of infidels who had condemned themselves to Hell by murdering a caliph and supporting a usurper.20 But if 'Uthman had fallen into error, he had forfeited the caliphate and been lawfully executed for his refusal to mend his ways or step down; 'All was in that case the true imam and it was his opponents who were "calling to hellfire".21 But maybe 'All had forfeited the caliphate as irrevocably as had 'Uthman, by-accepting the call to arbitration at SiffTn. This is what the Kharijites claimed: after all, it showed that he doubted his own entitlement to the office.22 In that case, both 'Uthman and 'All had turned into unbelievers. In short, to choose one's leader was to choose one's vehicle of salvation. It was because the civil 18. Zubayr b. Bakkar, Muwaffaqiyyat, 511, no. 326. 19. Ibn Sa'd, vii, 1, 82 (B, 1T4)- 20. Cf. MM, v, 16L, (iii, §1774), where Mu'awiya asks Hujr's followers to repent of their kufr. 21. Hence Hujr saw renunciation of allegiance to 'All as leading to dlikhul al-nar (MM, v, 17 (iii, $1774)). 22. Thus the Najdiyya (cf. Crone, 'Statement by the Najdiyya Kharijites', 63]. 2-4 First Civil War and Sect Formation war forced people to make up their minds about the whereabouts of this vehi-cle that it was known as fitna, test or trial: God was testing the believers to sei how many of them would come out of it with their sense of guidance intact, j The first civil war (656-61) was soon followed by a second, in which thd sons took over from their fathers. Mu'äwiya was succeeded by his son Yazldl in 680. The latter's caliphate was contested by a son of al-Zubayr's, 'Abdallah, and also by a son of 'All's, Husayn (for some reason no son of Talha's appeared).2' 'Abdalläh b. al-Zubayr, who ensconced himself in Mecca, refused to pay allegiance to Yazld I and, when YazTd died prematurely at the end of 683, claimed the caliphate for himself. 'Llthman's party thus split into two. Just as "Alawľ had come to mean an adherent of 'All - and later his kinsmen - to the exclusion of the Kharijites, so "Uthmänľ now came to mean an adherent of 'Uthmän and his Umayyad relatives to the exclusion of those who supported the Zubayrids.24'All's son, al-Husayn, set off for Kufa in 68o in the expectation of Kufan support against YazTd I, to be cut down by Umayyad troops at Kar-balä1. To Shľitcs, this became an event of almost the same importance (though not the same meaning) as the crucifixion of Jesus to Christians. When YazTd I died in 683, the Kufans repented of having left al-Husayn in the lurch. Some of them (known as the Tawwábun, 'Penitents') marched off on a suicidal mission against the Syrians. Others stayed in Kufa, where they accepted the leadership of a rebel by the name of al-Mukhtär, who proclaimed another son of 'All's by the name of Muhammad b. al-Hanafiyya to be the messiah (mabdi).25 Meanwhile the Kharijites were busy rebelling in Arabia and Iran under leaders of their own. Like the first civil war, the second was won by the Umayyads in Syria, now represented by 'Abd al-Malik (685-705), whose troops defeated and killed Ibn al-Zubayr in Mecca in 6 Ibn al-Zubayr declared in the course of negotiations with Kharijites,27 who had hoped for a different response. They told him that he hould "treat as an enemy him who was the first to institute error and make nnovations, and to depart from the judgement of the book (i.e. TJthman); if do that, you will satisfy your Lord and save yourself from painful chastisement"; he would forfeit his life in the next world if he did not, they said.28 "What do you say of Mus'ab (b. al-Zubayr)?," other Kharijites asked of Basran troops with reference to Ibn al-Zubayr's brother and governor of Iraq; "is he your friend (wait) in this world and the next . . . are you his friends (awliya1) in life and death . . .? What do you say about 'Abd al-Malik? . . . are you quit of him in this world and the next . . . are you his enemies in life and death?1'29 To the Kharijites, neithet the Zubayrids nor the Umayyads were rightly guided. In the same vein the famous exegete, Ibn al-'Abbas, reputedly declared the Zubayrids and the Umayyads alike to "call to hellfire".'0 One's imam in this world was still one's imam in the next. Choosing him was still to choose one's vehicle of salvation. Affiliation and dissociation The first civil wars thus split the community into rival communities (urn-man). Since each was seen by its members as a vehicle of salvation, each was a potential sect, and its members would certainly do their best to insulate themselves in social terms: having chosen their umma (or, as time passed, been born into it), they would publicly declare their loyalty to its imam and his adherents and dissociate from everyone else. Thus Ibn al-Zubayr would declare his loyalty for (wala, tawalla, also translated 'affiliate to', 'associate with') 'Uthman's party by declaring himself a friend (wall) of the latter, meaning that he accepted TJthman as a true imam, supported his cause and had a relationship of loyalty and support (walaya) with everyone else who did so. The Kharijites by contrast would dissociate from (tabarra'a min, also translated 'declare themselves quit of) 'Uthman's party. The people with whom one had walaya were those with whom one would socialize. One would eat with them, sit with them in the mosque, intermarry with them, allow them to inherit from other members of one's community, accept inheritances from them, help and assist them, pray for forgiveness for them, and hope to see them again in Paradise: they formed one's communitv in this world and the next. But those from whom one did bara'a 27. Ibid., ii, 517.2. 28. Ibid., ii, 515.10. 29. Ibid., ii, 821.13. 30. BA, v, r95f. z6 First Civil War and Seel Formation were tiot members of one's community, not believers, and not destined for Paradise, and one should not eat with them, sit with them in the mosqul intermarry with them, have relations of mutual inheritance with them, visit them, pray for forgiveness for them, greet them, or even smile at them.31 One might well curse them. Mu'awiya's governor of Kufa would curse 'All and all the murderers of 'Uthman, and pray for forgiveness for'Uthman and his party! (sbj'a);32 the adherents of 'All would curse 'Uthman or Mu'awiya, dissociate from them, and stay away from the Friday service and other public prayers led by Mu'awiya's governor." People suspected of Kharijism would be asked to curse the Kharijites: a young man in Basra refused to do so, just as he refused to declare himself 'aid din Mu'awiya, 'a follower of Mu'awiya's religion', or better, 'in a state of obedience to Mu'awiya'.34 When people declared themselves to be 'aid din fulan, 'followers of the religion of so-and-so', they did not mean that they had opted for a particular set of religious doctrines associated with that person, though this is precisely what later Muslims took them to mean. To them, the usage came across as offensive: there was no separate 'religion of 'AIT', all there was to 'All's religion was Muhammad's.35 But when first-century Muslims declared themselves to be 'aid din fulan, they simply meant that they accepted the person in question as the true imam and his party as the saving community. The declaration was synonymous with one of allegiance (ivalaya). "I, Ibn Shaddad, am in obedience to 'All {'aid din 'Alt) and not a friend iwali) of 'Uthman b. Arwa," as an adherent of al-Mukhtar in Kufa declared, derisively identifying 'Uthman with reference to his mother rather than his father.36 When a Syrian chief prepared for action to save the Umayyad house, he made a speech saying, "I bear witness that if obedience to Yazld b. Mu'awiya {din Yazid b. Mu'awiya) was right when he was alive, then it is still right today and his party are in the right too; and if Ibn al-Zubayr and his party stood for falsehood at that time, then he and his party still do so today."37 People would constantly declare their communal stance. Others would know where they stood and socialize accordingly. It was by mechanisms of this kind 31. For example NasafT, Radd, j8f. Ihn Bäbawayh, 'Vyün, ii, 2.01.7; Nagel, Recklleitung und Kalifat, 268, citing TB, viii, 144. 