The centrality of Islamic civilization

MICHAEL COOK

“Islamic civilization” is not a term calculated to endear itself to anyone with a systematic preference for splitting rather than lumping. For one thing, it commits us to the view that, in spite of a large amount of regional and local diversity, there is an overarching unity to be discerned in the lifestyles of Muslim peoples.

Perhaps the most obvious of these features is the sheer extent of the civilization by the end of our period, around 1500. The emergence of what by neolithic or bronze-age standards were very large human communities is a phenomenon familiar from the last three thousand years of human history in the Old World, but in no earlier case had such a community extended over so vast a territory - from West Africa to Southeast Asia, from East Africa to the bend in the Volga. Only the spread of European civilization across the globe since 1500 has eclipsed the Islamic expansion.

Another unprecedented feature of Islamic civilization is best approached by noting that earlier instances of very large human communities had taken one of two forms. They might involve the emergence of relative cultural homogeneity in particular regions, as with the spread of Hellenism in the eastern Mediterranean and Western Asia, the extension of northern Indian culture to the south, or the similar southward expansion of northern Chinese

I would like to thank Patricia Crone, Anatoly Khazanov, and the editors of this volume for their comments on a draft of this chapter, W.W.

Islamic civilization is also unusual in the relatively late date of its emer­gence. The major civilizations of the Old World typically took shape many centuries before the beginning of our era; Islamic civilization stands apart from its peers by virtue of the fact that it emerged well over a millennium later. It thus appeared at a time when most parts of the Old World with the requisite economic resources had already filled up with civilizations or equivalent cultural formations.

This last feature had significant implications for the interaction between Islamic and other civilizations. For one thing, it meant that the new civiliza­tion often superseded earlier ones. Such supersession was not unpreced­ented, nor was it total - earlier civilizations did not vanish without trace. And of course the expansion of Islamic civilization did not involve superses­sion everywhere. Being also a world religion, and one that originated in a region that was largely desert, Islam was well suited to spread among peoples whose limited economic resources or relative isolation placed them beyond the frontiers of existing civilizations; it did so in several parts of the Old World, such as West Africa and northern Eurasia. But overall, the supersession of older civilizations bulked unusually large in the history of Islamic civilization.

The theme of interaction is also central to the origins of Islamic civiliza­tion. Typically a civilization has its beginnings in a region that possesses the economic resources to support one but is not currently doing so: either it has never done so, as in the case of northern China in the second millennium before our era, or an earlier civilization has succumbed to the hazards of history, as in northwestern India in the same period. The Middle East, the region with the longest history of civilization anywhere on earth, had ceased to contain such regions long before the seventh century. It was not that all of the Middle East was civilized. Geographically speaking, much of Arabia is in effect an eastern extension of the Sahara Desert, and whatever its links to the civilizations that prevailed in the wider region, the constraints of the Arabian environment largely precluded its adoption of them. But if the Middle East was made up on the one hand of a region that already possessed well- established civilizations, and on the other of a region that lacked the resources to support one, where was a new civilization to take shape? There is a puzzle here; at the very least, we can say that the formation of Islamic civilization was not a development that anyone living prior to the seventh century had reason to anticipate.

But for the moment, in fact until near the end of this chapter, let us leave aside the puzzle of the emergence of Islamic civilization. Instead, let us start with a survey of the territories that made up the domain of this civilization - the Islamic world - as it was towards the end of our period. In the process we will also see how these varied regions had come to be part of this world (see Map 15. ι).

The regions of Islamic civilization around 1500:

the new lands

A glance at the map, showing the extent of the Islamic world around 1500, serves to underline two things. The first is the sheer size of the territory involved, a point we have already noted.

The process by which the Islamic world came to be so large can conveni­ently be divided into three unequal phases. The first was the lifetime of

Downloaded from https:Zwww.cambridge.org/core. Vanderbilt University Library, on 10 Jan 2017 at 00:01:56, subject to the Cambridge ( of use, available at https:/www.cambridge.org/core/terms. http://dx.doi.org/10.1017/CBO9780511667480.016

Map 15.i. The Islamic world

Muhammad, the Prophet of Islam, and more particularly the decade from 622 to his death in 632, during which he established and ruled a state in western Arabia that played an increasing role in Arab affairs.

Let us begin our survey with the territories added in the third phase. We can conveniently start in the far north among the Bulghars, and then proceed roughly clockwise.

The far north

The Bulghars were a Turkic people living around the bend in the Volga. Their case is not of major historical significance, but it is instructive.

In the early tenth century the Bulghar ruler was showing interest in adopting Islam and instituting diplomatic relations with the Caliph in Bagh­dad. The Caliph duly dispatched an embassy that included a certain Ibn Fadlan, who left a detailed account of his experiences which has fortunately survived. He found the Bulghar ruler ill-informed about Islam but anxious to learn. For example, he was making mistakes in connection with the ritual of Muslim prayer, including the Friday prayer, the major public ritual of the Muslim polity in which the ruler is mentioned by name and title. Ibn Fadlan accordingly coached him in the proper way to do things. Yet it is clear from the account of the mission that the ruler's openness to Islam was not just a matter of religious conviction. Like the emperor Constantine who had converted to Christianity six centuries earlier, the Bulghar ruler had other things on his mind. In his case a major concern was the Khazar threat. Between his country and the caliphate lay the empire of the Khazars, a rival Turkic people and the lords of a powerful nomadic state of the kind familiar in the history of the Eurasian steppes.

The Bulghar conversion extended beyond the ruler to the people at large. Ibn Fadlan was heartened by the signs he encountered of their willingness to adopt Islam. Yet their adherence was no more unconditional than that of their ruler. Thus Ibn Fadlan, a punctilious and puritanical Muslim, was deeply disturbed by the way in which the Bulghars would bathe in the Volga: they did so in the nude, and worse yet in mixed company. Tellingly, his attempts to put an end to this distressingly un-Islamic practice were a failure.

