Ninth-Century Muslim Anarchists

I: INTRODUCTION

The people with whom this paper is concerned were anarchists in the simple sense of believers in an-archy, 'no government'. They were not secularists, individualists, communists, social reformers, revolutionaries or terrorists, merely thinkers who held that Muslim society could function without what we would call the state. Their view is, however, of great interest from the point of view of early Islamic political thought and the history of anarchism alike. Since they are largely unknown even to Islamicists and have yet to be discovered by historians of anarchism, I am grateful for the opportunity to present them to a wider public here. All the anarchists came from Basra in southern Iraq or had their intellectual roots there, but they belonged to two quite different groups. Most of them were Mu'tazilites, that is members of a theological school of Basran origin distinguished by its reliance on reason. Mu'tazilites were not necessarily, or even usually, anarchists, but a ninth-century Mu'tazilite heresiographer presumed to be Ja'far ibn Harb (d. 850) implies that belief in the non-necessity of government was common among them in his days. Its adherents included al-Asamm (d. 816 or 817), al-Nazzam (d. between 835 and 845),4 Hisham al-Fuwat1 (d. 840s?) and his pupil 'Abbad ibn Sulayman (d. 870s?), all of whom lived or began their careers in Basra, as well as the so-called Mu'tazilite ascetics (sufiyyat al-muftazila), active in Baghdad. The other anarchists were Kharijites, that is to say, members of a mainly Basran sect which was notorious for its militant intolerance. The Kharijites were not normally anarchists either, but one sub-sect was, that is the Najdiyya, or Najadat, who had appeared in the seventh century and who seem to have survived into the tenth, possibly in Basra and possibly elsewhere. Whether Mu'tazilite or Kharijite, the views of the anarchists have been poorly preserved. Numerous sources mention that some Mu'tazilites and Kharijites denied the necessity of the imamate (roughly translatable as legitimate government), but it was not until van Ess published the heresiography now presumed to be Ja'far ibn Harb's (generally referred to as Pseudo-Nashi') that their laconic statements could be related to a context. This new source also provided a clue to the identity of unnamed anarchists who appear in a fragmentary epistle by al-Jahiz (d. 869), a famous litterateur and Mu'tazilite of the non-anarchist variety: they can now be plausibly identified as Mu'tazilites influenced by al-Asamm. In addition, van Ess has done an immense amount of groundwork on the anarchists (without ever using that term) in his Theologie und Gesellschaft, a monumental work which covers the doctrinal developments of the early Islamic world in four volumes of prosopography and analysis, and two of translations. Without Pseudo-Nashi' and Theologie und Gesellschaft this article could not have been written. But numerous problems of textual interpretation remain, and this, in conjunction with the need to provide information for readers in different fields, accounts for what may strike the reader as annoyingly dense annotation. Anarchism in the simple sense of belief in the dispensability of government appears to have a continuous history in the West from the Bohemian Taborites of the 1420s onwards. Outside the Western tradition it is difficult to find. There is a case for the view that Chuang Tzu (fourth century BC) and other early Taoists should be classified as anarchists, but much that looks like anarchism is not, and the only non-Western example known to date apart from the Taoists appears to be the Muslim thinkers under discussion here. As one would expect, the three types of anarchist arrived at their convictions by quite different intellectual routes, having started from different premisses. The Taoists will have to be left aside here, but we may start with a comparison of the Western and the Islamic routes.