32. Tab., ii, in. 33. Ibid., ii, T15,147, 234. 34. Abl '1-Aswad al-Du'all, Diwan, jzff., no. 47. 35. Thus 'All himself in Madelung, Succession, 178, citing Ihn Durayd. Of the young man in the previous note we are similarly told that he would only declare himself to be 'aid din Ibrahim. 36. BA, v, 233.2. 37. Tab., ii, 469^3. First Civil War and Sect Formation [ 27 that sects in the sense of bounded communities identified by different views on religious questions began to emerge. Needless to say, there must have been a good deal more to it than this schematized account suggests. The initial outcome of the debate over the caliphs seems to have been a confusing welter of small groups with slightly different views; amalgamation probably played as great a role in sect formation as splits. But whatever the details (which may prove beyond reconstruction), the reason why seemingly political disagreement led to the formation of sects is that the political leaders were religious guides and that their followers could not disagree about their identity without thereby assigning themselves to different vehicles of salvation. Fast caliphs All Muslims needed to be sure that the community they travelled in was that founded by the Prophet himself: they would not otherwise be Muslims. They thus needed to know the genealogy of their community, in the sense of the legitimate succession of rightly guided leaders from the Prophet to their own time. In tracing this genealogy all focused their attention on the point at which the disagreement about the succession set in, or, as one might also put it, the point at which the genealogy branched out. (For what follows, see charts 1 and z.) The Prophet was succeeded by Abu Bakr, who was succeeded by 'Umar, who was succeeded by "Uthman: there does not initially seem to have been any disagreement about the succession up to here; it was the civil war that started the disputes. The earliest sources to survive present the civil war and aftermath as a conflict between shi'at 'Uthman and shi'at 'Ali, no more and no less: the "Uthmams said that 'Uthman had remained a lawful caliph till the end, the Kharijites and ShTites said that he had not; the issue was whether one should attach oneself to the Prophet via 'Uthman or 'All; that the first two caliphs had been legitimate imams was taken for granted.38 But the possible lines of descent soon multiplied. The Kharijites attached their community to the Prophet via Abu Bakr and 'Umar without recognizing either 'Uthman or 'All as legitimate imams: though both had been rightly guided at the time of their accession, both had forfeited their imamate and indeed their status as believers by their sins. The Murji'ites ('suspenders of judgement"), who appeared c. 700 ad, also accepted Abu Bakr and TJmar as legitimate caliphs, but without pronouncing on the status of the two disputed caliphs who followed: it was no 38. Noted by Ibn Taymiyya, Minhäj al-sunna, i, 7. See further below, 7111. First Civil War and Sect Formation First Civil War and Sect Formation longer possible to establish whether TJthman or cAli had been right, they said; one should not affiliate to or dissociate from either of them. About the same] time some Shakes began to argue that the Prophet himself had designated 'Alt as his successor: they thus traced their community to the Prophet directly via! cAlT without going through Abu Bakr and TJmar, let alone "Uthman, rejecting all three as usurpers. In the early cAbbasid period the so-called Rawandiyya (named after a member of the 'Abbasid army) argued that the Prophet had designated his uncle al-'Abbas, from whom the 'Abbasid caliphs descended, as his successor, and cast his descendants as what one might call 'imams in exile' down to their open assumption of power. In principle, they thus rejected all the caliphs from the Prophet's death in 632 down to the "Abbasid revolution in 750I as usurpers (though they were unwilling to say so in practice).35 Not long after the Rawandiyya's appearance there were people who proposed a compromise between the TJthmanl and the 'AlawT positions: the rightly guided imams after the Prophet had been Abu Bakr, TJmar, 'Uthman, and CA1T. This was the famous four-caliphs thesis which spread at great speed during the ninth century to become the hallmark of all those Muslims who 'stuck to communal unity' (lazima al-jamd'a). JamacI Muslims The Muslims who stuck to communal unity will be referred to in what follows as jamd'l Muslims. The reference is to all those who refused to form separatist communities under present or future imams of their own even though they might regard the ruling dynasty as sinful - in effect all those who were not] Sh?ites or Kharijites. By the eleventh century one would call them Sunnis. One cannot use that adjective, however, until they had coalesced as a single party endowed with a shared understanding of the Sunna of the Prophet and its implications. In the early centuries they were divided into hostile groups that had little in common apart from their high appreciation of communal togetherness, and for this reason there is no single word for them. They did form a single party for some fifty years, from the first civil war to c. 700, and in that period one can call them TJthmanls. But then the Murji'ites appeared, followed by Partisans of Hadith and Mu'tazilites, while 'TJthmanf came to mean different things. All these people remained something of a single party to their ShTite and Kharijite opponents thereafter too, but this did not give them a name that we can use. The Shrltes and the Kharijites often called those who were prepared to live under sinful rulers 'MurjPites', at least from the ninth 39. Cf. below, 92t. century onwards. But this is confusing in that the Murji'ites were strictly speaking only one party committed to jamd'a and that the Traditionalists, the other major party, resented being btacketcd with them. Moreover, the Mu'tazilites were probably not included in the appellation; and even if they were, they shared with the Traditionalists the feature of not wanting to be: they called all their non-Shrite and non-Kharijite opponents 'Murjfites.'40 External observers could see that the various parties had something in common, but it was usually in a polemical vein that they grouped them together, and they always perceived their own parties as special. For this reason no term coined at the time is likely to help us. The use of the term jamcfi was pioneered by Hodgson half a century ago, when Islamicists routinely spoke of Sunnis from the first civil war onwards, or even from the time of the Prophet. As the first to change established usage, he felt obliged to retain a reference to the Sunnis in his new expression, so he opted for jama'i-sunm, regretfully noting that jamd'i on its own would have been a better term.'11 Now that it is generally accepted that the formation of Sunnism was a protracted process, we can adopt the more accurate (and less cumbersome) form. Jamd'i Muslims in this book are much the same people as the 'Murji'ites' of the Shrites and the Kharijites, but they include Mufazilitc adherents of the four-caliphs thesis, and from the eleventh century onwards I shall replace the expression with 'Sunnis', except when I wish to include the Mu'tazilites. When I have to use English terms, I shall translate jama'i as 'communitarian' or 'communalistic', refusing to sanction either by exclusive use since neither is very apt. The reader is asked to remember that in the context of this book a communitarian is the opposite of sectarian, not of a believer in individual rights; the term stands for somebody who wanted to keep the caravan together, not for somebody who wanted to shore up the family. Similarly, communahsm in this book has nothing to do with shared property, local autonomy, or devotion to ethnic and cultural subgroups at the expense of the community as a whole; on the contrary, devotion to the community as a whole was precisely what commitment to the jamd'a was about. Present Caliphs Initially, contemporary caliphs played the same role in the self-definition of the rival parties as those of the past. It was after all as a dispute about 40. Cf. Crone and Zimmcrmann, Epistle, 243. 