And yet there is no question that Islam took root among the Bulghars. For later centuries we have nothing as vivid as Ibn Fadlan's account, but we do have unquestionably Muslim tombstones inscribed in the Bulghar language that date from the fourteenth century. What we see in the Bulghar case is thus an example of a people who converted to Islam without being exposed to any military threat from a Muslim source, let alone a Muslim conquest; as the ruler remarked pointedly to Ibn Fadlan, there was no way the Caliph could project military power as far as the lands of the Bulghars. The factor most visibly at work in this conversion was geopolitical calculation, but to this we can add a couple of further factors that are surely relevant. One is the role of the fur trade in ensuring that the Bulghars already possessed some awareness of Islam, despite having no Muslim neighbours; this transit trade was based on the high level of demand in the Muslim world for the fur of animals hunted in the Arctic forests to the north. The other factor is the frequent readiness of pagan populations - and especially their rulers - to convert to a world religion when given the opportunity.

Meanwhile Islam had achieved some success among other elements in the western steppes. One by-product of this was that Hungary, the Christian country located at the far western end of the steppes, had a Muslim minority, though it had disappeared by the end of the fourteenth century; the hostility these Muslims encountered from the church tended to be balanced by the support they received from the rulers of the country, who regarded their Muslims as people who could perform useful services for the state. In cultural terms they seem to have been a rather assimilated minority. Some Hungarian Muslims were zealous enough to travel all the way to Aleppo and Jerusalem for a Muslim education. But these students described their community to a Muslim geographer as dressing like Christians, and even shaving their beards - a practice almost as un-Islamic as mixed bathing.

The thirteenth century, however, saw a major upheaval in the western steppes thanks to the arrival of the pagan Mongol conquerors. The result was a Mongol state, the Golden Horde, that lasted in some fashion into the sixteenth century. The Golden Horde was in fact one of several Mongol states established in the wake of the Mongol conquests, and all of them sooner or later converted to one of the world religions; in the first half of the fourteenth century the Golden Horde chose Islam. It is this conversion that lies behind the adoption of Islam by the Tatars, the Turkic-speaking Muslim populations that replaced the Bulghars around the bend in the Volga and also settled in the Crimea. There was even an isolated Muslim Tatar population far to the northwest, known as the Lithuanian Tatars. In many ways they assimilated to their Christian environment: they had coats of arms, spoke Polish or White Russian, and developed the view that for a thirsty man to down a drink or two is not actually forbidden so long as he does not get drunk. Yet they remained Muslims, and produced a religious literature composed in Polish or White Russian but written in the Arabic script.

The Turks

On his way to visit the Bulghars, Ibn Fadlan passed through the lands of yet another Turkic people, the Oghuz - often known simply as the Turks. Living on the fringes of Muslim Central Asia, these nomads were just beginning a process of Islamization that proved to be of major historical importance. Ibn Fadlan found them in general to be pagans, but with an interest in Islam that could lead to occasional conversion. Thus one Oghuz tribesman embraced Islam, only to renounce it when his fellow-tribesmen refused to accept him as their chief unless he did so. Over the next century Islam must have made major inroads among the Oghuz, for by the time they went on to invade Muslim Iran in the eleventh century they were either already Muslims or quickly became so. Though we have little information about the process, this too was clearly a case in which a pagan population was in no danger of being conquered by Muslims, but was nevertheless aware of Islam. This was perhaps due as much to proximity as to trade; the “silk route” between Iran and China may have played a large part in the eastward spread of Buddhism, Christianity, and Manichaeism, but it does not seem to have done much for Islam. Another difference between the Bulghar and Oghuz conversions is that among the Oghuz no one appears to have played the role of the Bulghar ruler.

Thanks in part to this conversion, the invasion of Iran by the Oghuz - or let us simply say the Turks - was far less traumatic than the arrival of the pagan Mongols two centuries later. Indeed one of the main effects of the Turkish conquest was to boost the military strength of the eastern Islamic world against its non-Muslim neighbours.

Thus to the west of Iran, it was the now Muslim Turks who ended the power of the Christian Byzantine Empire in Anatolia, which by the fifteenth century they had largely transformed from a land of Greek-speaking Chris­tians to one of Turkish-speaking Muslims. Already in the fourteenth century, one of the Turkish dynasties of Anatolia, the Ottomans, had expanded the frontiers of Islam yet further, crossing the Dardanelles and invading the Balkans. The result was to establish a Muslim dominion over southeastern Europe that lasted into the nineteenth century, and at its height extended far into Hungary. But unlike the Turkish invasion of Anatolia, the Ottoman conquest of the Balkans did not lead to the conversion of the population at large; today Albania is the only Balkan nation with a Muslim majority.

A parallel expansion took place to the southeast of Iran. Here the exten­sion of Muslim rule beyond Sindh had been begun by a Turkish dynasty that antedated the Oghuz invasion of Iran, the Ghaznawids. But the arrival of the Oghuz certainly boosted the process. By the end of our period Muslim rule was the norm in northern India, and was rapidly expanding southwards. Yet here, as in the Balkans, Islam was not adopted by a majority of the popula­tion. The exception in this case was the territory that is now Pakistan (in Bengal mass conversion seems to have taken place only after the end of our period). Elsewhere the Muslim population remained a minority, no more than the salt in a dish as one Muslim ruler put it; and after the twelfth century it was rare for Muslim rulers with possessions in India to have a base outside it. The result was that these rulers often had strong pragmatic reasons to accommodate their Hindu subjects, and this sometimes resulted in a degree of cultural and religious syncretism among the elite that we do not find in the Balkans. But the fact that Muslims were a minority wielding political power could also lead to fanatical Muslim intransigence.

Before we leave the Eurasian interior there is one other extension of the Islamic world that we should note: the conversion of the Uighurs of Chinese

Turkestan. North of the Indian Ocean this was the easternmost case of a large Muslim population. Still further to the east, China had a significant Muslim minority from the time of the Mongol conquest; like the kings of Hungary, the Mongol rulers of China found Muslims to be useful people, particularly as tax-collectors, and many arrived from the West in response to this opportunity. Their descendants became Chinese in language and to a large extent in culture, while retaining their Muslim faith; from the seven­teenth century, the Chinese-educated Muslim elite in southern and eastern China was producing a literature in Chinese that sought to reconcile Islam and Confucianism.