II: THE WESTERN PREMISSES

Western anarchism, medieval or modern, has its ultimate origins in the Western conviction that human society pre-dates the emergence of the state. The Western tradition abounds in claims that once upon a time humans lived together without coercive government in Paradise, the golden age, the state of nature, in primitive societies or before the development of agriculture. However formulated, the assumption is always the same: state and society are not inseparable, let alone identical. This may strike a modern reader as self-evident, but it is not. Its history takes us back to the Stoics. To the early Greek philosophers, including Plato and Aristotle, society and government developed together as two sides of the same coin: take away the one and you took away the other. But the Stoics distinguished between them in their famous accounts of what society would look like if it were based on natural law. Natural law was the right reason by which the universe was governed and on which the wise man would model his life. A society based on such reason would not have any law courts, private property, slavery, marriage or war; in other words, it would not have any structures of domination or organized violence: all these things were human conventions, not part of natural law. (Many other conventional institutions, including temples, education and coinage, would be absent too.) The Stoics were not anarchists. Their message was not that all these institutions could be, or ought to be, abolished. They did, however, lay the foundations for anarchism by assigning human sociability and human government to radically different sources: the one was natural, rational and good; the other not. The later Stoics said that in the golden age humans had actually lived in a society based on natural law, led by wise men; but then avarice had made its appearance, resulting in the development of private property, tyranny, slavery, war and so forth; in short, social and political inequality, coercion and strife had emerged. This view of human pre-history went into Cicero and other Latin sources that passed to the medieval West, and above all it went into the Latin Church Fathers, so that it became part and parcel of Latin Christianity itself. In its Christianized version it said that once upon a time, in Paradise or in some remote time on earth, humans had lived social lives without private property and slavery (though not without marriage), but that the Fall had so vitiated human beings that this was no longer possible. Kings had been instituted as a punishment for and remedy against sin; their authority derived from God Himself, however oppressively they behaved, and one had to obey them, but they did not form part of the original condition of innocence. The sinful, yet God-given, nature of power enabled medieval churchmen to stress the diabolical or celestial nature of government as they saw fit, and many held political subordination to have existed even in Paradise, where the existence of civil (as opposed to servile) subjection was to be explicitly endorsed by Thomas Aquinas (d. 1274) with reference to Aristotle's view of political organization as natural. But the view that government was unknown to God's original plan and to nature alike was too entrenched in Western thought to disappear, though it was often attacked. As a result, Westerners have always found it possible to think away the state. Some would think away society along with it, to illustrate how nasty, brutish and short life would be in the state of nature; but many would dream up societies from which the structures of domination had been removed, with reference to the remote past, the millenarian future, real or alleged primitive societies, or by way of construing utopias based on natural law or its socio-economic successor. In short, Western anarchism is in essence the belief that we can return to the condition of innocence from which we have fallen, or to some secularized version of it. Anarchist sentiments can thus be classified as endemic to the Western tradition, though they have rarely been epidemic. Differently put, if one thinks of an intellectual tradition as a box of conceptual tools with which every generation tries to carve some sense out of the world, the Western tradition has always had a tool labelled 'does God/nature really want us to have rulers?'