41. Hodgson, Venture of Islam, i, 278. The better term is not in fact entirely without support in the sources; cf. YB, 352.15: its inhabitants are not Kharijites ishurat) but rather jama'tyya. 30 First Civil War and Sect Formation contemporary claimants to the imamate that the first civil war had begun. Questions such as "what do you say about Mu^wiya?" or "what do you say about cAbd al-Malik?" were meant to elucidate whether contemporaries regarded these caliphs as imams of guidance, and at first sighr things continued unchanged after the second civil war. The notorious governor, al-Hajjaj, professed to believe not only in the unity of God and the messengership of Muhammad, but also in obedience to the caliph al-Walld I (705-15): "on this he would live, on this he would die, and on this he would be resurrected."42' Adherents of the Umayyads continued to accept the caliphs of their own time as imams down to the end of the dynasty, and indeed beyond.4j But for all that, it is clear that Muslim society was rapidly becoming too complex for the old fusion of the political and religious communities to be viable. In 744 civil war broke out again, provoked by the murder of the caliph al-Walid II in 744 and continuing as the cAbbasid revolution that toppled the Umayyads in 750. The third civil war was yet another fitna, or test, of the believers, but a good number of Muslims now held that the best reaction was to lock one's door and stay at home until the candidates for power had fought it out: to choose one's caliph was no longer to choose one's vehicle of salvation; the caliph was no longer an imam in the original sense of the word. He thus ceased to matter for purposes of the genealogy of the community: he was a mere guardian of the community now, not a link in the history of the transmission of right guidance from the Prophet down to today: Most Muslims ceased to regard the Umayyads as such a link in the course of the Umayyad period and never really accepted the 'Abbasids in that role either; as they saw it, the transmission of right guidance had passed to the religious scholars. Only the Shi%es continued to think that the ruler ought to be such a link, and would be if only he was chosen from among the rightful candidates. To the majority of Muslims, the debates about the genealogy of the community stopped involving the caliphs after 661, when the Umayyads took over. The de-palilicization of the community of believers That the caliph had ceased to define the community is neatly illustrated by the contrasting reactions to the Arabian and the Andalusian withdrawals of allegiance to him, in 632 and 756 respectively: When Muhammad died in 632, many Atab tribes stopped sending alms taxes to Medina. They did not all repudiate Islam, but they were branded as apostates (murtadds) regardless, for they 42. Crone and Hinds, God's Caliph, 41. 43. Cf. Nasr h. Sayyar's poem in DmawarT, 359; Tab., i, 2815.13; Azdl, 252.10; El1, s.v. 'cUthmaniyya'. First Civil War and Sect Formation 3' had ceased to be members of the Muslim community by their refusal to recognize Muhammad's successor in Medina. A little over a century later, in 756, the ruler of al-Andalus, cAbd al-Rahman I, stopped acknowledging the 'Abbasid caliph as his overlord by omitting mention of him in the Friday oration. In principle the Andalusians had thus also ceased to be members of Muhammad's community, and it would have been easy enough for the 'Abbasids to brand them as apostates. But it does not seem to have occurred to them, or for that matter anyone else, to do so. One could now be a member of Muhammad's community without paying allegiance to Muhammad's successor. This point of view is also attested in the account, by the Kufan historian Sayf b. "Ulnar (d. between 78(1 and 809), of the Arab conquest of Iran. On the eve of the battle of Qadisiyya (between 635 and 637), according to Sayf, the Arabs assured the Persians that they had no wish to conquer their land, only to bring them to the truth; if the Persians would accept Islam voluntarily, the Arabs would go home, leaving them to tule themselves in alliance with Medina. "Choose Islam and we'll leave you alone on your land," they said; "we'll be loosely associated with one another, but your land will be yours and your affairs will be in your hands"; "we'll leave you with the Book of God and put you in charge of it on the understanding that you'll govern according to its laws. We'll leave your country and let you deal with its affairs as you please."44 This is apologetic, of course: it is directed against charges by the conquered peoples that the Arab conquests were motivated by worldly ambitions rather than a desire to spread the truth.4' But its interest here lies in the fact that by the later eighth century it was possible to imagine a world in which there were two sovereign Muslim states within the community of believers. If the Persians had not been so stubborn, according to Sayf, an Arab and a Persian polity would have coexisted in amity, ruled by a successor of Muhammad and a Khusraw respectively, travelling along the same paths of truth without amalgamating in political terms. Others, too, were entertaining the idea of a Muslim world divided into a plurality of allied or federated states about this time.'"' There was only one community of believers, as all agreed even though they disagreed about its identity; but as most Muslims now saw it, it was defined by-its beliefs and its religious leaders, not by its political leadership. By fixing the number of rightly guided caliphs at four, mainstream Muslims laid down that religious guidance could never be concentrated in the head of state again (cf. chs. 11, 16). As noted, the ShI'ites disagreed. But one section of them, the so-called Imamis (or Ithna-'asharls, 'Twelvers') declared 44. Tab., i, 2240, 2272, 2280, ef. 2271, 2273, 228 45. Cf. below, 374. 46'. Cf. al-Asamm below, 68. 31 . First Civil War and Sect Formation their imam to have gone into hiding in 874, thereby in effect laying down that the religious guide could never become a head of state again (cf. further below, ch. 10). Most Kharijites and all ShTites of the type known as ZaydTs continued to consider it possible to unite the two roles in a single man, but they never succeeded in taking control of the Muslim world at large and such imamates as they succeeded in establishing on the fringes were intermittent. Within two centuries of the conquests, the vast majority of Muslims thus found themselves ruled by caliphs whom they did not consider to be embodiments of right and wrong. They might grudgingly recognize them as legitimate, accepting them as what one might call quasi-caliphs - caliphs without the epithet of Rashidun (rightly guided)4' - or they might positively denounce them as kings and tyrants, but either way they could not model themselves on them. For guidance to salvation they had to look elsewhere. Their religious fate thus ceased to be bound up with that of the ruler. Even quasi-caliphs could be righteous; the same was eventually found to be true even of kings. But one could not assure oneself of salvation by paying allegiance to them. In short, the community ceased to be in the nature of a caravan which travelled along a single path under the leadership of a single religio-political leader to a single destination. One still had to choose one's community, but one no longer did so by paying allegiance to a caliph. The end of simplicity By c. 800 the community of believers came in several rival versions and had at least two different types of leadership within each version, one political and the others religious. The proliferation of the types of leadership was not to stop there, for the caliphs soon