The Indian Ocean

Around the Indian Ocean, far more than in the Eurasian interior, the spread of Islam was linked to long-distance trade routes that had already developed centuries before the time of Muhammad. With the rise of Islam we soon find Muslim merchants playing a prominent part in this commercial scene. As early as the eighth century we know that they were sailing all the way to the ports of southern China; in 758 they sacked Canton and then escaped by sea, thereby attracting the notice of the Chinese historians. In 869 a major rebellion of slaves from East Africa broke out in Iraq, thereby securing the attention of the Arabic historians, and providing indirect but powerful evidence of large-scale commercial relations with East Africa, almost cer­tainly conducted by Muslim merchants. East Africans, doubtless slaves, also appear in ninth-century Java - no surprise since they had already reached southern China in the eighth century. Given this pattern of long-distance Indian Ocean trade, it was only a matter of time before a combination of Muslim commercial settlement and local conversion in response to it would lead to the spread of Islam around the rim of the Indian Ocean.

Since those who chronicle historical events give much more attention to the doings of rulers than to those of merchants, we have little direct infor­mation about this maritime spread of Islam. But we know the outcome. To the east, Muslim populations appeared along the coasts of India and Ceylon, and on the Arakanese coast of northern Burma. To the south, a similar ribbon development took place down the coast of East Africa, where it is likely that Muslim settlement and interaction with the native populations had issued in the formation of the lingua franca known as Swahili well before the end of our period.

The spread of Islam around the shores of the Indian Ocean also initiated the Islamization of a large part of island Southeast Asia, together with the Malay peninsula. Here the arrival of Islam was manifestly the result of Indian Ocean commerce, though once established, the new religion often spread further through conquest. InJava especially, the incoming Islamic civilization superseded an older culture of Indian origin, but not without a degree of admixture of the two cultural traditions that went beyond anything we find further west. For example, one seventeenth-century Javanese court adopted a calendric reform whereby the year was now defined as the Muslim lunar year - and yet they continued to count these lunar years according to their traditional Indian era, not the Muslim era one might have expected them to adopt.

What this maritime spread of Islam across the Indian Ocean did not do was to generate large Muslim populations in the continental interiors. This was as true of East Africa as it was of India and Southeast Asia. Only in the desert lands of the Horn of Africa did such a development take place; and in the highlands behind them Ethiopia itself was able to remain Christian, though it only just escaped Muslim conquest in the early sixteenth century.

The African interior

Further west, however, conquest proved effective in spreading Islam up the Nile valley. The seventh-century Arab conquest of Egypt had stopped short of Nubia, where independent Christian kingdoms held out for several centuries more. But a combination of harrying by warlike Arab tribes and full-scale military expeditions launched by the Mamluk rulers of Egypt meant that by the end of our period Islam had spread far to the south; this is the source of the Arab Muslim population that is the core of northern Sudan today. To the west of the Nile valley, Islam was sooner or later adopted by all the tribal peoples of the Sahara, though elements of their pre-Islamic cultures often survived. Thus in the eleventh century, the Berber nomads of the western Sahara still maintained some conspicuously un-Islamic practices (such as marrying more than four wives), though they were Muslim enough to serve as the vehicle of a puritanical Muslim reformer whose followers, the Almoravids, conquered Morocco and Muslim Spain. Even today the Touareg nomads retain the use of a script derived from the Phoenician alphabet as it was used by the Carthaginians.

In the agricultural belt south of the Sahara - the Sahel - our best evidence comes from the lands along the northern course of the River Niger. Archae­ology gives us reason to believe that there were states in this part of West Africa long before the rise of Islam, but it provides no indication that they came under the influence of the civilizations and religions that prevailed to the north of the Sahara. The rise of Islam ended this isolation, a development that must bear some relation to the familiarity of the Arabs with desert conditions. By the ninth century the Sahara had become sufficiently trans­parent for us to learn from an Arabic author the name of an African kingdom in the Sahel: Ghana. A detailed account of Ghana dating from the eleventh century reveals a pagan kingdom with a significant Muslim commercial presence. There was a separate Muslim city a few miles from the pagan capital. Our account makes it clear that the rulers of Ghana regarded the Muslims as useful people, no doubt in part because of the goods they brought from the lands to the north of the Sahara, and in part because they possessed a skill of literacy that had no native counterpart. The king thus made use of Muslims as ministers, one of them having charge of the royal treasury. What is not clear from this account is whether there were any indigenous African converts to Islam in Ghana.

Just how Islam spread to the mass of the native population we do not know, but by the fourteenth century the account of the Moroccan traveller Ibn Battuta describes a Muslim state in this region, that of Mali, that had been in place since the previous century. In some ways he found the inhabitants of Mali commendably zealous in their practice of Islam; for example, they would encourage their children to memorize the Quran by putting them in chains till they did so. But in other ways he was as shocked as Ibn Fadlan among the Bulghars; for example, he noted court rituals that struck him as absurd, and described how young women would walk around naked. This latter practice was still current in the late fifteenth century, when a pious ruler of Mali was sufficiently perturbed by it to consult a Moroccan religious scholar on the subject. Since then things have changed in the region, though the prominence of women in the marketplace continues to distinguish West African societies from those of the Muslim heartlands.

The Atlantic

The original Arab conquest had brought Islamic civilization to the western coasts of Morocco and the Iberian peninsula. But here, in contrast to the Indian Ocean, there was to be no maritime spread of Islam. Even the Canary Islands, not far off the coast of Morocco, remained outside the Islamic world, eventually to be colonized and Christianized by the Spanish in the fifteenth century. The Arab conqueror of Morocco in the later seventh century is reputed to have ridden his horse out into the Atlantic, telling God that only the sea prevented him from taking the fight further. In the very different conditions prevailing in 1600, a Moroccan ruler wanted to set up a joint venture with the English to acquire territory in the Americas - a plausible alliance, since Spain, the major power in the New World, was their common enemy. But nothing came of this initiative. Even in the Old World, there was no maritime spread of Islam down the west coast of Africa to mirror what was happening in East Africa; still less was there a coastal spread of this kind in western Europe north of Muslim Spain.

Indeed, the Christian lands of western Europe as a whole were remarkable for the absence of a Muslim commercial diaspora; instead it was the Euro­peans - mainly the Italians - who established trading outposts along the Muslim coastlands of the Mediterranean. This fits with the fact that while three of the five larger Mediterranean islands came under Muslim rule at one time or another - including Sicily from the ninth to the eleventh century - only Cyprus retains a significant Muslim population at the present day, one dating from the Ottoman conquest of the sixteenth century.

