In the summer or autumn of AD 661, Mu‘awiya b. Abi Sufyan, governor of Syria since 639 and already acclaimed by his Syrian followers as caliph (khalifa), religious and political leader of the Muslim state, entered the Iraqi garrison town of Kufa. In historical tradition this event is seen as bringing to an end a bitter period of civil war among the Arabs, achieving the reunification under one ruler of all the territories conquered by them, and initiating the caliphate of the Umayyad dynasty of which Mu‘awiya was the founder. The dynasty was to rule for 90 years or so until its overthrow and replacement by that of the ‘Abbasids in 749–50.
The Umayyad dynasty was the first to emerge in the Middle East following the conquest of the region by the Arabs, a conquest which had begun in the 630s and was still continuing for much of the Umayyad period. Apart from this fact, however, what was the importance of the period of Umayyad rule, a period which in its details is often complex and confusing, and how has it traditionally been regarded by Muslims in relation to the history of Islam? The answer to the first part of this question is provided by discussion of the two concepts of islamisation and arabisation, referring to two related but essentially distinct historical processes.
The term ‘islamisation’ refers both to the extension of the area under Muslim rule and to the acceptance of Islam as their religion by peoples of different faiths, but in the Umayyad period the question is further complicated by the fact that Islam itself was developing from its still to us not completely understood origins into something approaching the religion with which we are familiar. One should not imagine that Islam as we know it came fully formed out of Arabia with the Arabs at the time of their conquest of the Middle East and was then accepted or rejected, as the case might be, by the non-Arab peoples. Although many of the details are obscure and often controversial, it seems clear that Islam as we know it is largely a result of the interaction between the Arabs and the peoples they conquered during the first two centuries or so of the Islamic era which began in AD 622.1 During the Umayyad period, therefore, the spread of Islam and the development of Islam were taking place at the same time, and a discussion of islamisation has to begin with some consideration of the importance of the Umayyad period for the development of Islam.
In the first place, it was under the Umayyads that there began to emerge that class of religious scholars which eventually became the leading authority within Sunni Islam and which is chiefly responsible for shaping the historical and religious tradition which has come down to us. In effect, it was this class which led the development of Islam as we know it, and it is important to remember that it emerged largely in opposition to the Umayyad government. The Umayyads had their own conception of Islam, itself developing with time and different circumstances, but on the whole we see the religion from the viewpoint of the religious scholars.
In the emergence of this class the most important region was Iraq, and in Iraq Kufa was the leading centre. Other regions tended to follow its lead. Building on and reacting against the ideas and practices available in Kufa and other centres, from the second half of the Umayyad period onwards groups of Muslim scholars tried to develop and put on a sound footing what they saw as a true form of Islam. In doing so they frequently accused the Umayyads of impious or unislamic behaviour.
The main concept which these scholars developed and worked with was that of the Sunna. This idea went through several stages but increasingly came to be identified with the custom and practice of the Prophet Muhammad, which was to serve as the ideal norm of behaviour for his followers, and was eventually accepted as the major source of Muslim law alongside the Koran. Increasingly, Muslim ideas, practices and institutions came to be justified by reference to the Sunna, the words and deeds of Muhammad as transmitted by his companions to later generations. The proponents of the Sunna as thus understood became increasingly influential, and political and religious developments after the Umayyads had been overthrown resulted in the final crystallisation of the Sunni form of Islam with the religious scholars, the guardians of the Sunna, as its leading authority.2
Not all Muslims, though, accepted the primacy or even the legitimacy of the Sunna, and the Umayyad period also saw the emergence of the two other main forms of Islam, Shi‘ism and Kharijism. Tradition dates the fragmentation of a previously united Islam into the three main forms which we know today (Sunnis, Shi‘ites and Kharijites) to the time of the first civil war (656–61), which ended with the accession of Mu‘awiya to the caliphate. However, just as the development of Sunni Islam was a slow process which only began under the Umayyads, so too Shi‘ism and Kharijism were not born in one instant. They too developed in opposition to the Umayyads, in a number of distinct movements which each had individual characteristics, and again Iraq was of prime importance.