came to coexist with kings and sultans, while the religious scholars soon came to coexist with philosophers and Sufis (mystics) too. The simplicity of Muhammad's multi-purpose polity had gone. It is sometimes said that the schisms over the imamate arose because the Prophet had made no provisions for the succession, but this is difficult to accept. For one thing, no serious disagreement arose over the succession until 'Uthman was killed. For another thing, the Muslims were unlikely to retain their simple socio-political order and broad consensus on political and religious matters when the bulk of them left Arabia for the fleshpots of Egypt, Syria, Iraq, and Iran. The very fact that the imam represented their worldly and otherworldly interests alike meant that all clashes of interests were likely to display themselves in disputes about his identity and the nature of his office. It is to these disputes that we may now turn. 47. Cf. below, 139. CHAPTER 3 THE UMAYYADS As the Roman expansion had undermined the Roman tepublic, so the Muslim conquest of the Middle East destroyed the patriarchal regime in Medina. In both cases, civil war was followed by the emergence of an increasingly authoritarian monarchy. The Muslim counterpart to Augustus was Mu'awiya (661-80), who moved the capital to Syria and founded the Umayyad dynasty (661-750), undet whom the embryonic state founded by the Prophet acquired a more developed form. But the developments unleashed by the conquests continued to transform Muslim society, rapidly making the political organization of the Umayyads obsolete, their orientation outmoded, and the dynasty itself heartily disliked. Within three generations they had come to be denounced as impious survivors from the pagan past who had somehow managed to hijack the Islamic enterprise. They were ousted in the third civil war, more precisely that part of it known as the 'Abbasid revolution. But contrary to what many had hoped, the trend towards more authoritatian government was not reversed. A fully-fledged, if shortlived, empire emerged under the 'Abbasids (effectively 750-861; faineance 861-12.58). All the fundamental questions first raised under the Umayyads continued to be debated down to the effective end of the 'Abbasid empire some hundred years after the revolution. Legitimacy By what right did the Umayyads tule? An extraordinary amount of medieval Islamic political thought is devoted to legitimation of the dynasty in power, and this was so already in Umayyad times. [33] 34 The Umayyads I he Umayyads grounded their right to the caliphate in the legitimacy o| 'Uthman. He had been lawfully elected by consultation (shura) and unlawful), killed, indeed martyred; his Umayyad kinsmen and avengers had taken over hi's position as imam of guidance.1 'All had never been a caliph, merely a rebel or imam of error whose family the idolatrous Iraqis continued to worship as the Israelites had worshiped the golden calf.2 In short, the Umayyads we«j Tthmanls who cast their accession as a Restoration: the royalists had won, tU legitimate monarchy was back, the rebel interlude was over. In addition, every Umayyad ruler was a man of unsurpassed merit (a\ afdal), indeed the best man alive (kkayr al-nas):3 so numerous poets from the time of cAbd al-Malik onwards assure us in the conviction that monarchy was government by the most virtuous individual. This was an idea with long roots. "Nothing can be found better than the rule of the one best man," as Herodotus had made the Persian emperor Darius (d. 486 bc) declare. A man far superior to everyone else would have to be obeyed as king, or alternatively ostracized, as Aristotle (d. }zz rc) said. "Among human beings the highest merit means the highest position," as the Stoic philosopher Seneca (d. 65) succinctly put it, voicing what had by then become a Hellenistic commonplace. It was still current in late antiquity, and not just on the Greek side of the border.4 When the philosopher Paul the Persian dedicated a treatise on logic to the Persian emperor Khusraw I (531-79), he addressed him as the best of men (tava de-gavre).3 But what if a man initially rejected as inferior proved capable of mak-| ing everyone obey him? One response would be to accept that he had thi requisite merit after all and hence that his power was right. It appears to have been by some such reasoning from might to right that the Arabs came to terms! with Mu'awiya. It was certainly how the court poets tended to think: the sheer The Umayyads [35 e i s rim the Umayyads had power showed that they deserved it; since God had Jsputably raised them up, they had to have qualities which pleased Him.6 The fact that the Umayyads saw their power as decreed by God has often been taken by modern Islamicists to mean that they supported determinism 1st the doctrine of free will (known as Qadarism) in an effort to deprive their subjects of a right to resist: God having decreed that they should rule, there was nothing the believers could do about it.7 But although it is undoubtedly true that the Umayyads were determimsts, it has to be remembered that so were most of their enemies, bc they Kharijites, Murjfites, or ShTites. Since the political debate took place within a generally determinist ambience, the Umayyads had no reason to give much thought to the relationship between divine omnipotence and human action for purposes of justifying their regime, nor do they seem to have done so. They simply took their success to mean that God was on their side.8 "So long as He will have them to reign . . . kings are armed with authority from God," as Calvin said, meaning that a particular ruler was to bc obeyed so long as he had the power to impose obedience.9 "God helps the strongest," as a popular maxim quoted by Luther more succinctly puts it.10 This is precisely what the Umayyads meant, and how early Muslims in general seem to have looked at things. The obvious determinist response was to deprive the Umayyads of their divine authority by defeating them, and a fair number of determinist rebels tried to do just that. God eventually helped the strongest by toppling the Umayyads at the hands of Khurasan! revolutionaries whose views, if any, on determinism are unknown. But the fact that the Umayyads had expressed themselves in a determinist vein came in handy for the advocates of free will, who used it to discredit the determinist position, perhaps already in the Umayyad period and certainly in later times. This is what accounts for the modern impression that the Umayyads supported determinism in opposition to the doctrine of free will for political reasons.11 1. Cf. Tab., 1, 3243.5-12; Crone and Hinds, God's Caliph, }iL; Nadler, Umayyaden-kalifen, 65ft. 2. Crone, Slaves, 204, note 30; Tab. ii, 1774; Cook, 'Apocalyptic Chronicle', 27. Cf. also the attitude to 'All in early 'Abbasid Syria (van Ess, TG, 1, 7of.). 3. Crone and Hinds, God's Caliph, 30, notes 23, 24; Nadler, Umayyadenkalifen, 77, T45, 170, 175. 4. Dvornik, Political Philosophy, i, 175 (citing Herodotus, iii, St), and index, s.v. 'best man', theory of; Aristotle, Politics, i284a-b, cf. 1288a; Seneca, Letters, no. 90. Cf. also Themistius, Kisala, 34, J6/961,100, translated (if his it is) long after the idea had surfaced on its own in Islam. 5. Land, Anecdota Syriaca, iv, 1 = 1. 6. Compare the story in which Sulayman takes the very misery of lepers to prove that they deserve it: God would not otherwise have inflicted such a fate on them (Conrad, 'First Islamie Hospital', 234t.). 7. Thus Goldziher, Theology and Law, 8j£; Wart, Political Thought, 38; 'God's Caliph', 570 (and elsewhere); van Ess, Zwischen Hadit und Theologie, i8iff. (and elsewhere); but cf. now TG, i, 24. Further references in Murad, 'Jabr and Qadar', ttj. 8. Cf. Crone and Hinds, God's Caliph, ii-i.; Crone and Zimmermann, Epistle, eh. 5; Nadler, Umayyadenkalifen, 194, 275; Murad, 'Jabr and Qadar . 9. Walzer, Revolution of the Saints, 38. 10. Luther, 'Secular Authority', 40. 