The regions of Islamic civilization around 1500: the old lands

The five coins shown in Figure 15.ι are at first sight hard to tell apart. All five share the same aniconic design and are inscribed in the same Arabic script; indeed the inscriptions are largely identical. Minor features of design apart, they differ only in two particulars, contained in one of the inscriptions: the name of the mint and the date of minting. From these we learn that the first coin was minted in Muslim Spain in 725-6, the second in Tunisia in 729-30, the third in Syria in 738-9, the fourth in Iraq in 739-40, and the fifth in what is now northern Afghanistan in 733-4. Thus all were minted within a 15-year period, but in places spread out over a distance of some four thousand miles. The coordination that is evident here could only be the work of a state imposing a uniform monetary system over the whole expanse of its territory, and the state in question was in fact the Umayyad Caliphate (661-750), the first hereditary dynasty of Islamic history.

Remarkable as this uniformity was, it was superimposed on tremendous diversity. In the first wave of expansion, in the 630s and 640s, the Muslim state had conquered the entire Persian Empire and the southern provinces of the Byzantine Empire. In later decades the conquests were extended west­wards across North Africa to Spain, passing beyond the limits of former Byzantine rule; at the same time they were extended eastwards to Transox- ania and Sindh, moving beyond the limits of former Persian rule. The Muslim state that minted the five coins was thus ruling over an unpreced­ented assortment of territories.

Figure 15.ι Islamic coins, obverse and reverse (photographs by Alan Stahl)

Figure 15.ι (cont.)

The former Persian Empire

The Persian Empire under the rule of the Sasanian dynasty (224-651) repre­sented a geopolitical synergy of some antiquity: it combined the military manpower of the Iranian plateau with the fiscal resources of the lowlands of Mesopotamia - or to use the Arabic term, Iraq. This pattern first appeared with the Achaemenid dynasty in the sixth century before our era, and was revived for a century or so in Islamic times by the early rulers of the ‘ Abbasid dynasty (749/50-1258); it worked well as long as Iraq retained its agricultural wealth. One result was that the Persian Empire was inevitably a cultural hybrid. There were the traditions of the plateau, which at least at an elite level enshrined a proud Iranian identity; this identity was associated with languages such as Persian and Parthian, with memories of empire projected back onto mythical rulers antedating even the Achaemenids, and with the Zoroastrian religion and its sacred text, the Avesta. The realities at ground level, to the extent that we know about them, were rather different and more diverse. Alongside these traditions of the plateau were those of Mesopota­mia, where there was no shared sense of an ancient political identity, but rather a variety of religious traditions each associated with a literature written in a different dialect of Aramaic. The culture of the Mesopotamian lowlands was thus more fragmented, though also more cosmopolitan and sophisticated, than that of the Iranian plateau.

That the disorganized tribes of Arabia should conquer the Persian Empire would have struck any political analyst of the sixth century as highly improbable - as improbable as the formation of a new civilization. But the first wave of Arab conquest coincided with a low point in Persian political and military fortunes: the Persians had ended up the losers in a prolonged war with the Byzantines, and while the victors made no attempt to annex the defeated empire, they left it politically unstable and devoid of effective leadership. Moreover the Persian capital was located at Ctesiphon in the Mesopotamian lowlands, territory similar to what the Arabs were accus­tomed to in Arabia; once they had taken the capital city, the chances of coordinated resistance on the plateau were poor. The last Persian “King of Kings,” Yazdagird III, held out till his death in 651, but thereafter the Arabs enjoyed more or less undisputed possession of the former Sasanian territor­ies. Their most formidable enemies in this quarter were now the Khazar state to the north of the Caucasus and a Turkic empire in Central Asia (ruled by a pagan Turkic people, nomads like the Khazars, but distinct from the Oghuz whom we met above).

What then became of the former Sasanian territories under the new Islamic dispensation? In one respect Iraq and Iran were no different: in the long run - perhaps by the year 1000 or so - the great majority of the population of both regions converted to Islam, predominantly in its Sunni form (it is only in recent centuries that the majority of the population in Iraq and Iran has shifted to Shi'ite Islam). In another respect the two regions fared very differently. In Iraq, over roughly the same period, conversion was accompanied by a linguistic change whereby Arabic replaced Aramaic, even among the non-Muslim population; whereas in Iran Persian held out as the spoken language, and spread at the expense of other Iranian languages. This disparity was matched by a larger cultural contrast. The Muslims of Iraq knew very little about the pre-Islamic past of their land, and had no sense that this past was part of their heritage; for them, the coming of the new religion and the formation of the new civilization had more or less wiped the slate clean. The Muslims of Iran, by contrast, did not just continue to use Persian as their spoken language; from the ninth or tenth century onwards they also wrote it, albeit in the Arabic script and with many Arabic loanwords (a precedent later followed by the Turks and others). The result was the emergence of an elaborate literature in Persian. Some of this literature invoked the memory of the pre-Islamic past; for the last thousand years Persian-speakers have been in proud possession of an epic account of the pre- Islamic history of Iran, the “Book of Kings” (Shahnama), composed by the poet Firdawsi. Yet Muslim Iran was still very much part of the wider Islamic civilization; indeed many of the most famous works of Arabic literature, history, scholarship, and science were written by men whose vernacular was Persian or some other Iranian language, and the idea of using a Persian translation of the Quran in a ritual context never made much headway. Iran nonetheless maintained a cultural continuity with a pre-Islamic past that it persisted in regarding as its own.

It is not hard to suggest a couple of reasons for the differing outcomes of the Arab conquest in Iraq and Iran. One is the contrast between the cultures of the two regions, with Iran conspicuously high on identity and Iraq notably low on it. The other is the fact that the Arabs came from Arabia: Iraq was a region at a similar elevation and with similar climatic conditions, a land in which not just Arabs but whole Arab tribes could readily make themselves at home. The Iranian highland with its frigid winters was a very different environment, unlikely to attract large-scale Arab settlement, which in any case was militarily unnecessary given the rapid collapse of resistance. The strongest Arab presence in Iran was in the northeast, near the Turkish front, and even these Arabs had a marked tendency to become Persian-speaking - though scattered pockets of Arabic-speakers survived into modern times.