III: THE MUSLIM PREMISSES

But the Muslims started with a very different set of conceptual tools. As they saw it, structures of domination had always existed and always would, for the universe itself was a kingdom, in the most literal sense of the word. The king of the universe was God, who ruled by legislating. At first sight, divine law as conceived by the Muslims looks much the same as the natural law of the Stoics (who often called theirs divine as well); but the conceptions are quite different. The natural law of the Stoics was something built into nature, exemplified by nature, and available to all humans by virtue of their possession of reason; it was 'written into their hearts', as St Paul put it, and thus wholly independent of human government. But the divine law of the Muslims was envisaged on the model of positive law as something that had to be enacted, promulgated and enforced within a particular community: the King had to send messengers in order for people to know it, and He had to raise up deputies of one kind or another in order to have it executed. Far from being independent of human government, divine law engendered it. You acknowledged God as your king by accepting membership of His polity, to live by His law as brought and executed by His agents. God's government was coercive. He would not, of course, have to use force if His subjects would obey Him of their own accord, but for some reason or other they all tended to be rebellious. There was nothing special about humans in this respect. God had sent armies against disobedient creatures even before humans had been created, and the human fall plays no role whatever in the Muslim view of why coercive government exists. Government had always existed and always would; it was an inescapable feature of the universe. Consequently, the Muslim golden age myth is not about the absence of government, but rather about its ideal form. The myth is set in Medina in the time of the Prophet and the first caliphs, from 622 to 656 (or earlier), that is in a well-remembered historical period rather than the hoary past, and what it offers is an idealized version of that period rather than complete fiction. Like the Stoic account, it describes a simple society which was guided by wise men until things went wrong, as they did when the first civil war broke out in 656, if not before. But unlike its Stoic what counterpart, it starts with the foundation of a polity, and it illustrates is not a contrast between divine law and human government, but on the contrary their fusion. The Prophet and the first caliphs who bring and execute God's law are unambigu- ously envisaged as rulers, not just as wise men. They impose penalties, conduct campaigns, suppress revolts and start the wars of conquest; in short, they use institutionalized violence. But they always do so in accordance with God's law. Nothing is wrong with coercive was institutions as long as they are properly used: that the basic position. Ideal government was government by an imam, a communal leader who modelled himself on God's law and who thus set an example to be imitated. The first imam in human history was Adam. The first imam in Islamic history was Muhammad; the imams after him adopted the title of caliph, and their position was thereafter known now as the imamate (which stressed its legitimate nature) and now as the caliphate (which stressed its political reality). But they were all rulers of the same kind. Everything else was a corruption, in two opposite directions. On the one hand, some people transgressed against God by arrogating His power to themselves, leading to tyranny. This was the were condition under which the non-Arabs had lived until they conquered by the Muslims. More precisely, they had lived under kings, but all kings other than God Himself were tyrants, for a king was somebody who wanted power at God's expense, like Pharaoh, the paradigmatic example. On the other hand, there were people who forgot about God and His law altogether and so had no government at all. Statelessness was the condition in which the Arabs had lived before the rise of Islam. The Greeks had rather admired them for their ability to do without rulers, and they had certainly admired themselves for it: they boast endlessly of their refusal to submit to kings or anyone else in their poetry. But after the rise of Islam they realized that they had lived in pagan ignorance and barbarism, Jahiliyya, a state of amorality and disorder, not a condition of innocence, let alone one which established a natural right to freedom from subordination. Obligation, subordination and order all came with the revelation, for a religion was first and foremost a set of legal and moral obligations whereby human society was ordered. The Medinese caliphs steered a middle course between tyranny and anarchy by adhering to God's law. To the vast majority of Muslims they represented the political ideal, as indeed they still do. In short, coercive government was not a mere human convention, except in so far as it had been perverted by kings. In its authentic form it was a sacred institution which reflected the absolute. You could not have a moral order without a revealed law, and you could not have a revealed law without an imam to enforce it. This was the premiss with which the Muslims started. It is not easy to see how they could get to anarchism from there.

IV: FROM IMAMATE TO KINGSHIP

Like everyone else, however, the Muslims soon discovered that divine law and human government tended to be at loggerheads. By about 800 Medina had long ceased to be the capital, the Muslim polity had long lost its simplicity and the imams had long ceased to be wise men dispensing friendly guidance, in so far as they ever were. The 'Abbasid caliphs ruled a vast empire from Baghdad in a style all too reminiscent of Pharaoh and his likes. The imamate had turned into kingship, as people said; in other words, it had turned into tyranny. The question was what one should do about it. Islam had originated as an activist religion, and there were still people who said that wherever you saw people act wrongly, you should take action against them, with the sword if necessary: if the ruler misbehaved, one had to rebel and replace him with another, provided that there was a reasonable chance of success. Most ninth-century Mu'tazilites were of this opinion, as were all Kharijites of the non-anarchist variety. But the religious scholars who came to be the bearers of Sunni Islam were quietists, like the churchmen of the medieval West. In their view, civil war was more destructive for the community than such wrongs as tyrants could inflict on it, and preserving the community was more important than setting its leadership right; you had to obey the ruler, however sinful he was, unless he ordered you to disobey God Himself, in which case you had to adopt passive resistance. There were even some who argued in the Christian style that tyrannical rulers were a punishment for sins. But the anarchists proposed a third solution. We may start with the Mu'tazilites, who will get the bulk of the attention.