Kufa was the centre of the development of Shi‘ism in the Umayyad period. As early as 670, but especially after the revolt of Mukhtar in 685–7, Kufa saw a number of movements aimed at overthrowing the Umayyads and appointing a relative of the Prophet, usually a descendant of his cousin and son-in-law ‘Ali, as imam, which title the Shi‘ites tend to prefer to caliph. Where these Shi‘ite movements differed from one another was in the particular member of the Prophet’s family whom they favoured and in certain other doctrines they developed; what they had in common was devotion to the Prophet’s family and insistence that membership of it was a sine qua non for the imam. Some of them developed more extreme beliefs, such as acceptance of the imam as an incarnation of God and a doctrine of the transmigration of souls. It seems that from an early date the conquered non-Arab peoples were attracted to the Shi‘ite movements, and it may be that some of their doctrines were influenced by the previous beliefs of these non-Arab supporters. Shi‘ism has a long and complex history which extends well beyond the Umayyad period, but it was then that its basic character was established.3
The basic principle of Kharijism was a demand for piety and religious excellence as the only necessary qualification for the imam, and a rejection of the view that he should belong to the family of the Prophet, as the Shi‘ites demanded, or to the tribe of the Prophet (Quraysh), as the Sunnis required. Like Shi‘ism, Kharijism too was manifested in a number of movements, some relatively moderate and others more extreme. The extremists tended to insist on the rejection of all other Muslims, regarding them as infidels and therefore liable to be killed unless they ‘repented’ and ‘accepted Islam’, that is, unless they recognised the Kharijite imam and accepted the Kharijite form of Islam. This fierce rejection of other Muslims, however, involving the duty of rebellion against what was regarded as an illegitimate government, became increasingly difficult to maintain except in areas remote from the authority of the government or in times when the authority of the government for some reason collapsed. In Basra, the second of the Iraqi garrison towns, on the other hand, a more moderate form of Kharijism was elaborated and spread to eastern Arabia and North Africa. It is this form of Kharijism which has survived into the modern world.4
Each of these three main Muslim groups came to hold that Islam should be open to all peoples and that all should enjoy the same status within it regarding rights and duties. The development of this idea too, of Islam as a universal religion, can be traced to the Umayyad period, again in circles opposed to the dynasty.
Although it can be debated whether the Koran was addressed to all men or to the Arabs only, the Umayyads and the Arab tribesmen who first conquered the Middle East regarded their religion as largely exclusive of the conquered peoples. There was no sustained attempt to force or even persuade the conquered peoples to accept Islam, and it was assumed that they would remain in their own communities paying taxes to support the conquerors. Although from the start there was some movement of the conquered into the community of the conquerors, the separation of Arabs from non-Arabs was a basic principle of the state established as a result of the conquests. This is clear both from the procedure which a non-Arab had to adopt in order to enter Islam and from the fact that there were, from time to time, official measures designed to prevent such changes of status. Islam was in fact regarded as the property of the conquering aristocracy.
In order to attach himself to the religion and society of the Arabs, a non-Arab had to become the client (mawla, pl. mawali) of an Arab tribe. In other words, in order to become a Muslim, something which it is possible to see as a social or political as much as a religious move, he had to acquire an Arab patron and become a sort of honorary member of his patron’s tribe, adding the tribal name to his own new Muslim one, even though he and his descendants were in some ways treated as second-class Muslims. It is evident, therefore, that membership of Islam was equated with possession of an Arab ethnic identity.5
Nevertheless, association with the elite in this way did have advantages for some, and at various times in different places we hear of large numbers of non-Arabs attempting to enter Islam by becoming mawali but being prevented from doing so, or at least from having their changed status recognised, by local Umayyad governors. Probably the best-known example was in Iraq around 700 when large numbers of local non-Arab cultivators sought to abandon their lands and flee into the Arab garrison towns to enter Islam as mawali, only to be forced back by the Umayyad governor al-Hajjaj who refused to recognise their claims.
In the long run it proved impossible to maintain the isolation of conquerors and conquered from one another in this way, and attempts to do so only served to alienate further those Muslim groups which had come to see Islam as a religion open to all. The problem for the Umayyads was that they had come to power as leaders of a conquering Arab elite and to have allowed the conquered peoples to enter Islam en masse would have abolished or at least weakened the distinction between the elite and the masses. The crucial privileges of Islam, from this point of view, were in the area of taxation. In principle the Arabs were to be the recipients of the taxes paid by the non-Arabs. If the conquered peoples were allowed to become Muslims, and to change their position from that of payers to that of recipients of taxes, the whole system upon which the Umayyads depended would collapse. But as the pressure from the non-Arabs built up, and the universalist notion of Islam became stronger, this problem became increasingly urgent for the dynasty and played a major part in the generally negative attitude of Muslims towards the Umayyad dynasty.6
How far the development of Islam in the Umayyad period involved radical changes in religious practices or beliefs is not easy to say. Broadly speaking, Muslim tradition assumes that the fundamental institutions of Islam—such things as belief in Muhammad as a prophet, acceptance of the Koran in the form in which we know it as the word of God, and performance of the main rituals such as the five times daily prayer (salat) and the annual pilgrimage to Mecca (hajj) —existed at the beginning of the Umayyad period and were accepted equally by the Umayyads and their opponents. The difficulty is to decide how far our Muslim sources, which are relatively late in the form in which we have them, are reading back later conditions into an earlier period.
Sometimes, certainly, we have hints that the situation was not so static or so uniform as the tradition generally implies. For example we are told that Muslim rebels supporting Ibn al-Ash‘ath against the Umayyads in the early years of the eighth century accused the caliph of ‘murdering’ the ritual prayer (salat) and called for vengeance for it, although what this meant and what exactly was involved, if anything specific, is not spelled out.7 Even such tantalisingly obscure hints are relatively scarce, and when we do sometimes have more substantial information its significance seems often to be limited in one of two ways.