11. Cf. al-Jubbä'i in van Ess, Anfänge, 241h; Zimmermann, review of Anfänge, 440L 56' The Umayyads Dynastic succession Shortly before his death Mu'awiya designated his son YazTd as his successor! and made the leading men of the garrison cities pay allegiance to him. Whed YazTd I died prematurely in 683, the Syrian chiefs chose Marwan I, from a dif.' ferent branch of the Umayyad house, as his successor; and Marwan I in his turn designated two sons as his successors, one as heir and one to spare. Though at times the presence of two heirs proved almost as problematic as none, this became standard practice for the rest of the Umayyad period. The Umayyads themselves took pride in their dynastic succession. As their court poets present it, 'Uthman's election by shurd showed that the founder of the lii eage was legitimate while the hereditary nature of the caliphate thereafter showed that the Umayyads were highborn rather than upstart rulers.12 But to the enemies of the dynasty, it was above all by their inrroduction of dynastic succession that the Umayyads condemned themselves. How much resentment Mu'awiya's designation of Yazld actually aroused at the time is difficult to say: the main opposition seems to have come from Qurashls in Medina (such as Ibn al-Zubayr and al-Husayn) who found themselves excluded from consideration. But whatever the initial reaction, there is no doubt that the resentment grew thereafter. Automatic succession within a single family was quite rightly seen as a first step towards the imperial form of] government which was to culminate under the cAbbasids. The sources for the Umayyad period abound in calls for the election of the caliph, whether fromi among all Muslims, all Arabs, all QurashTs, or all members of the Prophet's family, by consultation (shurd). Indeed, when YazTd III rebelled against al-WalTd II to ascend the throne himself in 744, he too called for shurd and volunteered to step down if a more meritorious candidate could be found. For all that, there are two puzzling problems here. First, the Umayyad mode of succession continued under the 'Abbasids, who sometimes designated one heir and sometimes two. The chroniclers accept this without demur, but nonetheless continue to blame the Umayyads in strident tones for their introduction of dynastic succession. Secondly, the Umayyad mode of succession was adopted by the Imami ShT'ites too, without the heir to spare; but m polemics against the ShT'ites, the Sunnis will insist, not that the ShTcitcs are copying their mode of succession, but on the contrary that their own mode of succession is elective. The secondary literature usually copes with these problems by postulating that succession in early Islam was elective and that this is what Sunni law iz. Crone and Hinds, Cod's Caliph, 51. Cf. also the Umayyad claims in Kumayt, I Hasimijjat, no. 2:37. r preserve ;oJ underB The Umayyads 37 /es; hereditary succession was allegedly introduced by the Umayyads Bvzantine or Persian influence,13 or alternatively by way of reversion to Arabian ideas,14 in any case by way of departure from Islamic norms; the Sunnis learnt to live with it, but they always maintained a principled stance against it, which they displayed by insisting on the elective nature of their own svstem and condemning the Umayyads for changing it. But this is not quite on target. It is certainly true that succession to high office in pre-Islamic Arabia was normally hereditary. Tribal chiefs, judges, and guardians of sanctuaries in pre-Islamic Arabia came from the same families for generations on end; the sources envisage Muhammad's own tribe of Quraysh as hereditary guardians of the Ka'ba. Chiefs were not necessarily followed by their sons, however, still less was there a rule of primogeniture; rather, the most suitable candidate from the chiefly family would be 'elected', in the sense that the person capable of asserting his authority would be accepted. (God helped the strongest in pre-Islamic times as well.) There were not usually any formal procedures, though thete cannot be much doubt that the pre-Islamic Arabs were familiar with shurd, the practice of letting rival candidates come to an agreement among themselves on the basis of estimates (acquired by consultation) of their relative support." Another component of the explanation is undoubtedly off target, however: we do not have to await the Umayyads for the pre-Islamic mode of succession to reappear in Islam. In a sense, the very first Muslims could be said to have used it. Abu Bakr, cUmar and all six members of the shurd, including TJthman, were Qurashi emigrants to Medina. So too were all the contenders for the caliphate in the first civil war unleashed by 'Uthman's death, including Mu'awiya. Some Syrians are said to have liked the idea of 'Abd al-Rahman b. Khalid b. al-WalTd as Mu'awiya's successor: he too was a QurashT, and the son of an emigrant to Medina.16 So too were all al-Husayn, Ibn al-Zubayr, and the men considered eligible for the caliphate in Syria at the time of Mu'awiya's designation of YazTd.17 With the exception of the Kharijites, all northern Arabs continued to expect the caliph to be chosen from Quraysh thereafter,18 and 13. Tyan, Califat, 243-8. 14. Watt, Political Thought, 39. 15. EI1, s.v. Crone, 'Sbürä as an Elective Institution'. lö.Aghäni, xvi, 197. 17. Ibid, xx, 212. A Syrian chief who said that they could easily find a replacement for Mu'awiya specified that they could do so ft qawmihi, 'in his tribe' (Tab., ii, 144). 18. A north Arabian participant in the revolt of Ibn al-Ash'ath is admittedly said to have hoped that the latter would become caliph, but the reader is expected to find it absurd (Tab., ii, mi). A late Christian source alleges that non-Qurashl Yazld b. al-Muhallab 3« The Umayyads some had apparently formulated it as a doctrinal requirement by the mi, Umayyad period.'9 In short, it would appear that the Qurashl emigrants Medina were regarded as the equivalent of a chiefly family. They were the on! kinsmen that the Prophet had left after cutting his ties with Mecca (whd many genealogically closer relatives of his remained), and they were the rnf with whom he had come to Medina. Within this group the caliph would [ chosen by one traditional procedure or other: Abu Bakr was 'elected'; cUmar said to have been nominated by Abu Bakr, which may have been an innov; tion, but he may also have succeeded simply because he was the obvious sue ,0 and the shura which he in his turn appointed was another pre-Isl amic cessor; procedure Still, succession within a lineage is not quite what is normally understood! by hereditary succession, and the lineage broadened to become an entire tribcB as all Quraysh converted. But succession from father to son (or other clo: kinsman) appeared already in the first civil war. This war pitted Qurashl lead ers against each other. Obviously, these men were not going to regard the. rivals as candidates for the succession; on the contrary, both they and their fol lowers would have a strong interest in keeping the office within their ow house. Accordingly, when 'All was killed, his followers elevared his son al Hasan to the caliphate. Towards the end of his life Mu'awiya similarly designated his son as his successor. All this was repeated in the second civil war, when sons reappear in their father's shoes on the Umayyad, the 'Alid, and the Zubayrid side alike. And when YazTd died a mere three years after his accession, the Syrian chiefs behind Mu'awiya's regime reacted, like the Iraqi chiefs behind 'All's, by elevating a kinsman of their deceased caliph to the throne, as has been seen. Three points follow from this. First, in a sense the imamate could be said to have been conceived as a hereditary institution from the start, but heredity did not determine the heir and thus did not exclude election. The Arabs combined heredity and election in much the same manner as the Franks of Merovingian Gaul, with whom they were contemporary. Secondly, the Umayyads narrowed down the group within which it was hereditary to their claimed the caliphate when he rebelled, but the claim is presumably wrong (Marl b. Sulaymän, 65.16). 