V: THE MU'TAZILITE ARGUMENTS

The Mu'tazilites offered a variety of arguments in favour of anarchism, but only one is quoted in full, that of the Mu'tazilite ascetics. It went as follows. Islam is different from other religions, for other religious communities have kings who enslave their subjects, but the Prophet was not a king, nor were his successors, and if an imam tsere to turn into a king, by ceasing to govern in accordance with the law, then the Muslims would be legally obliged to fight him and depose him (as the activists said). But civil war was indeed terrible; it split the community and led to more violation of the law without guaranteeing a better outcome (as the quietists said). Since imams kept turning into kings, the best solution was not to set them up in the first place. The Mu'tazilite ascetics did not deny that there might be a legitimate ruler in the future: they seem to have thought that he would have to be an 'Alid, or in other words a descendant of the Prophet. But in the absence of such a ruler it was better to have none. Al-Asamm's argument, which has to be pieced together from diverse passages, was based on the premiss that the imam was a ruler on whom all members of the community agreed; without such consensus he would not be an imam at all. This was widely accepted (and also the premiss of the Najdiyya). Originally, the caliphs ruled with communal agreement and had thus been true imams (according to al-Asamm, though not the Najdiyya), but nowadays they did not: the community had grown too big. Like the ascetics, al-Asamm seems to have kept open the possibility that there could be a true imam again one day, though he cannot have regarded it as likely if he saw size as the key problem; and he certainly did not think that the imam would have to be an 'Alid, or even an Arab. In any case, one had to look for alternatives while such a ruler was absent. Hisham al-Fuwatl subscribed to a variant version of alAsamm's argument. According to him, the community only needed an imam when it was unanimous and righteous, by which he appears to have meant that it was only under such circumstances that it was possible (and obligatory?) to elect one. In the past it had been possible, but nowadays it was not: the community had grown disunited and sinful. His pupil 'Abbad ibn Sulayman went so far as to declare in categorical terms that there never could be an imam again. Here too, it followed that one had to look for alternatives. The Mu'tazilite anarchists were clearly regretful anarchists. They would not have agreed with Emma Goldman (d. 1940) that 'all forms of government rest on violence, and are therefore wrong and harmful, as well as unnecessary'. In their view, this was only true of government in the sense of kingship; the imamate had been an exception. But it was no longer an exception. It had indeed come to rest on violence, and thus to be wrong, harmful and unnecessary. This was the problem they were grappling with. The Mu'tazilites (as also the Najdiyya) declared the imamate unnecessary by denying that it was prescribed by the religious law. A Muslim had to pray, fast and fulfil other duties laid down by the law, but having an imam was not a duty of that kind, they said. They demonstrated this in different ways. Al-Asamm, followed by al-Nazzam, famously declared that people would not need an imam if they would obey the law of their own accord. By this he does not seem to have meant that such a situation could actually be brought about, but rather that since one could envisage a situation in which the imamate was superfluous, one could not identify the institution as obligatory on the basis of reason. Nor was it prescribed in the Qur'an (as all or most nonShi' ites seem to have agreed at the time); and according to al-Asamm's presumed followers described by al-Jahiz, the behaviour of the Prophet's Companions after his death also ruled out that it had been prescribed by him. In short, no legal obligation to have the institution existed. 'Abbad ibn Sulayman, perhaps echoing Hisham al-Fuwatl, used the very fact that doubts about the possibility of having a legitimate imam had arisen to demonstrate that none could appear any more, presumably inferring that therefore no obligation to have the imamate could exist (any more?); it was generally agreed that God did not impose impossible duties on the believers. How the ascetics argued we do not know, but one way or the other they all denied that the imamate was God-given. In other words, they all desacralized it: it did not reflect the absolute; it was just a fallible human institution like any other (min mu'amalat al-nas, as al-Asamm put it). They did not say that it was a bad institution or that God originally meant people to live without it; they merely denied that God had any views on its desirability or otherwise. Given that the imamate was simply a human convention, one could have it or not as one saw fit: people had had it in the past, and it had worked very well, but nowadays it was preferable, or even necessary, to do without it. In short, they cut the link between the imamate and the law on which Islamic society rested. That was how they made anarchism possible.

VI: DOING WITHOUT THE IMAM

But how would one manage without the would apply the law and imam? Above all, who dispense the penalties of which it was so-called hudud, that is generally agreed that they could only be applied by the imam or his circumstances no private representatives? Under normal another believer. The person was allowed to kill or maim law did, however, stipulate that certain crimes (such as theft, adultery and wine-drinking) were punish- able by death, amputation or flogging; these penalties were among 'God's rights' (huquq allah), i.e. they were required for the greater good, and a public figurehead was required for their This obviously suggested execution. that having an imam zvas a legal duty, and itis presumably for this reason that the hudud loom large in the military surviving matters discussions. without an The question of how one might manage imam attracts less attention, and there is no discussion at all of where one would find religious guidance, for all that the ninth-century Mu'tazilites generally saw the imam as religious guide or teacher. But maybe this simply reflects the fragmentary state of the evidence. In any case, the Mu'tazilites gramme of moral responded partly with a proAs regards the former, rearmament and partly with practical proposals. they harped on the theme of cooperation and taking duties seriously. degree 'People's welfare lies entirely in the to which they Everybody had to cooperate', as some of them pointed out. participate, no people guilty of crimes had shirking was allowed; even to do their bit, by giving themselves up voluntarily. As regards the latter, the proposals ranged from dissolution of public complete authority to drastic decentralization.