First, the information may centre on a point which seems to be relatively minor. For instance, much play is made with the charge that the Umayyads insisted on delivering the khutba (in the early period a speech or sermon given usually in the mosque by the caliph or his representative and often dealing with secular as well as more purely religious affairs) while sitting, contrary to what is alleged to have been the practice established by the Prophet and his immediate successors. This is supposed to be a sign of the haughtiness of the Umayyads, refusing to stand before their subjects and preferring, like kings, to remain seated. Even though the detail may have lost some of its significance because of the later decline in importance of the khutba and its associated institutions and ceremonies, however, it is difficult to see arguments about the correct posture for the khutba as of fundamental importance for the development of Islam. In the way in which the practice is presented by Muslim tradition, it does not provide grounds for arguing that the outward forms of Islam underwent great and radical changes under the Umayyads.8
Secondly, even when the information is apparently more weighty, the impression is usually given that the Umayyads were perverting some orthodox practice or belief which already existed and was widely accepted by Muslims. There is no suggestion that basic religious ideas were still in a state of flux and that ‘orthodoxy’ (an ambiguous term in Islam since there is no central authority to say what is and what is not orthodox) was only slowly developing. We are told, for instance, that some of the Umayyads tried to make Jerusalem a centre of pilgrimage, but the sources imply that this was against the background of an already generally accepted practice of annual pilgrimage to Mecca which had been established as the cultic centre of Islam from the time of the Prophet. The reader should be aware of such preconceptions in the sources and consider the possibility that there may not have been, as yet, any firmly established cultic centre in Islam.9
Any attempt to argue that there were during the Umayyad period more fundamental religious developments than the sources allow for, therefore, involves a certain amount of ‘reading between the lines’ of Muslim tradition and using whatever evidence is available outside the Muslim literary sources. A recent discussion using such methods has questioned whether the name ‘Islam’, as the designation for the religion of the Arabs, existed much before the end of the seventh century.10 Muslim tradition itself, though, has proved remarkably impervious to analysis with such questions in mind, and one’s attitude to the question of the extent of the religious development of Islam in the Umayyad period must depend greatly on one’s attitude to the value of Muslim sources for the history of the period, and especially the earlier part.
The spread of Islam during this period, as already indicated, has to be viewed on two levels, that of its territorial expansion and that of its acceptance by the conquered non-Arab peoples from a variety of religious backgrounds.
Muslim tradition is generally more concerned with the former process. When an area is under Muslim rule and subject to Muslim law, that area is regarded as a part of the Muslim world (dar al-Islam), even though the majority of its population may remain non-Muslim. Strictly speaking, only Christians, Jews and Zoroastrians (these last known as majus) were to be allowed to refuse to accept Islam and maintain their existence as separate religious communities under Muslim rule, but in practice toleration was frequently extended more widely.
From this point of view, then, the extensive conquests made under the Umayyads were an extension of Islam. At the beginning of the Umayyad period Arab Muslim rule did not extend much further west than modern Libya or further east than the eastern regions of Iran, and even within these areas many regions must have been held only precariously or merely nominally. By the end of the dynasty all of North Africa and southern and central Spain were included in the boundaries of the Muslim world, and in the east the extension of control into central Asia and northern India prepared the way for later advances in those areas.
In the west the garrison town of Qayrawan was founded about 670 in Ifriqiya (modern Tunisia), and this served as the base for further westward expansion. ‘Uqba b. Nafi‘ is subsequently said to have marched as far as the Atlantic before being killed by the still unsubdued Berbers, but it was not until the end of the century that regions of modern Algeria and Morocco were substantially pacified and the Berbers brought into Islam, but keeping their own language and tribal system. This development is associated with the governorship of Hassan b. Nu‘man in Ifriqiya (683–707). It was Hassan’s successor, Musa b. Nusayr, who initiated the invasion of Spain in 711, sending his Berber client (mawla) Tariq to lead the expedition. It is from this Tariq that Gibraltar takes its name (Jabal Tariq, ‘the hill of Tariq’).
In the east too the years around 700 saw major advances. Al-Hajjaj, governor of the eastern part of the Umayyad territories from 694 to 714, sent his generals Ibn al-Ash‘ath against the ruler of Kabul, Qutayba b. Muslim into the territories lying beyond the river Oxus (Jayhun or Amu Darya in Muslim works), and Muhammad b. al-Qasim into northern India. Qutayba is said to have reached the borders of China and sent an embassy demanding submission from the ‘king of China’. The extent and effectiveness of these expeditions may sometimes be open to question, but it is clear that Arab Muslim control was extended and consolidated in the east under the Umayyads.11
The spread of Islam among the non-Arab peoples of the conquered regions is much less explicitly described in our sources. At the outset of the Umayyad period it is clear that very few of the conquered peoples had accepted Islam, however we understand this last phrase (islam literally means ‘submission’). But by the end of the period, in spite of the initial attempt by the Arabs to keep themselves apart religiously and socially from their subjects, and in spite of the refusal by caliphs and governors to allow the non-Arabs to enjoy the advantages of acceptance of Islam, large numbers of the subject peoples had come to identify themselves as Muslims.
The spread of Islam vertically in this way is clearly a complex process, depending on a variety of factors which were not the same in every area or among every group of the non-Arab population, and resulting in divergent rates of progress. Because of the silence or ambiguity of the sources we are often reduced to speculation about causes and the spread of the process. For example, we know very little about the islamisation of Syria and there are only one or two references in non-Muslim sources which seem to indicate substantial islamisation of the local peoples during the Umayyad period. On the other hand, the Muslim sources have many references to the difficulties caused to Umayyad governors of Iraq and Khurasan when large numbers of non-Arab non-Muslims attempted to accept Islam by becoming mawali in the early decades of the eighth century, but they still leave many questions unanswered or answered at best ambiguously.