The former Byzantine provinces

Unlike the Persian Empire, the Byzantine Empire was fortunate in having its capital city located far away from Arabia. To reach Constantinople the Arabs had either to cross the Anatolian plateau - a territory as geographically alien to them as the Iranian plateau - or to develop sufficient naval power to reach the city by sea. They did both, though only twice, and each time the formidable defenses of Constantinople saved the day for the Byzantines. The result was that while Anatolia became a war-zone, consolidated con­quests on this front were limited. The Byzantine Empire thus survived the Arab onslaught in a reduced form. But if the Arabs found it hard to push north, they had little difficulty expanding to the west. Egypt was taken in the first wave of conquest, which also saw a venture into Tunisia; the definitive conquest of Tunisia came in the later seventh century, at a time when the Arabs were extending their reach in North Africa further to the west than Byzantine rule had ever extended.

What this meant was that the Arabs had annexed a set of Byzantine provinces, entities more like Iraq than Iran, with no counterparts of the features that enabled Iran to retain its distinctiveness. Beyond that, each case was different. In cultural terms Syria was a western Iraq: its population spoke Aramaic, and was divided into a variety of religious groups that might base their written languages on different Aramaic dialects. Here too, both Islam and Arabic prevailed among the majority of the population by sometime around the year 1000. Egypt was set apart by geography (the Nile valley was insulated on both sides by desert) and language (the population spoke Coptic, a late form of ancient Egyptian), but the outcome was essentially the same. In Tunisia too, conversion to Islam was overwhelming, and Arabic equally successful; after the twelfth century we hear little of the indigenous Christians, and today they have entirely disappeared.

The second wave of conquests

Finally we come to the lands that became part of the Islamic world during the second wave of conquests in the early eighth century. Here, as already indicated, the Arabs were moving beyond the frontiers of the former Persian and Byzantine Empires, though into territories that had historical and cul­tural links with one or other of them.

Transoxania in the early Islamic period was inhabited by people speaking Sogdian, an Iranian language very different from Persian; they were divided among a number of small principalities, and adhered to a variety of religions - Zoroastrianism, Buddhism, Christianity, and Manichaeism. The Arab con­quest led to the spread of an Arabic elite culture, but also, more paradoxic­ally, of Persian as a spoken and eventually written language. In addition it brought about some contact with China, where remnants of the Sasanian elite had taken shelter after the Arab conquest of Iran; but there was no significant interaction with Chinese culture.

Where there was interaction with an alien civilization was in Sindh. Though this northwestern corner of India had experienced Persian rule, it was culturally Indian, and its conquest marked the earliest phase of conflict and accommodation between Islam and Hinduism. However the Arabic historiographical tradition, while deeply interested in the conquest of Trans- oxania, regarded that of Sindh as a side-show, and the interaction with Hinduism seems to have remained of largely local interest. It is not till the time of the Ghaznawids that we get a real sense of the clash of civilizations on this front; and by then the Muslim conquerors were Turks rather than Arabs. It was in this Ghaznawid context that the polymath Biruni visited India, and learnt Sanskrit the better to understand its elite culture.

On the western front, when the Arab conquerors went beyond Egypt they were entering territory largely inhabited by the Berbers. Like the pre-Islamic Arabs, the Berbers were a tribal people with limited tolerance for centralized government; west of Tunisia they put up a fierce resistance to Arab con­quest, and frequently rebelled thereafter. But they seem to have found Islam itself attractive. Many Berbers, eventually all of them, converted to Islam, though at first often to heretical forms of the religion that were at odds with the Umayyad and ‘Abbasid Caliphs. In the early centuries we also hear of new religions appearing among the Berbers that mimicked Islam. Thus until the eleventh century the Berber tribesmen of the Atlantic plains of Morocco followed a faith brought to them by a Berber prophet; they had a scripture like the Quran, but composed in Berber, and celebrated a weekly communal prayer, but on Thursday rather than Friday. The one thing we do not hear about is any long-term survival of Berber pagans or Christians. On the other hand, their conversion to Islam, like that of the Iranians, does not seem to have been accompanied by the adoption of Arabic. But in the long run their fate diverged considerably from that of the Iranians. In the first place, they had no memory of a glorious past when the Berbers had ruled the world; there was no Berber Firdawsi. In the second place, the North African lowlands, though not the mountains, were the kind of territory in which Arab tribes felt at home; by the end of our period the westward and eventually southward spread of these tribes had laid the foundations for the present-day status of Algeria, Morocco, and Mauritania as Arab countries, albeit with significant Berber minorities in Algeria and Morocco.

The original Arab conquest, however, did not continue to the south of Morocco. It did continue into Spain, despite the barrier posed by the Straits of Gibraltar. Here the Arabs - and their Berber converts - encountered a solidly Christian population made up of Romance-speakers ruled by a Visigothic monarchy and aristocracy. The conquerors seized some two thirds of the country and put an end to Visigothic domination; but Christian political power survived in the north, and eventually played a large part in the Christian reconquest that was completed shortly before the end of our period. By then Muslims in Spain were a persecuted minority, soon to be forcibly converted to Christianity and finally, in 1609-14, expelled from the country. In the late seventeenth century a Moroccan ambassador to Spain met people who still remembered nostalgically that their ancestors had been Muslims; but for all their interest in hearing about Islam, they were no longer Muslims themselves.

All these regions were far from the centre of the Islamic world - which is why they were conquered only in the second wave. But they differed considerably in how they related to the the central lands. Transoxania, with northeastern Iran, was effectively part of the metropolitan cultural scene in pre-Mongol times. Sindh was not. Nor was North Africa west of Tunisia for most of our period, though by the fifteenth century Morocco had produced a Sufi writer who would be read as far east as Southeast Asia. Muslim Spain was known mainly for the care with which it imitated the culture of the central Islamic world; a tenth-century Iranian vizier who procured a copy of a literary anthology composed in Spain was disap­pointed to see that the book was “just our own goods they’re sending back to us.” But by the eleventh century some Spanish scholars were repack­aging eastern goods so well that their books became standard works in the central Islamic lands.