19. When Jazlran Khärijites considered an alliance with Mutarrif b. al-MughTra in 77/6961, they were up against the fact that the latter restricted the caliphate to Qurashls: thus Abu Mikhnaf (d. 157/774) in Tab., ii, 984ft., and he may well be right, for the Shfite poet al-Kumayt fd. 126/743) also knew that the caliphate was restricted to Quraysh (he took it to mean that the Prophet's family must have the best right to it; Häsimijjät, no. 2:55). zo. CI. Rotter, Bürgerkrieg, 7. The. Umayyads 39 ■n family and replaced election within the group by designation, to be followed in this by the 'Abbasids and some of the 'Alids. Thirdly, the Umayyads were not the first to attempt succession from father to son, as the ill-starred reign of al-Hasan shows; they were only the first to do so successfully: The reason why the sources continue to lambast them for their introduction of dynastic succession long after this institution had come to be taken for granted in its '■Vbbasid form, or accepted as a positive article of faith by the Shl'ites, is not that people disliked the institution but rathet that they disliked the Umayyads: the one point on which all enemies of the Umayyads could agree was that the imamate should never have been hereditary in their family. So much for the first problem. As regards the second, the reason why the Sunnis describe themselves as adherents of election (ikhtiyar), as distinct from those who made the office hereditary (mawrutha), is not that they were clinging to Islamic norms which had come to be violated in practice, but rather that they were trying to distinguish theit own mode of succession from that of the Shl'ites .21 In tenth-century Sunni theory succession rested on a combination of heredity (within Quraysh) and either election or designarion. The authors of this theory accepted dynastic succession within the 'Abbasid family, even from father to son, without the slightest reluctance, whatever they might say about the Umayyads. They could not declare the imamate to be hereditary, however, without thereby appearing to agree with the Shl'ites that the imamate was a personal quality transmitted by descent rather than an office bestowed by the community.22 Moreover, if rhe succession was hereditary in the narrow sense of passing from father to son, or other close kinsman, as a matter of principle, not just in mundane practice, then the caliphate had passed from the Prophet to his cousin, 'A17, and his descendants, even though others had taken it in actual fact, or alternatively it had passed to his uncle, al-'Abbas, and his descendants: in the former case all actual caliphs orhcr than 'All had been usurpers, in the latter case 'All had been a usurper too, though all caliphs had been legitimate from the'Abbasid revolution onwards. Since the Sunnis did not agree with either view they said that the imamate was elective within Quraysh. What they meant was, firsr, that Abu Bakr, 'Umar, and 'Uthman had been legitimate caliphs even though they wete not members of the Prophet's family (i.e. the Hashimites), and secondly, that the dynastic nature of the succession undet the 'Abbasids was not of the slightest religious significance to them. 21. Similarly Mikhail, Mawardi and After, 2T. 22. Cf. below, ch. r6, zi6i. 4° The Umayyads The caliphal role The Umayyads 41 The Umayyads conceived their role as one of multi-purpose imams in the old style. As they saw it, the community of believers owed its legal existence to the imam, who guided it in what to us are both political and religious terms: he: defended it against infidel enemies, sought to expand its domain, maintained internal order and formulated God's law as the ultimate authority on the right, eous behaviour that the believers should seek to adopt. In the past, God had used prophets to convey His message; now He was using another type of agent to administer it. The caliph was God's deputy or vicar on earth (khalifat alldh ft ardihi) with all the fullness of power that this implied.23 Like the assumption that the ruler had to be the best of men, this idea seems to be rooted in Hellenistic conceptions. The king was living law (nomas empsychos), as people said from Hellenistic to Byzantine times; he ought to be sinless and perfect in argument and deed, a certain Musonius (d.-c. 120) told a Syrian king. It was his duty to imitate God, so that his subjects could imitate him, as everyone agreed. "There is no need for any compulsion or menace ... when they see with their own eyes a conspicuous and shining example of virtue in the life of their ruler," the moral philosopher Plutarch (d. after 120) explained.24 The king had been fashioned on the model of his maker; he was the image of God, the archetype of the true king, as the pagans said and the Christians started agreeing when Constantine converted.25 He was living law in the sense of embodying and exemplifying the law, not in the sense of overruling it. On the contrary, he had to be ruled by God's law in order to rule' legitimately himself.26 All this was close to the Muslim conception, except that they did not usually believe in human images of God. It was as God's deputy (khalifa), not as his image or 'after His likeness', that Adam had been created according to the Qur'an;27 and it was as God's deputy rather than His image that the caliph was expected to provide a model for imitation by his subjects. But like their Hellenistic predecessors, they saw their imam as living law and held him to display his legal insight, among other things, by adjudicating (cf. Q. 38:25). Like them, too, they saw the ruler's position as depending on his relationship with God, 23. Crone and Hinds, God's Caliph, ch. 3. 24. Dvornik, Political Philosophy, ii, 548, 552!.; Hunger, Prooimon, 58ft". 25. Ecphant1.1s in Dvornik, Political Philosophy, i, 253; Beskow, Rex Gloriae, 263 (Eusebius); cf. also Baynes, 'Eusebius and the Christian Empire'. 2.6. Thcmistius in Dagron, 'Empire romain', rijff.; Aalders, 'Nomos Empsychos', 322! and note 28; Agapetos in Barker, Social and Political Thought, 59 (cap. 37); Henry, 'A Mirror for Justinian', jot; Dagron, Empereur el pretre, 37. 2.7. Q. 2:20, paraphrasing Gen. 1:26. at ;s his moral rectitude: leadership was a reflection of personal virtue rather him a'n office. "It is not wealth . . . that makes a king, nor his purple cloak, nor nara and scepter . . . nor numerous hoplites and myriad cavalry; not though 11 men should gather together and acknowledge him for their king, because virtue they cannot bestow on him, but only power," as Julian the Apostate (d. . ^ said. If a virtuous man seized power (monarchid), his virtue would at once make it lawful sovereignty (hasileia), as the eclectic philosopher Plutarch had it23 It was thanks to this line of thought that the debate about TJthman [,ad centered on his moral status rather than the relative rights of rulers and subjects. It is thanks to the same view of things that people would deny that the Umayyads were caliphs when they held them to be wrong, or that conversely the Shicites came to view their imam as the ruler even though he did not have any political power, or again (after the further injection of such ideas by means of translations) that the philosophers would insist that the true king was not "someone possessing a cavalcade, a retinue or a realm: what is meant, rather, is one truly deserving of kingship, even if nobody pays him any attention."29 Of course, the ruler's moral status displayed itself above all in his treatment of his subjects, so people did indirectly discuss such rights as well. In fact, the relative rights of ruler and subjects figure prominently in the political debate of the Umayyad period, but only in the sense that whenever moral evaluations are spelt out in concrete terms, it turns out to be such rights that are involved. It should be noted that court poetry also describes the Umayyad caliphs in terms that can only be characterized as messianic: they dispersed darkness, made the blind sec, made the road plain, loosened burdens, healed sicknesses of the breast, slaked people's thirst, filled the earth with mercy, justice, and light; each one was a mahdi, a rightly guided deliverer from evil; some were even hailed as the Mahdi, the Redeemer.30 There was a Hellenistic tradition for casting the ruler as a saviour (soter), but it was not meant in a transcendental sense and is unlikely to account for the Muslim conception. The significance of the messianic overtones will be taken up in Chapter 7. Exalted though his position was, the caliph was not credited with any personal sanctity, nor was his court dominated by elaborate ceremonial designed to assure people that they formed part of an eternal and 'profoundly correct' polity maintained directly by God.31 He did fulfil a major cultic function as the leader, in person or through his governors, of the weekly Friday prayer, and 28. Dvornik, Political Philosophy, 11, 547, 662. (with oddly off-rarget comments). 