Complete dissolution of public authority seems to be what Hisham al-Fuwat. and 'Abbad ibn Sulayman had in mind. Since no legitimate authority existed, people should take the law into their own hands whenever they could to ensure that the law was applied: self-help was encouraged even when it entailed killing, and even when the killing had to be done on the sly. Better still, people should rebel and openly take over the implementation of the law, including the amputation of thieving hands, the execution of murderers and everything else that imams used to do. This sounds like a prescription for anarchy in the normal sense of chaos and disorder. In a slightly less anarchic vein, other Mu'tazilites proposed that trustworthy and learned leaders of households, districts, tribes and towns should apply the law within their jurisdiction, and that they were qualified to carry out the hudud. In other words, power should revert to patriarchs and local leaders domestic tyrants and local thugs in modern parlance. But others were reluctant to do without public authority altogether. In their view one could elect temporary imams. This could be done whenever legal disputes arose or crimes were committed, or when the enemy invaded; the imam would lose his position as soon as he had finished the job, just as an imam in the sense of prayer leader (another meaning of the word) loses his authority the moment the prayer is over. One assumes that it was the above-mentioned leaders of households, districts, tribes and towns that they had in mind as candidates, but in any case these Mu'tazilites (apparently the ascetics) clearly wanted government to be taken over by elected officials. Al-Asamm played in sufficient numbers to minimize the danger of bias and collu- sion, they could replace the imam for purposes of maintaining the law and applying the hudud. In other words, one could have government by executive committee. Al-Asamm also advanced a proposal for extreme decentraliza- tion: one could have several, semi-independent imams. One imam had been fine in the days of the first caliphs, he said, but nowadays there could no longer be real unanimity on just one man, and he could not know meritorious people in distant provinces, meaning, as al-Asamm saw it, that he could not collaborate with provincial elites; this, he said, was frustrating for those who wished to participate in government. Hence it would be better to have several imams, and this was perfectly lawful. He ought to have continued that since the imamate is a human convention, we can do what we like with it, but at this point he seems to have lost his nerve: he tried to legitimate his proposal by invoking Prophetic precedent. He claimed that the Prophet's governors in Arabia had in effect been independent imams. Each one had collected taxes, maintained order, conducted defence and taught people the law; and when the Prophet died, the inhabitants of each provincial centre had inherited the right to appoint such governors of their own. In short, all provinces were now entitled to elect their own semi-independent rulers, who would, of course, have to cooperate. Al-Jahiz scoffed at this proposal (as known to him from al-Asamm's followers): who ever heard of neigh- bouring rulers who did not fight? But what al-Asamm was grappling with was clearly the concept of federation. No such concept existed, and he did not quite arrive at it either, for he did not explain what would hold the governors together now that the Prophet had died. But it is none the less a remarkable piece of political thinking, and it is hardly surprising that he lost his nerve, for to propose that the Muslim community should be divided up among a number of imams was even more heretical than saying that it could do without imams altogether. Al-Asamm never dared to publish this proposal; he merely told his close friends and pupils (khazvass ashabihi) about it.

VII: GREEK OR TRIBAL ROOTS?