The Arabian homeland

Our survey of the Islamic world is now complete with the exception of Arabia, where it all began. The Arabs of pre-Islamic times were not in fact confined to the peninsula; Arab nomads were already present in the Syrian desert, the Sinai peninsula, and the eastern desert of Egypt, thus foreshadow­ing the Islamic expansion. But Arabia was their homeland, and it was here that Islam had its origins.

The fundamental feature of the Arabian environment was aridity. There was some rainfall in the southwestern corner of the peninsula, and to a lesser extent in the southeast, in each case thanks to the presence of mountains; but elsewhere the water supply, such as it was, mostly took the form of oases. Nowhere in Arabia was there a river like the Tigris, the Euphrates, or the Nile. This aridity imposed a low ceiling on the economic, social, and political life of the region. The population was overwhelmingly tribal and to a large extent nomadic. Commercially it had little to offer the outside world in the period that concerns us: the incense trade was over before the rise of Islam, the Yemeni coffee trade did not begin till after our period was over, and Saudi oil was of course far in the future. It is therefore unsurprising that the tribes of Arabia lacked the steep social stratification common among the tribes of the steppes, where nobles and commoners were distinct social groups. Such states as there were in Arabia tended to be located on the fringes of the peninsula, where the supply of rainfall was greater (as in Yemen and Oman) or imperial subsidies more generous (as on the borders of the Persian and Byzantine Empires).

Within this generally arid environment, the interior of western Arabia with its scattered oases - the land known as the Hijaz - hardly seemed particularly promising. But it was here that Muhammad proclaimed his religious message - a monotheism that unlike Judaism or Christianity was customized to resonate with Arab identity. Moreover, he went on to use his message to found a state that extended its influence over Arabia, and after his death far beyond it. Yet once the power of the Muslim state was firmly established outside Arabia, the brief exposure of the peninsula to the geopol­itical limelight came to an end. The center of power shifted to the richer lands of the Fertile Crescent - Iraq and Syria - in the civil war of the later 650s, and it never returned to the arid soil of Arabia.

The impact of Islam on the peninsula was nonetheless profound. The Arabian population converted early and easily to Islam; it was their religion, in a way in which it was not for the conquered peoples outside Arabia. Yemen and Oman continued to matter owing to their combination of relatively greater agricultural resources and commercial frontage on the Indian Ocean; merchants from both regions were active in the trade with East Africa. Despite its lack of productive resources, the Hijaz was now able to join these relatively favored regions thanks to its possession of two of Islam's holiest sanctuaries: Mecca where Muhammad was born, and Medina where he built his state and was buried. This assured the Hijaz a flow of resources from the economically more privileged regions of the Muslim world. Elsewhere Islam showed its unusual ability to generate and maintain networks of religious scholars, even in desert environments where the full panoply of Islamic civilization was unsustainable. This made a real difference in Arabia, as it also did in such parts of the Sahara as Mauritania. But it was not till the coming of oil that Arabia was once again to achieve geopolitical prominence.

The components of Islamic civilization

To speak of Islamic civilization is to say that Islam is more than just a part of a larger package; rather, it is the core around which the civilization is built. That we do not readily speak in the same way of Christian or Buddhist civilization implies that Islam is unique in performing this role. This is a large claim, but a plausible one.

By way of illustration, consider the coins we looked at earlier. They have to be seen against the background of the standard model of coinage that had previously predominated in the Old World west of China: a round, flat piece of metal with the ruler's head on one side and a religious symbol on the other. Our Islamic coins diverge sharply from the standard model in two obvious ways. The first is that they are aniconic, reflecting the Islamic prohibition of images of living beings. The second concerns the substance of the inscriptions. With the exception of the one that states where and when the coin was minted, they consist entirely of Quranic material selected to proclaim God's message to mankind, with no ruler or governor so much as mentioned. They are still round, flat, metallic objects, and immediately recognizable as coins; but their content has been radically Islamized.

Our assignment now is to glance at some other aspects of Islamic civiliza­tion in a similar perspective. For convenience, let us look first at institutions and then at culture.

Institutions

What are the central institutions of Islamic civilization, and how do they relate to Islam? Let us start with what the religion prescribes. In the first place, it requires an Islamic political order: the Muslim community is to be under the authority of a single Muslim ruler, the Caliph - a title understood to refer to his role as either deputy of God or successor of the Prophet. Eligibility for this office is governed by certain rules, for example that the Caliph should belong to Quraysh, the Arab tribe into which Muhammad was born; at the same time there are procedural rules about how the Caliph is to be chosen - there is nothing like a rule of dynastic succession in Sunni Islam. The duties of the Caliph are both domestic and foreign. On the domestic side, he is to see to the performance of the basic public rituals of Islam, notably the Friday prayer; this involves building and maintaining suitable mosques in urban centers. At the same time he is to establish and maintain an Islamic legal order; this entails appointing judges whose learning and integ­rity qualifies them to give judgment according to Islamic law. But unlike most rulers, he does not have the power to make new law: God is the only legislator. On the foreign side, the Caliph's primary responsibility is military. He has to defend the frontiers of the Muslim world against any invasion by unbelievers from outside, and to attack these unbelievers from time to time with a view to extending the frontiers of Islam. Alongside this political order there must also be an Islamic social order, that is to say a Muslim society conducting its affairs in accordance with Islamic law. Here two characteristic features are worth mentioning. One is that there must be a group of people learned in Islamic matters, the scholars. The other feature is that women are subject to certain norms of subordination and segregation that limit their interaction with males outside the family and restrict their role in public space. There is much we could add to extend and refine this simple account, but for our purposes this outline suffices.

All this, of course, is prescription, so we should now ask how far it corresponds to historical reality even in the central lands of Islam. For simplicity, we can think in terms of three possibilities.