29. NasTr al-DTn Tusi, Ethics, 192; cf. further below, ch. 14. 30. Crone and Hinds, God's Caliph, }6f. 31. Cf. Bowersock et al. (eds), Late Antiquity, s.v. 'emperor'. 42. ] The Umayyads also of the annual pilgrimage (usually delegated to members of his family). But he owed this function to his position as head of the community, which was seen as coming together in its entirety on these occasions and whose togetherness he symbolized, not to any personal magic or charisma [baraka). He had no physical inviolability, no special powers or healing touch, and no special ritual status either; on the contrary, he was subject to the same rules of purification, worship, and marriage as everyone else. He did play a prominent role in prayers for rain,32 and the later Umayyad caliphs were the object of more extravagant claims than the first, in terms of rheir persons and their religious insight alike;33 but by and large the caliph remained an ordinary human being in the Umayyad period. The redemptive overtones of his guidance notwithstanding, his role was conceived in a sober vein: as God's caravaneer, he kept the believers together in safety and order while leading them along the right path to their ultimate destination, no more and no less. The Umayyads con- I stantly stressed the importance of sticking to the jama'a, the collective body or compact majority. "Satan is with the individual," as al-YX'alTd I reputedly said;3'1 scattering, dispersing, going separate ways and following individual whims (fot which there was a rich vocabulary) meant going astray and perishing. One stuck to the jama'a, and thus to the safe path, by obeying the Umayyad imams of guidance, who were way-marks and lodestars to their followers and who never tire of enjoining obedience (ta'a) in their official letters. The rise of the scholars The manner in which the Umayyads fulfilled their duties as religious guides strikes the reader from a Christian background as odd. Contrary to what one would expect, they did not create a hierarchy of officials charged with the definition, maintenance, and dissemination of right belief or practice. The same was to be true of the ShT'ite imams, though the latter's formal role as religious guides was more pronounced than that of the Umayyads. But then the Christian church developed as a separate organization within the Roman empire; it 31. Ringgrcn, 'Some Religious Aspects of the Caliphate', 740 (where one line seems to suggest that the caliph was a factor behind fertility in general); cf. EI-, s.v. 'Istiska''. 33. Cf. K. al-Vyun, iii, 101 (where Hisham is told that caliphs were immune to plague, without however being convinced); Crone and Hinds, God's Caliph, 56 (where the Marwanids are credited with superhuman insight and immunity against minor mistakes); Qadi, 'Religious Foundations', 230 (where the late Umayyad secretary cAbd al-Hamld b. Yahya credits them with ilham, divine inspiration; compare Abu '1-Jarud on the ZaydT imam and Ibn al-Muqaffa' on the 'Abbasids, below, 104, i3of.). 34. Tab., ii, 1178.11. The Umayyads [ 43 modelled its own ecclesiastical hierarchy on the secular chain of command in the Roman empire, but it could not identify the two. Islam, by contrast, developed by taking over from the empire, imposing its own chain of command, and the idea that religious authority could be channelled through a separate hierarchy of offices does not appear to have occurred to the early Muslims. Rather, they seem to have assumed that since the imam was the best man of his time, everyone would automatically follow him of his own accord, as everyone had followed Abu Bakr and "Umar: one went where he went and did as he did, by following his kinsmen and governors and other local substitutes for him who were assumed to do as he did, too. There was no need for a separate hierarchy because right guidance was built into the power structure. In line with this, the Umayyads saw themselves as guiding the community by exemplifying the law in everything they did, in person or through their governors and judges, rather than by issuing directives to a clergy: They did make explicit pronouncements on the law when they adjudicated disputes and answered questions submitted to them by petitioners; they also sent legal instructions to judges on particular questions from time to time, and issued edicts on legal questions of a public nature, such as cUmar IPs famous edict on the status of converts and other fiscal matters. Bur they did not appoint ministers to represent them among their subjects at large. The task of answering everyday questions thus fell to men outside the government apparatus. Obviously, people could not travel to the imam's residence every time they needed to know whether a particular form of prayer was valid, whether maintenance was owed to a former wife, whether a particular food could be lawfully eaten, or the like. They could turn to the governor, who was officially charged with teaching the religion,3^ or to his deputy, the judge; and no doubt they did, especially for questions with a public angle. But neither the governor nor the judge was a local minister intimately involved with the lives of his parishioners after the fashion of a village priest. Where such intimate familiarity was desired, people would turn to local friends and neighbours credited with a better understanding of the religion than the rest. The latter developed into local scholars. They would answer questions on the basis of what they remembered upright men 111 the past to have done, what appeared right to them, or by second-guessing the imam's views, meaning that they would devise their own solutions and attribute them to him. The Umayyad caliphs themselves would consult such scholars at times, regarding them as their memory bank, and they seem to have granted some of them official authorization to 3>. This emerges from scatemenrs attributed to cUmar (lab., i, 2740.14, 2742.2) and al-Asamm's image of the Prophet's governors (Ps.-Nashi, §103). 