So far as the evidence enables us to judge, and leaving aside the Berbers whose society and way of life made them likely allies for the Arabs in the wars of conquest, it seems to have been in lower Iraq, Khurasan and Syria that Islam made the most significant advances among the subjects peoples in the Umayyad period. In western Persia and Egypt, on the other hand, it seems that islamisation in this sense was relatively slow and that it was not until after the dynasty had been overthrown that Islam became the religion of the majority in these areas.12
In spite of our uncertainties, it seems clear that the Umayyad period was crucial for the process of Islamisation in all its forms.
By ‘arabisation’ I mean the spread of a culture characterised above all by its use of the Arabic language in the area which had become subject to Arab Muslim rule. Although associated with the process of islamisation, arabisation is a distinct movement as can be seen from the fact that important communities of Jews and Christians survived in the Islamic Middle East into modern times. These communities maintained their religious traditions in spite of the fact that they had renounced the everyday languages which they had used before the Arab conquest and had adopted Arabic. Conversely, Persia presents a striking example of a region which largely accepted Islam as its religion but maintained its pre-Islamic language at first in everyday and later in literary use, although, of course, the language underwent significant changes in the early Islamic period.
Again one has to take into account that Arabic itself changed as it spread and was elaborated in the process of interaction between Arabs and non-Arabs. Put crudely, as the non-Arab peoples adopted Arabic, so their own linguistic habits and backgrounds affected the language, leading to significant changes and to the formation of different dialects. The result of this evolution is usually described as Middle Arabic as opposed to Classical Arabic, which is identified with the language of the Koran and of the poetry which, it is claimed, originated in pre-Islamic Arabia. The origin and nature of Classical Arabic itself, though, is to some extent a topic of controversy. What led to the adoption or rejection of Arabic by non-Arabic speakers is obviously a very complex question involving consideration of political and social relationships as well as more purely linguistic ones.
In attempting to chart the progress of arabisation the difficulties again arise from the lack of explicit information on the topic in our literary sources and from the paucity of written material surviving from the Umayyad period. For instance, although it has been suggested that Jews of all sorts began to speak Arabic as early as the seventh century, the process of change must have been gradual and our earliest texts written in Judaeo-Arabic (that is, the form of Middle Arabic used by Jews and written in Hebrew rather than Arabic script) come from the ninth century. Our earliest Christian Arabic texts (Arabic written in the Greek script) have been dated to the eighth century, but there has been some argument about the dating. On the other hand, from later developments we know that Persian must have survived as the spoken language of the majority of Iranians during the Umayyad period, but our sources only rarely and ambiguously let us see that it was so, and almost all of our source material on the history of Persia under the Umayyads is in Arabic.
More concrete evidence is provided by the administrative papyri which have survived from Egypt. In spite of the limited range of subjects with which they are concerned, they at least enable us to see a gradual change from Greek to Arabic in the language of the administration. Furthermore, our literary sources report that around 700 it was ordered that henceforth the government administration should use Arabic rather than the languages which had been used before the Arab conquest and which had continued in use thus far. This could indicate that there was at that time a significant number of non-Arabs with sufficient command of Arabic at least for the purposes of administration, since the bureaucracy continued to rely overwhelmingly on non-Arabs. The change of language in the bureaucracy did not happen overnight, and the sources are not unanimous about when it was ordered, but in the development of arabisation it seems to have been a significant step.
Why and how Arabic, and with it the other features which seem to make Islamic culture in the Middle East significantly Arab and distinguish it from others, spread is, therefore, still debatable. Eventually, as we know, the adoption of Arabic for most purposes became general in Syria, Iraq and Egypt while the Berbers and Persians, in spite of their acceptance of Islam and therefore of Arabic as their sacred language, continued to use their own languages for everyday purposes. We can assume that arabisation, like islamisation, progressed a long way under the Umayyads, but precise evidence is hard to come by.13
The second question asked at the beginning of this chapter concerned the way in which the Umayyad dynasty has been regarded by Muslim tradition and how it has been seen in the context of Islamic history generally. Discussion of this question, which involves some consideration of the way in which our Muslim sources for the period came to be formed, is a necessary condition for an understanding of the narrative history which the remainder of this work undertakes.
Even allowing for the qualifications which will be made shortly, there is no doubt that in its broad outlines as well as in its details Muslim tradition is generally hostile to the Umayyads. When the two can be distinguished, Shi‘ite tradition is more hostile than that of the Sunnis, but many of our sources contain material which reflects both Shi‘ite and Sunni points of view so that there is some justification, for our purposes here, in talking about Muslim tradition as a whole.14 The hostility of tradition is reflected in both what the tradition reports and the way in which it reports it.