One possibility was that historical practice would more or less ignore Islamic prescriptions, leaving the scholars to save the appearances as best they could. A case in point is the principle that the Muslim community should form a single state ruled by the Caliph. This was still a realistic requirement in the first decades after Muhammad's death, and one message of our five coins is that as late as the first half of the eighth century this unity was still intact. But an empire stretching from Spain to Afghanistan was in geopolitical terms unsustainable: such a configuration had never existed before, and was never to be reconstituted. Already in the second half of the eighth century the Islamic world was beginning to break up into a number of regional states, and by the tenth century the political disinte­gration was complete; there were still contacts between Muslims from distant parts of the Islamic world, especially in the context of the pilgrimage to Mecca, but they now paid their taxes to many different rulers. What then was to become of the office of Caliph in the absence of a unitary state? One could have a shadow caliphate, a solution best represented by Egypt towards the end of our period; here an ‘ Abbasid Caliph went through the motions of reigning while a Turkish or Circassian sultan actually ruled. Or one could bestow the caliphate on all Muslim rulers, or on none. Or the ruler of the largest and most powerful state in the Muslim world might claim the caliphate, and receive some degree of recognition outside his own lands; from the sixteenth century this came to be the case with the Ottoman sultan, despite the awkward fact that he did not descend from the tribe of Quraysh.

Meanwhile the prime concern of the scholars was to construe the political realities of the day in such a way as to avoid the distressing conclusion that a legitimate Islamic order no longer existed; they did not usually take the drastic step of seeking to change the realities on the ground in order to bring them into line with the principles enshrined in the religious tradition.

Another possibility was that Islamic institutions could exist side-by-side with non-Islamic ones. An obvious case of this was the administration of justice. Rulers did indeed appoint judges, and the judges did indeed apply Islamic law in one or other of its recognized forms. But while this was part of the story, it was not the whole story. A large part of the population, especially those living outside the major cities, conducted their legal affairs in accordance with whatever system of customary law was accepted locally. The difference mattered: for example, in cases of inheritance Islamic law assigned shares to women, whereas customary law regularly excluded them. At the same time rulers might infringe on God's legislative monopoly by making laws of their own; for a long time the Ottoman sultans did this, producing an elaborate body of law that was known as Qanun (an Arabized form of the same Greek word that gives us the English “canon”). This too could make a substantive difference: for example, Qanun frequently imposed fines as punishments, something Islamic law does not do. We can easily find parallels to this situation in other aspects of Islamic civilization. One clear-cut example concerns the calendar. The formal calendar of Muslim societies was the Islamic calendar; its unique feature was that it defined a year as consisting invariably of twelve lunar months, thus making it some ten days shorter than the solar year, with the result that it was not in phase with the seasons. But in real life the Islamic lunar calendar regularly coexisted with some solar calendar, typically one inherited from the pre-Islamic past of the region in question. Another example takes us back to the prohibition of images. Though the law made no such distinction, the tendency was for the prohib­ition to be observed in public contexts and disregarded in private ones. But there was no hard and fast boundary: in the period following the Turkish invasion of the Middle East, there was a remarkable efflorescence of coins with images on them.

The third possibility was that Islamic principles might be more-or-less honored in practice. For example, most Islamic societies really did contain groups of Islamic scholars learned in the religious tradition.

Two points are worth adding to this reality check. The first is that the Islamic prescriptions often left much undefined. For example, they required rulers to build and maintain mosques in which the Friday prayer could be performed. But they did not tell them what these mosques should look like, for example by prescribing that all mosques should be built on the model of Muhammad's mosque in Medina. As a result, no serious tensions arose over the fact that different regions of the Islamic world developed very different traditions of religious architecture (no one would mistake an Ottoman mosque for a traditional North African one). Likewise there had to be scholars, but the law made no institutional provision for their education. In the early Islamic period they received it by joining informal study circles around individual teachers; from the tenth century onwards, this pattern tended to give ground to a new religious institution, the madrasa or college, which was endowed by the rich and powerful, paid salaries to its teachers, and provided support for its students. The madrasa thus became a wide­spread feature of Islamic civilization, but it was not the only way to arrange things. In the same way Islamic norms did not tell Muslim rulers how their armies should be organized, nor did they say much about how Muslim societies should be stratified.

The second point relates to the way in which the non-Islamic elements we have noted presented themselves. It was one thing for a non-Islamic practice to coexist quietly with Islamic norms, but another for such a practice to flaunt itself in some flagrantly un-Islamic fashion. For example, rulers who were not Caliphs, or claimed the caliphate merely as an add-on, might go by a variety of titles; but if they had the temerity to include the title “King of Kings” (the historic title of the pre-Islamic rulers of Iran) in the ritual of the Friday prayer, this could put them on a collision course with pious Muslims. Likewise if rulers made law and called it by the rather colourless term “Qanun,” this was less likely to raise objections than if they called it “Yasa,” a Mongolian word closely associated with the pagan Mongol con­queror Chinggis Khan. In the same way, Muslim societies were regularly seen as divided into the elite and the masses, but any assertion of formal caste divisions was open to challenge as un-Islamic.

Culture

Let us now shift our attention from institutions to culture, with particular reference to the high culture of the pre-modern Islamic world. We can best begin with the tradition of religious learning created and transmitted by the scholars, though the fact that our surviving sources were written mainly by these scholars means that their activities are likely to loom larger in hindsight than they did in real life (indeed the scholars themselves were given to complaining that no one paid any attention to what they said). The starting point of this tradition was of course divine revelation: the precisely bounded corpus of God's speech, the Quran, and the less well-defined corpus of authentic reports of Muhammad's inspired sayings and doings. As might be expected, the meaning of these texts, and their implications for the lives of subsequent Muslims, were not always transparent, leading to elaborate commentarial genres that proceeded verse by verse, report by report. But the scholars were not confined to following the thread of the revealed texts. They also composed systematic works that were arranged thematically. Thus they wrote treatises on substantive law, treatises on the theory of law, treatises on dogma, and so forth, while the mystics among them wrote works on Sufism. The scholars also wrote abundantly about them­selves: collections of biographies of scholars are a stock-in-trade of Islamic scholarship.

At the same time, the scholars wrote on fields of knowledge less intrinsic to the religious tradition. For example, they collected and commented on poetry, including that composed in pre-Islamic Arabia, and they wrote poetry themselves. Such pursuits could be denounced by the pious as frivolous, but they could also be defended in terms of the benefits they offered to properly religious scholarship: a pagan poet's use of a word could illuminate its meaning in a Quranic verse, since the scripture was revealed in the language of the Arabs of Muhammad's day. What is clear is that whatever validity this rationale might possess, those who wrote and read such works liked poetry for its own sake. Historical writing was a somewhat analogous case. Writing the history of Muhammad and his immediate, authoritative successors was a contribution to properly religious scholarship; but the religious justification for continuing the story down to one's own day was more tenuous, and the bottom line is that people in Muslim societies found history interesting to read about. The case for belles-lettres was weaker still, but a man who was ignorant of the genre could not pass for educated in polite society. By no means all those who wrote poetry, history, or belles-lettres were at the same time scholars of the religious tradition, but many were.