Some eighty years ago Goldziher proposed that al-Asamm and Hisham al-Fuwatl found their anarchism in a Greek letter supposedly written by Aristotle to Alexander after the latter's conquest of Iran; but this is unlikely and has rightly been rejected by van Ess. The letter makes two points reflecting the Hellenistic debate for and against monarchy (rather than government as such): first, 'many people think that a ruler upholding the law is only necessary in times of war; when the war is over and security and calm prevail, one can do without him'; and secondly, 'some think that people should all be equal, without any ruler or subject among them', while others go to the opposite extreme of deeming it acceptable for the ruler to be 'coercive in disregard of the law'. The letter was probably translated into Arabic in Syria in the 730s-740s, and a formulation reminiscent of the second point reappears in a work by the Iraqi secretary Ibn al-Muqaffa' (d. c.757), suggesting that it was read in Iraq well before the time of al-Asamm and Hisham. But it is hard to see how it could have inspired them, for the concept of the ruler as an emergency leader in war was alien to both of them, and neither wished to dispense with the imam on the grounds that people were, or ought to be, equal. The postulate of Greek influence is in any case unnecessary. Al-Asamm and Hisham formulated their ideas in interaction with their Kharijite neighbours (who were not given to reading Greek works), and both the Mu'tazilite and the Najdite anarchists were clearly drawing on tribal tradition which lies behind all early Islamic political thought of the type which may be loosely identified as libertarian. It should be stressed, however, that tribal ideas were at work only in the sense that they formed part of the value system of early Muslims, not as a model in their own right. There was no tradition for crediting tribesmen with the preservation of political virtues that the members of civilized societies had lost. Much later, in a fourteenth-century school text from Iran, one does encounter the argument that people do not need the imamate, for the bedouin manage perfectly well without rulers; but there is no way of telling where this argument comes from or how early it is. It could be of Najdite origin. But the Mu'tazilites never invoke the bedouin in the surviving texts, and the chances are that they envisaged tribal statelessness as every bit as bad as tyranny in that neither was based on God's law. They do not in fact invoke any concrete example of statelessness at all, and this is what is so remarkable about them: they were not thinking in terms of a return to some original condition; all were discussing new forms of political organization for which they had no example in either real or imagined history. In so far as one can tell, they simply reasoned their way to the view that one could live by the law alone, in conjunction with some local administration.

It could be argued that most of the Mu'tazilite proposals are not really anarchist in that most of them replace one type of government with another instead of abolishing it altogether. But this is true of most anarchist proposals: the alternative to the state is more often than not authoritarianism of another, and frequently more thoroughgoing, kind. The main difference between Mu'tazilite and Western anarchism is that the Mu'tazilites only proposed political alternatives to the imamate, whereas Western anarchists have usually proposed social regimentation and/or extreme simplicity of life in order to do without the state. Western anarchism has always been as much about socio-economic reorganization as political reform, almost always in an egalitarian vein, and usually communist (thus already the Taborites). Neither government nor private property had existed in the state of nature; the former had come into existence for the protection of the latter, as everyone knew from Cicero, so the two had to be abolished together. But Mu'tazilite anarchism was not concerned with social reorganization at all, nor was it egalitarian, let alone communist. The Mu'tazilite ascetics did postulate a connection between government and property, but what they said was not that both were intrinsically wrong, only that both were wrong unless they were based on Islamic law. This they no longer were. The abode of Islam had turned into an abode of unbelief, as one of them declared, meaning that collective life no longer had any legal or moral foundations. The illegitimacy of the ruler vitiated all titles to property: all ownership was really usurpation until the rightful imam appeared; making a living in any manner involving buying and selling was forbidden; all income was sinful, apart from such scraps as one received by begging in extreme need. (How they viewed living off the land we are not told: all were clearly urbanites. ) Had the head of state been legitimate, property and income therefrom would have been lawful too. In so far as the Mu'tazilites postulated a relationship between property and their political problems, their view was thus that wrongful government made property immoral, not that property engendered wrongful government. Far from construing their inegalitarian society as a source of caliphal tyranny, they all saw it as an alternative to it: society would be fine if only it were left alone; patriarchs and local leaders would dispense the law. Their anarchism consisted in thinking away the head of state and, implicitly, his army and bureaucracy too, in order to replace the whole apparatus of central government with either provincial imams in federation or local imams elected for a term, or with executive committees, or simply with the leaders of households and tribes as they were, or with straightforward self-help. But as regretful anarchists, not one of them condemned the state on principle. What they minded was not the existence of coercive power but rather its distribution. The distribution of power in the ninth-century caliphate was in fact extremely lopsided. The 'Abbasids tended to recruit their soldiers and governors in one province, eastern Iran, and their bureaucrats in another, lower Iraq; by and large, all others were excluded from decision-making at a central level, however influential, wealthy or meritorious they might be in local terms. This was to get worse, for instead of broadening their power base the caliphs decided, from the mid-ninth century onwards, to import Turkish tribesmen as slaves and to train them as soldiers and government servants, so that central government came to have even less anchorage in Muslim society than before. This had not happened by the time al-Asamm wrote, but it was where things were going, and his federation was undoubtedly meant to counteract this trend. He wanted more local participation. The same would appear to be true of the other Mu'tazilite anarchists.