We are told that before Islam the Umayyad family was prominent in the opposition to Muhammad among the Meccans and that most of the members of the family only accepted Islam at the last moment when it became clear that the Prophet was going to be victorious. Once inside the Muslim community, however, they exploited circumstances, and, by skilful political manipulation not entirely free from trickery, they obtained power, displacing those whose claims to the leadership were based on long service to Islam, piety, and relationship to the Prophet. In power they pursued policies which at best paid no regard to the requirements of Islam and at worst were positively anti-Islamic. Among the charges brought against them, some of the most prominent are that they made the caliphate hereditary within the Umayyad family; that they oppressed and even caused the death of numerous men of religion and of the Prophet’s family, most notably of the Prophet’s grandson Husayn; that they attacked the holy cities of Mecca and Medina, going so far as to bombard Mecca with catapults on two occasions— an image which may well symbolise the conception of the Umayyads in tradition; and that they prevented non-Muslims from accepting Islam and obtaining the rights due to them. They ruled by force and tyranny. Literary works came to be produced devoted to cataloguing the crimes of the Umayyads, singing the praises of their opponents, and explaining why God allowed the community to fall under the sway of these godless tyrants. The best-known of these works are those of Jahiz in the ninth and Maqrizi in the fifteenth centuries.15
Tradition expresses its hostility to the dynasty above all by insisting that they were merely kings and refusing to recognise them, with one exception, as caliphs. The caliphate, according to tradition, emerged in Medina on the death of Muhammad in order to provide a leader for the Muslims in succession to him. The title khalifa is interpreted as meaning ‘successor of the Prophet’, in full khalifat rasul Allah, and the caliph was to be motivated solely by the interests of the Muslims. The Muslim theory of the caliphate took time to evolve and was never static, but two ideas in particular came to be prominent. First, the caliph was to be chosen, from among those with the necessary qualifications, by some sort of election. How this election was to be carried out was never agreed on but the feeling was that the caliph should not simply seize the office by force or be appointed by one man with no consultation of the Muslims. Secondly, the caliph’s authority was to be limited, in particular in the sphere of religion, where the real authorities, the guardians of the Sunna and the heirs of the Prophet, were the religious scholars (the ‘ulama’). In effect, the caliph was simply to maintain the conditions in which the religious scholars could get on with their task. (All this, of course, refers primarily to the Sunni view of the caliphate. The Shi‘ites and Kharijites had different ideas.)16
A sharp distinction is then made between the idea of a caliph and that of a king, between caliphate (khilafa) and kingship (mulk).
Unlike the caliph, the king (malik, pl. muluk) is an arbitrary, worldly ruler whose power depends ultimately on force. The symbolic type of king for Muslim tradition is the Byzantine emperor (Qaysar, i.e.,
‘Caesar’) and the Sasanid shah (Kisra, i.e., ‘Chosroes’, ‘Khusraw’).
When tradition denigrates Umayyad rule as kingship, therefore, it is putting the Umayyads in the same category as all the other kings of this world and contrasting them with its own ideal of Islamic government.
It is not the personal qualities or defects of a ruler which determine primarily whether he is to be accorded the status of caliph or discarded as a king, although the personal piety or wickedness of an individual could affect the question. There were some personally upright Umayyads just as there were corrupt and debauched members of the ‘Abbasid dynasty which took over the caliphate when the Umayyads were overthrown. The latter, however, are all accepted as caliphs by Sunni tradition while the former, with the one exception, are merely kings. Nor does it depend on the self-designation of the dynasty. The Umayyads do not appear to have used the title malik (king) and they did not, at least in the earlier Umayyad period, affect in a very marked way the paraphernalia of kingship such as a crown, throne or sceptre. In contrast to them, the early ‘Abbasid rule was associated much more with the symbols of a traditional oriental despotism.17
In fact it was the Umayyads’ use of the title khalifa which probably played an important part in the tradition’s classification of them as kings. Whereas Muslim tradition regards the title as an abbreviation of khalifat rasul Allah, signifying successor of the Prophet, the Umayyads, as evidenced by coins and inscriptions, used the title khalifat Allah. While it is not completely impossible to reconcile the use of this title with the traditional understanding of khalifa, it does seem likely that the Umayyads’ conception of the title and the office was different. Khalifat Allah (Caliph of God) almost certainly means that they regarded themselves as deputies of God rather than as mere successors to the Prophet, since it is unlikely that khalifa here means successor (one cannot be a successor of God) and elsewhere khalifa is frequently met with in the sense of deputy. In other words, the title implies that the Umayyads regarded themselves as God’s representatives at the head of the community and saw no need to share their religious power with, or delegate it to, the emergent class of religious scholars.18
Above all the charge of kingship is connected with the decision of Mu‘awiya to appoint his own son Yazid as his successor to the caliphate during his own lifetime. This event, more than anything else, seems to be behind the accusation that Mu‘awiya perverted the caliphate into a kingship. The episode will be considered more fully later, but, in the light of the Sunni conception of the nature of the caliphate, what was wrong with Mu‘awiya’s appointment of Yazid was that one man took it upon himself to choose a caliph with no consultation with the representatives of Islam (whoever they might be) and without even a token nod to the idea that the office should be elective. It is probable that such ideas were not generally held, even if they yet existed, in the time of Mu‘awiya. But according to tradition he acted as a king in this matter, introducing the hereditary principle into the caliphate, and the dynasty which he thus founded, and which maintained the general principle that the ruler nominated his successor, was thus a line of kings. Yazid’s personal failings, which are certainly underlined by tradition, merely seem to reinforce the message and are not really the source of opposition to his appointment.19