44 ] The Umayyads issue legal responsa {fatwds).16 Caliphs and scholars were not initially rivals. But this was to change when the Umayyad claim to legitimacy was rejected. Umayyad mulk The rejection of the Umayyads was commonly formulated as an accusation that they had petverted the caliphate (or imamate) by turning it into kingship {mulk). We do not know when this accusation first appeared. It is often directed at Mu'awiya, who is said to have effected the transformation by designating his own son as his successor, or by introducing jails, bodyguards, forced labour and the like, or in ways unspecified.3'' But all these statements are clearly retrospective: the founder of the dynasty here stands for the Umayyads at large.38 Since our sources were compiled in 'Abbasid times, it has been suggested that their hostility to the Umayyads reflects 'Abbasid prejudice against the fallen dynasty, but this is unlikely to be right: the hostility to the Umayyads is too pervasive in the sources to reflect the change of dynasty. More probably, the denunciation of the Umayyads as mere kings began in Iraq in the mid-Umayyad period, when Umayyad government became more openly authoritarian than it had been in Mu'awiya's days and when scholars hostile to the regime emerged, to reject the Umayyad view of themselves as leaders of the vehicle of salvation.39 The equation of kings with oppressive rulers, which the reader has encountered before in this book, had both tribal and monotheist roots.'10 The prc-Islamic Arabs conceived of public life as an arena of competition between free men, who would fiercely resist the encroachments of others while trying to accumulate more power than everyone else themselves,41 and their attitude to kings [muluk, sing, malik) was correspondingly ambivalent. On the one hand, a king was an insufferable tyrant. "Since the time of 'Ad we have thought well 36. Cf. Crone and Hinds, God's Caliph, s;if.; Dhahahi, Tal'rikh, v, 185, where YazTd b. AbT HabTb (d. lzfth^t.) is described as one of the three men appointed by 'Umar II ro give responsa in Egypt. 37. AghanJ, xvii, 357; YT, ii, 276 (cf. 'And al-Malik b. Hahib, Ta'rikh, no. 343); BA, iv, a, 125; Goldziher, MS, ii, 3if. 38. Cf. SafTna's statement that "the Umayyads are the worst kings, and the first king was Mu'awiya" (in Ibn Abu Shayba, xiv, no. 17854, and elsewhere). 39. Schacht, Introduction, 23 (followed by Lambton, State and Government, 47); Crone and Hinds, God's Caliph, 23. 40. Above, 7; Lewis, Political Language, \3ff. 41. bor the coexistence of these attitudes, and an illuminating comparison with modern America, see Lindholm, 'Kinship Structure and Political Authority' and 'Quandaries of Command in Egalitatian Societies'; much of it also in The Islamic Middle East, ioff. The Umayyads \ 45 of capturing kings, killing them and fighting them," a pre-Islamic poet boasts; "we will tell you of the days long and glorious when we rebelled against the king and would not serve him," another echoes. When the poet al-Tirimmah al-Ta'I advised al-Husayn, in 680, to seek refuge the mountains of Aja, where "we held out against the kings of Ghassan, I.Iimyar, al-Nu'tnan b. al-Mundhir and everybody altogether," he was expressing himself in a wholly JahilT vein.42 But on the other hand, a king was also somebody to be admired, for there was nothing more prestigious than being muta', obeyed by others rather than obedient to them. The kings of Kinda, Hlra, and Ghassan were as respected and envied as they were resisted. 'King' could thus be a complimentary term no less than one of abuse. A pre-Islamic ruler hyperbolically identifies himself as "king of all the Arabs" in the famous Nemara inscription of 328, and the term often appears in a flattering vein in Umayyad court poetry, and indeed in other contexts in which it is not implicitly or explicitly contrasted with imam or khalifa.^ The negative evaluation of kingship was further developed when the Arabs converted to Islam. This, one assumes, is when they came to equate it with the pursuit of private interests at the expense of public welfare (when kings enter a town, they corrupt it, as the Queen of Sheba notes in Q. 27:34, though curiously the exegetes make nothing of this verse). The Prophet identified God as king and reserved all power for Him, thereby rendering all human use of it illegitimate unless it was used for God's ends. God was practically synonymous with the community. He was perceived as a higher authority who had miraculously intervened to regulate the competition between selfish human beings by setting the rules of the game, and He stood for the interests that the believers had in common.44 Wherever God had a claim to something, humans had none: this was why the idea of being governed by God struck the Muslims as so liberating. An imam or caliph was a ruler who recognized all power as God's and used it in accordance with His will, and government by such a person accorded with the wishes of his subjects, for all believers accepted God's rules. By contrast, a king would treat his power as his private property and use it to further his own individual interests (ahwc?, whims), which necessarily meant that he had to subjugate other people by force, depriving them of their freedom to act, inasmuch as their interests would conflict with his. "I am not a king who will 42. Abu Tammam, Hamasae Carmina, i/i, 195; cAmr b. Kulthüm in Arhcrry (rr.), Seven Odes, 205; Tab. ii, 304. Cf. also the Basran who boasted, in 684, of "how many a brutal and violent tyrant we have killed" (Tab. ii, 456); Arhamina, 'Tribal Kings', 36f. 43. Beeston, 'Nemara and Faw', 3; also cited in Lewis, Political Language, 54; Tyan, Califat, 3«2f. 44. See further below, 393. 46 The Umayyads enslave you, only a servant of God who has been offered a trust," as 'Umar is reputed to have said.4j In short, imama/khilafa stood for theocracy, government by God, whereas mulk stood for autocracy; government by selfish, arbitrary, and shortsighted human beings. Thanks to Islam, kingship came to be seen as both an infidel and a non-Arab institution. It must of course have been strongly associated with non-Arabs already in pagan times, by far the most famous kings being those of the Persians and the Byzantines; but non-Arabs were not branded as infidels in those days, and the Arabs themselves had been ruled by kings from time to time. The Muslim polity was however created without recourse to the institution, and the Muslims soon set out to defeat the kings of this world, with the result that kingship came to seem incompatible with both Islam and the Arab tradition. "Servants of God, a king in the Tihama!," a pre-Islamic Meccan, here cast as a proto-Muslim, indignantly exclaims in a story of how a QurashI tried to have himself accepted as client king on behalf of the Byzantines in Mecca.46 "He has gone Christian!," the troops exclaim in equal indignation in a story of how the son of the conqueror of Spain was persuaded by his Visigothic wife to wear a crown.47 It was with good reason that Mu'awiya wore no crown.48 Kingship was illegitimate human self-aggrandizement which the Muslims had come to eliminate. On all this the Muslims of the Umayyad period were agreed. But God did not rule the believers directly, so who represented Him in the here and now? This was the question that became deeply contentious. The Umayyads claimed that they did, but their enemies held them to have forfeited their position by allowing private interests to override those of the community and generally allowing power to be used in the manner characteristic of the conquered nations before the coming of Islam. This was the force of the accusation that they had turned the caliphate into kingship - kingship like that of the Persians and Byzantines (kisrawi, birqaliiqaysan), as some added.49 But if the Umayyads had no right to the caliphate, who did, and with precisely what competence in matters of religious guidance? Was it lawful to rebel against the 45. Tab., i, 2368. Cf. also Ibn Sa'd, iii/i, 2.21 (B, 3061.), where 'Umar is told that a caliph only takes what is due and spends it where it is meant to be spent, whereas a king oppresses people by arbitrarily taking from some and giving to others. 46. Fast, Shiftf al-gharam, ii, 171.-5. 47. Akhbar majnwfa, 20; Ibn 'Idhan, Bayan, ii, 23f. 48. Cf. rhc Maronite chronicle in Hoyland, Seeing Islam as Others Saiv It, 136. 49. AgbSm, xvii, 357; Jahiz, 'H 'l-nabita', in his Kasifil, ii, lot; cf. BA, iv, a, 125.17. Kisra (Khusraw), Hirqal (Heraclius) and Qaysar (Caesar) arc here generic terms for the Persian and Byzantine emperors. The Umayyads head of state to enthrone the deserving candidate, or should one put up with oppressive government for the sake of communal togetherness? And if right guidance was not to be found with the head of state, where was one to look for it? These were the questions that increasingly preoccupied people as the Umayyad period wore on, especially in Iraq, which had the largest concentration of Arabs outside Arabia without being the centre of power.