IX: PUTTING ANARCHISM INTO PRACTICE

None of the Mu'tazilite anarchists explained how one was to do away with the state. All seem to have made their proposals in what Dawson calls a 'low utopian' vein, meaning that their programme of radical reform was destined for eventual implementation, if possible, but that meanwhile it served the eminently useful function (shared by all utopias) of providing a critique of existing institutions. Since all were scholars devoid of political experience, they may have held the practicalities of implementation to be for others to work out. We know next to nothing about their social status or sources of income, but most of them seem to have been happy to enjoy the comforts available under the protective cover of the state, however despotic or illegitimate it might be. Hisham al-Fuwat.l, possibly a wealthy trader, is even said to have frequented the court. Only the Mu'tazilite ascetics of Baghdad appear to have kept their distance from rulers and the normal comforts of life alike, but their withdrawal gave them greater affinities with the mystics, with whom the term sufiyyat al-mu'tazila brackets them, than with political reformers. None of them displayed any interest in fomenting rebellion, whether they considered it lawful or not; and al-Asamm positively ruled it out. An opportunity did, none the less, arise. In 817 the government collapsed of its own accord in Baghdad. There had been a civil war (the fourth); the new caliph al-Ma'mun was still absent, and his governor could not maintain order. The result was complete lawlessness, to which a certain Sahl ibn Salama responded by founding a famous vigilante group that proved quite effective. This man has turned out to be a Baghdad Mu'tazilite, possibly of the anarchist variety. His brief career certainly made a deep impression on the anarchists. 'At a time when government disintegrated and the plebs and ruffians took over . . . we saw a small number of people of integrity and standing get up in their district, tribe, street and quarter to . . . subdue the ... ruffians so that the weak could once more move freely . . . and so that merchants could go around again', they boast in al-Jahiz's account of them. This was devolution in action. The anarchists concluded that when people are forced to rely on themselves, they discover talents they did not know they had. People should wake up: the so-called shepherd would resume oppression as soon as he recovered his strength. He did in fact recover his strength, so that was the end of that.

X: THE NAJDIYYA

This brings us to the Najdiyya, who can be dealt with rather more briefly.79 The Najdiyya were almost certainly the first to deny the necessity of the imamate, and the reason why they did so, in so far as one can tell, is that they wished to shed the obligation to rebel. By origin they were activists: one had to fight to replace the illegitimate caliph of today with a true imam. They had in fact started their history by rebelling, in the late seventh century, but their revolt had been suppressed, which left them with the choice between trying again or modifying their doctrine. They chose the latter, we do not know exactly when: their belief in the dispensability of the imamate is not attested until the end of the ninth century. But the implications are clear enough: if the law did not prescribe an imam, one did not have to rebel to set one up; one could be a Kharijite without committing oneself to establishing a true imamate; one could live under the illegitimate 'Abbasid caliphs without endangering one's chances of salvation. But the Naidiyya soon developed a further reason to deny the necessity of the imamate: they did not want an imam to lay down the law to them. It is this second concern which is uppermost in their argument as we have it. Unlike the Mu'tazilite anarchists, who merely held the imamate to have become inoperable, the Najdiyya denied that it had ever existed. An imam was someone on whom everyone agreed, they said, but perfect agreement was inconceivable in theory and had never been seen in practice: even the very first caliph, Abu Bakr, had encountered opposition, as everyone knew. He had not been an imam, then, though he had certainly been a good man. They classified him as a chief (ra>ts). From this they inferred that the obligatory nature of the imamate was a myth. By denying that the imamate had ever existed, they also denied that it was the only form of government compatible with Islam, so they did not thereby declare themselves to be anarchists: rightly guided chieftainship was still an option. But the question of political government did not interest them much, since they would have had to rebel in order to establish a ruler of their own whether they classified him as an imam or not. They said that if one were to establish a polity of one's own, then one could have a chief, though one was not obliged to have one, adding that he would have to be elected by the community, supervised by it and deposed by it if he was found to stray: he would merely be the community's agent. But this was the standard Kharijite view of the imam (a term the Najdiyya sometimes fell into using of their rightly guided chief as well). As far as religious guidance was concerned, however, their dismissal of the imamate was certainly meant in an anarchist vein. Even if a ruler existed, he would only be a chief, not an imam, meaning that he would not be empowered over anybody else in religious terms. All believers were entitled to their own opinions on law and doctrine on the basis of iitihad, independent reasoning, for all of them were equally authoritative. The believers were 'like the teeth of a comb', or 'like a hundred male camels without a single female riding camel among them': why should they defer to someone just like themselves? Just as there could never be sufficient agreement to establish an imamate, so there could never be enough to establish law: consensus (iima') was not a source of law at all. Everybody was responsible for his own road to salvation. Najdite Islam was a do-it-yourself religion. Politically and intellectually a Najdite would have no master apart from God. This was radical libertarianism, and it was achieved at a cost. The Naidiyya held themselves to be the only Muslims. A11 others were infidels who could in principle be enslaved, dispossessed and exterminated by the Najdiyya, should the latter choose to rebel.83 In practice the Najdiyya seem to have lived in perfect amity with their so-called infidel neighbours. But they continued to regard the latter as outsiders, and this meant that they did not have to consider them in their political thought. What they were writing about was a tiny community in which people probably knew each other face to face. They could allow a high degree of independent reasoning because there was almost certainly a high degree of consensus anyway, and they could dream of libertarian politics because they had no polity of their own. As a solution to the problems of how one might keep the Muslims from India to Spain together in a single political and/or religious community, the Najdite vision was no use at all.