It should be clear, then, that tradition is generally hostile to the Umayyad dynasty. It is, nevertheless, true that the same Muslim tradition transmits some material which is more ambiguous, sometimes even overtly favourable to the Umayyads. For example, the administrative and political ability of caliphs like Mu‘awiya and ‘Abd al-Malik is admitted, and some of the ‘Abbasids are said to have expressed admiration for this aspect of their predecessors’ work. Even on more strictly religious questions, the tradition sometimes seems less clear-cut than one would expect. The name ‘the year of the (reestablishment of the) community’, which is applied both to the year in which Mu‘awiya received acknowledgment in Kufa after his defeat of ‘Ali and to that in which ‘Abd al-Malik similarly ended the second civil war, recognises the virtues of these two caliphs in rescuing the community from a period of internal dissension. Indeed, one often finds in tradition a fearfulness for the fate of the community under such enemies of the Umayyads as ‘Ali and Ibn al-Zubayr, whatever their personal merits might have been. In legal traditions some Umayyads, notably Marwan, himself caliph for a short time and ancestor of one of the two branches of the Umayyad family to acquire the caliphate, are frequently referred to as makers of legal rulings, and they often come out quite favourably even in comparison with some of the most important of the Prophet’s companions. On occasion a maxim which one tradition ascribes to, say, Marwan will appear elsewhere as a maxim of the Prophet himself. Even the bombardment of Mecca and the consequent damage to the Ka‘ba, which is a key point in the traditional complaints against the dynasty, can be toned down. Among the various reports of these events, some say that the fire which damaged the Ka‘ba while Mecca was being bombarded came about accidentally, and some even say that it was caused by the carelessness of one of the defenders of Mecca, even Ibn al-Zubayr himself being named. Here we are not concerned with the historical accuracy of these reports, merely with the fact that they are transmitted even though the tenor of Muslim tradition is broadly anti-Umayyad.20
Even the treatment of the one Umayyad caliph who is recognised as such in tradition and exempted from the accusation of kingship levelled at the others, ‘Umar b. ‘Abd al-‘Aziz (‘Umar II, 717–20), may be ambiguous. In one way to nominate him as the only caliph in a line of kings serves, of course, to underline the contrast between the pious ‘Umar and the rest of the dynasty, but equally it could be argued that the existence of ‘Umar to some extent rescues the dynasty from complete condemnation. While the traditions about him emphasise the links on his mother’s side with ‘Umar I, the second successor of Muhammad and one of the four Rightly Guided Caliphs, they also do not hide the fact that on his father’s side he was a leading member of the Umayyad family. His father was brother of the caliph ‘Abd al-Malik and governor of Egypt for most of the latter’s caliphate. Evidently, therefore, the Umayyads could produce a genuine caliph and one could conclude that there was nothing inherently bad in the family.21
In order to understand both the generally negative attitude towards the Umayyads in Muslim tradition and the fact that the tradition transmits material which is apparently more favourable to the dynasty, it is necessary to understand the way in which the tradition came to be formed—the way in which our Muslim literary sources originated, were transmitted, collected and finally committed to writing in the form in which we know them.
It seems likely that it was not until the later part of the Umayyad period that traditions, religious or historical (and the distinction is not always clear), came to be committed to writing with any frequency. Before that time they were generally transmitted orally in short, separate reports which were self-contained and relatively easy to memorise. As it became more common to put them in a written form, however, these short reports could be united into more complex units, compiled around a theme or organised in a narrative framework. In the later Umayyad and early ‘Abbasid period, then, scholars such as Abu Mikhnaf (d. 774), Ibn Ishaq (d. 761), or ‘Awana (d. 764) began to compile ‘books’ by collecting the traditions available and organising them around a theme such as the battle of the Camel, the second civil war, or even the history of the caliphate. They may have simply dictated the relevant material to their disciples, which would account for the different versions of works attributed to a particular scholar which have come down to us from different disciples, or they may have put it in writing themselves.
The material thus collected was then transmitted to later generations which treated it in a variety of ways. It might be again broken up and put together with material from different sources in order to make it relate to a different theme; long narratives might be abridged by omitting material considered irrelevant; short narratives might be filled out by interpolation or by linking material together without making it clear where the link occurs or even that it has been made; material might fall out of circulation or it might be reshaped consciously or subconsciously by substitution of words or phrases, by the addition of glosses, or even by formulating entirely new material. It is obvious, therefore, that there was plenty of scope for the material to change in the course of its transmission, and it would be natural that it should change in accordance with changing political, social and religious circumstances. Generally speaking, the material would have been constantly revised to make it relevant and acceptable, and the original significance and context of the material would come to be forgotten.
This process continued for some generations until, in the ninth and tenth centuries, written versions of the material were produced which have survived as our earliest Muslim literary sources, our earliest examples of Muslim historical writing, biography, Koranic exegesis, and so on. In fact, of course, the process continued even beyond the ninth and tenth centuries: even in the fifteenth and sixteenth centuries Muslim scholars were selecting from and reshaping the works of their predecessors, but, when we have the material in both its early and its later form, we can clearly see what has happened to it in the course of transmission. Our problem with our earliest sources for the Umayyad period is that the material prior to the ninth and tenth centuries has been lost and we have to depend on relatively late versions of it transmitted to us by scholars such as Baladhuri (d. 892) and Tabari (d. 923).22
An important point is that a decisive role in the collection, transmission and reduction to writing of the material was played by scholars representative of the opposition to the Umayyads. That is, scholars associated with the Muslim circles hostile to the dynasty, predominantly in Iraq, took a leading role in collecting, arranging and editing the material. If we add to this the fact that the written material which has come down to us was produced in the period after the Umayyads had been overthrown, under the caliphate of the ‘Abbasids who had supplanted them, it is not hard to understand why it has the fundamental hostility to the Umayyads which has been indicated. It is not a question of the ‘Abbasids employing scholars to produce deliberate justifications for ‘Abbasid rule, rather that the scholars involved inherited material from, and were themselves part of the tradition of, Muslim opposition to the Umayyads.
Although we often refer to scholars like Baladhuri and Tabari as historians inasmuch as they were concerned with producing a picture of the past and its relationship to their own times, objectivity, which has been regarded as at least a desideratum of the historian since the nineteenth century, is not to be expected from them. Fundamentally they were religious scholars and it is useful to remember that Tabari, whose Ta’rikh (a mixture of history and chronicle) is one of our fullest sources of information on early Islam and the Umayyad period, wrote a Koranic commentary which is even more voluminous and which, regarding the life of Muhammad, often provides more ‘historical’ information than is available in the Ta’rikh.
If the outlook of these scholars was likely to make them generally hostile to the Umayyads, however, certain things mitigated this hostility and help to explain the more ambiguous material which has been noted. Most importantly, the material collected and transmitted by any individual scholar may be traced ultimately to a wide variety of sources, including even pro-Umayyad sources, and there was no central directory imposing a censorship on the scholars. It used to be thought, following Wellhausen, that the scholars could all be classified as the representatives of one or another ‘school’, that the material associated with the name of a particular scholar would be biased to support the geographical and religious viewpoint of the ‘school’ to which he belonged. So Abu Mikhnaf was regarded as a representative of the Iraqis, Ibn Ishaq of the Medinese, and so on. But it is now recognised that one will find many different shades of opinion represented in the material transmitted under the name of any individual. Even the earliest of them already had an amount of material from which to select, and we cannot point to a particular time or individual as being decisive in the formation of the tradition. Any analysis of the tradition needs to take into account both its final editing and arranging and its earlier transmission.23
Secondly, the scholars were strongly aware of the element of continuity in the history of Islam, and to have been too hostile to the Umayyads, portraying them as non-Muslims for example, would have been incompatible with this sense of continuity. It may be that the traditions about ‘Umar II, linking the Umayyads with the period of Rightly Guided Caliphs, are particularly influenced by this sense of continuity. Those scholars representing the Sunni tendency had a particular problem. If the legitimacy of the Umayyads was questioned too sharply, ammunition might be provided for the Shi‘ites, most of whom came to see ‘Ali as having been cheated not only by Mu‘awiya but also by the first two caliphs, Abu Bakr and ‘Umar, who are of central importance for the Sunni concept of the transmission of the Prophet’s Sunna to the later community. Furthermore, Mu‘awiya himself was a companion of Muhammad, his secretary according to tradition, and one of the characteristics of Sunni Islam is its championing of the companions as sources of authoritative teaching, as against the Shi‘ites who viewed them in general with suspicion and as enemies of ‘Ali and the imams.
Muslim tradition is virtually our only detailed source for the history of the Umayyad state. It should be obvious, therefore, that the nature of the tradition has to be borne in mind constantly when attempting to discuss the history of the period.
For modern treatment of the Umayyads, see Appendix 2.
1. C.H.Becker was one of the first to insist on the distinction between islamisation and arabisation, and he stressed too the crucial importance for the development of Islam as we know it of the interaction between Arabs and conquered peoples outside Arabia in the period after the Arab conquests. See his Islamstudien, i, 66–145, and in English his ‘The expansion of the Saracens’ in the Cambridge Mediaeval History, 1st edition 1911–36, ii, chapters 11 and 12.
2. This understanding of the emergence of the schools of religious scholars and their elaboration of the notion of Sunna depends on the results of J.Schacht’s persuasive, but still controversial, studies of early Muslim jurisprudence. See his Introduction to Islamic law, especially chapters 5 and 6, and the article ‘Fikh’ in EI2; for a more conservative analysis of the concept of Sunna, taking issue with Schacht, M.M. Bravmann, The spiritual background, 179 ff.; R.B.Serjeant in Arabic literature to the end of the Umayyad period, ed. A.F.L.Beeston et al., Cambridge 1983, 142–7.