XI: CONCLUSION

Aristotle's Greeks and the very first Muslims were political animals in much the same sense: both assumed the highest form of human life to consist in participation in the public affairs of a politically organized society, the polis (city-state) in the Greek case, the umma (the community founded by Muhammad) in the Muslim case. As the city-state was the only polity in which one could be free according to the Greeks, so the community founded by the Prophet was the only polity in which one could be a slave of God's, as the Muslims put it, meaning free of subjection to mere humans in this world and saved in the next. In both cases the conception was undermined by world conquest. What Alexander did to the polis, the Muslim conquerors did to their own community in Medina. People now had to come to terms with empires. The original types of polity survived, of course: the polis continued zuithin Alexander's empire, the umma continued as an empire. But they ceased to be coterminous with the arena in which people found the meaning of their lives. Real politics now meant kingship, which the Greeks and Muslims alike equated with enslavement. Real freedom now meant transcending politics, to find the meaning of one's life elsewhere. This is the ultimate background to the anarchists, and it is also what doomed them to extinction. The Mu'tazilites had not transcended politics. Unlike the Cynics and the Stoics, they were not saying that people should change their attitude to government and understand how unimportant it was in terms of the ultimate order of the universe. It was the Sufis who took that line. The Mu'tazilite anarchists were saying that people should change government itself. Al-Asamm and his pupils wanted their political participation back, even if it meant sacrificing the imamate. The Najdiyya wanted to keep their intellectual autonomy, even if it meant remaining a tiny minority. Both were to that extent backward-looking. The Mu'tazilites were perfectly realistic in their recognition that the imamate could no longer function as it had done in Medina, and al-Asamm was also right that political decentralization was on the cards. By the end of the ninth century the caliphate had broken up under semi-independent governors, very much as he said it should. But it did not break up in accordance with his prescription. The semi-independent rulers were military leaders who fought each other as much as al-Jahiz had said they would, and they were simply miniature versions of the caliph the tyrant that al-Asamm had wished to replace. By the tenth century these rulers had officially taken to calling themselves kings in a flattering sense. By the twelfth century the sources will routinely make statements such as that the king must ensure that his subjects 'do not take out the ring of slavery from their ears'. The anarchists must have turned in their graves. But whether one tried to live with this development or sought to transcend it, anarchism was dead. Of people contemplating life without mon- archs on a permanent basis in the Islamic world I do not know a single unambiguous example after the Mu'tazilite and Najdite anarchists had disappeared.

Institute for Advanced Study,

Patricia Crone