3. J.Wellhausen, The religio-political factions in early Islam; M.Hodgson, ‘How did the early Shi‘a become sectarian?’, JAOS, 75 (1955); S.Moscati, ‘Per una storia dell’ antica Ši‘a’, RSO, 30 (1955); W.M.Watt, ‘Shi‘ism under the Umayyads’, JRAS, (1960); W.F.Tucker, ‘Bayan b. Sam‘an and the Bayaniyya: Shi‘ite extremists of Umayyad Iraq’, MW, 65 (1975); idem, ‘Rebels and gnostics: al-Mugira ibn Sa‘id and the Mugiriyya’, Arabica, 22 (1975); idem, ‘Abu Mansur al-‘Ijli and the Mansuriyya: a study in medieval terrorism’, Isl., 54 (1977); idem, “Abd Allah b. Mu‘awiya and the Janahiyya: rebels and ideologues of the late Umayyad period’, SI, 51 (1980); S.M. Jafri, The origins and early development of Shi‘a Islam.
4. J.Wellhausen, Factions; W.Thomson, ‘Kharijitism and the Kharijites’, in The MacDonald presentation volume, Princeton and London 1933; W.M.Watt, ‘Kharijite thought in the Umayyad period’, Isl., 36 (1961); articles ‘Azarika’, ‘Ibadiyya’ and ‘Kharijites’ in EI2.
5. I.Goldziher, Muslim Studies, i, 101 ff.; P.Crone, Slaves on horses, 49– 57.
6. See pp. 70–1, 76–81, 85–6, 105–7.
7. See p. 70.
8. I Goldziher, Muslim Studies, ii, 49 ff.; H.Lammens, Mo‘awia 1er, 202 ff.; on the development of the khutba and associated features, article ‘Khutba’ in EI2.
9. Cf. I.Goldziher, Muslim Studies, ii, 48 ff. and S.D.Goitein, ‘The sanctity of Jerusalem and Palestine’ in his Studies in Islamic history and institutions.
10. P.Crone and M.A.Cook, Hagarism, 8, 19–20.
11. C.H.Becker, ‘The expansion of the saracens’; H.A.R.Gibb, The Arab conquests in central Asia; F.McGraw Donner, The early Islamic conquests.
12. D.C.Dennett, Conversion and the poll-tax in early Islam; M.Lapidus, ‘The conversion of Egypt to Islam’, IOS, (1972); M.Brett, ‘The islamisation of North Africa’, Islam and modernisation in North Africa, ed. M.Brett; N.Levtzion (ed.), Conversion to Islam; R.Bulliet, Conversion to Islam in the medieval period.
13. A Poliak. ‘L’arabisation de l’orient semitique’. REI, 12 (1938); M. Sprengling, ‘Persian into Arabic’, AJSL (1939, 1940); J.Blau, The emergence and linguistic background of Judaeo-Arabic; article “Arabiyya’ in EI2; G.Lazard, ‘The rise of the New Persian language’, in R.N.Frye (ed.), The Cambridge History of Iran, iv, London 1975.
14. For Shi‘ite views of the Umayyads, E.Kohlberg, ‘Some Imami Shi‘i interpretations of Umayyad history’, in G.H.A.Juynboll (ed.), Studies on the first century of Islamic society, 145ff.
15. Jahiz, Risala fi Bani Umayya (=Risala fi’l-nabita), French trans. Ch. Pellat, AIEOr. Alger (1952); Maqrizi, Al-Niza‘ wa’l-takhasum fima bayna Bani Umayya wa-Bani Hashim, English trans. C.E.Bosworth, Al-Maqrizi’s ‘Book of contention and strife’.
16. Article ‘Khalifa’ in EI2; for discussion of one of the most important statements of the qualifications, powers and duties of the caliph, see H.A.R.Gibb, ‘Al-Mawardi’s theory of the khilafa’ in his Studies on the civilization of Islam.
17. I.Goldziher, Muslim Studies, ii, 38 ff.; G.E.von Grunebaum, Medieval Islam, 156 ff.; A.Abel, ‘Le Khalife, presence sacrée’, SI, 7 (1957); O.Grabar, ‘Notes sur les ceremonies umayyades’, in Studies in memory of Gaston Wiet, ed. Myriam Rosen-Ayalon, Jerusalem 1977.
18. Cf. I.Goldziher, ‘Du sens propre des expressions Ombre de Dieu, etc.’ RHR, 35 (1897); W.M.Watt, ‘God’s caliph. Quranic interpretations and Umayyad claims’ in Iran and Islam, ed. C.E.Bosworth, Edinburgh 1971.
19. Note that it is Mu‘awiya rather than Yazid who bears the brunt of the charge of corrupting the khalifa to mulk.
20. G.R.Hawting, ‘The Umayyads and the Hijaz’, Proceedings of the fifth seminar for Arabian Studies, London 1972.
21. C.H.Becker, ‘Studien zur Omajjadengeschichte. a) ‘Omar II’, ZA, 15 (1900).
22. Article ‘Ta’rikh’ in EI1 Supplement; P.Crone, Slaves on horses, ‘Historiographical introduction’; A.A.Duri, The rise of historical writing among the Arabs.
23. A Noth, Quellenkritische Studien zu Themen, Formen, und Tendenzen frühislamischen Geschichtsüberlieferung.