Rather different was the case of scholarly genres that manifestly originated outside the Arabic and Islamic tradition. Such was the case with philosophy, by which is meant in this context the philosophical tradition of ancient Greece, translated into Arabic by Christians for the ninth-century Muslim elite. This tradition came to be an established component of Islamic civiliza­tion, but its alien origin was not forgotten and could always be held against it. Something similar can be said of the sciences, which came to the Muslims largely from Greek sources, but also from Middle Persian and Sanskrit. One example is the Greek medical tradition identified with Hippocrates and Galen. This became the standard medical tradition of Islamic civilization, but like philosophy it was open to challenge for its alien origins. Some pietists thus preferred a rival tradition, the remedies reported to have been pre­scribed by Muhammad, and there was contention about this; one might hold such remedies to be a part of God's revelation, or one might dismiss them as mere folk medicine that Muhammad had picked up from his environment. Indirect support for such dismissal was available in a report describing how Muhammad once gave advice regarding the cultivation of date-palms. The cultivators duly followed his advice, with disastrous results; he then told them that anything he said on such matters was just his own human opinion, and carried no stamp of divine approval. Astronomy was another science that could be contentious. The astronomers believed the earth to be spherical, and had no regard for those who considered it flat; but their view was in awkward tension with the literal sense of revealed texts. Moreover astron­omers often doubled as astrologers, protagonists of a set of beliefs about the influence of the heavenly bodies on earthly events that was horribly un- Islamic, though very widespread in Muslim, as in other, societies.

Islamic civilization as we find it in the core lands of Islam was thus a tense but rather stable mixture of fully Islamic elements, quasi-Islamic elements, and transparently un-Islamic elements. Outside the core, it was some version of this mix that was spread by the later expansion of Islam. The outcome in any given region depended on a number of factors: the precise version of Islamic civilization in play, the force with which it arrived, and the strength of local traditions antedating its arrival. An obvious example of the importance of version is the fact that the new lands dependent on Iran received Persian literary culture as part of the package; this applies to Anatolia, Central Asia, and India, but not to Southeast Asia or Africa. As to the role of force, the limited part played by conquest in Southeast Asia and West Africa is surely related to the persistence in both those regions of Muslim societies retaining systems of matrilineal inheritance from the pre-Islamic past. Finally, a clear example of the part played by the strength of local traditions is the admixture of Islamic and indigenous elite culture that we find in some measure in India, and to a much greater extent in Java and among the Muslim minority in southern and eastern China; significantly, both Java and China were lands in which Muslims not only used the indigenous languages, but wrote them in the indigenous scripts.

Against such diversity we can set a striking linguistic unity in one key respect. Religions that spread widely have to decide whether to confer on translations of their foundational texts the same ritual standing as the origin­als possess. Both Buddhism and Christianity did this at some points in their pre-modern history but not at others. Islam, with one still-born exception, did not do it at all.

The origins of Islamic civilization

It would be easy to imagine a counterfactual history in which the outcome of the Arab expansion was the absorption of the conquerors by the peoples they had conquered; such an outcome would have been favored by the disparity of numbers, and has many historical parallels. But instead it was the con­quering Arabs who were the assimilators, ultimately spreading their lan­guage from Mauritania to Iraq, and their religion over an even wider area. Nor was this all: as we have seen, in the wake of the Arab conquests a new civilization took shape. The paradox is that the key role in this drama should have been played by conquerors coming from the least civilized part of the Middle East. How do we resolve this paradox?

Part of the answer must have to do with the character of the peoples the Arabs conquered in the Fertile Crescent. It was here that Arab settlement was densest, and here that the new Arab rulers had their capital cities, notably Damascus under the Umayyads and Baghdad under the ‘ Abbasids. So it stands to reason that this was the crucial region for the formation of Islamic civilization. As we have seen, the native inhabitants of Iraq and Syria were speakers of varieties of Aramaic, which like Arabic was a Semitic language. Many of them readily joined the conquerors and adopted Arabic. Culturally they had much to bring with them: they possessed elaborate literary traditions of a kind that Arabia lacked. But a millennium of foreign rule had long ago eroded any strong identity of the Iranian type, at once ethnic, religious, and political.

The other half of the answer must relate to the character of the Arab conquerors. In contrast to the peoples of the Fertile Crescent, they were low on elite culture but amply endowed with just the kind of identity that the Aramaic-speakers lacked. Like the Iranians, the Arabs were now in possession of an identity that was at once ethnic, religious, and political. The ethnic component was an old sense of Arab identity, the religious component was the Islamic customization of monotheism that highlighted this identity, and the political component was participation in the immensely rewarding enter­prise of extending the sway of the Islamic state far beyond the borders of Arabia.

There was thus a certain complementarity between the conquered and the conquerors. The Arabs brought such things as their language, their traditions of poetry and public speaking, their consuming interest in genealogy, and their as yet uncanonized scripture. But the process that turned all this into a comprehensive literary culture surely owed more to the non-Arab converts of the Fertile Crescent. And yet it has to be emphasized that any account of the way in which Islamic civilization emerged is likely to remain in large measure speculative. In the end we cannot say much more than that this civilization was not yet there at the beginning of the seventh century, and had crystallized by the end of the eighth. Nor does it help that we are dealing with an event to which we have no historical parallel. In short, the dramatic changes that issued in the formation of the new civilization can fairly be seen as one of the great black-swan events of history: utterly unlikely in prospect, and immensely consequential in retrospect.

FURTHER READING

1. General historical surveys and reference

Encyclopaedia Iranica, ed. Ehsan Yarshater. London: Routledge & Kegan Paul, 1985-. The Encyclopaedia of Islam, second edn. Leiden: Brill, 1960-2009.

Humphreys, R. Stephen. Islamic History: A Framework for Inquiry, revised edn. Princeton University Press, 1991.

Kennedy, Hugh. The Prophet and the Age of the Caliphates: The Islamic Near East from the Sixth to the Eleventh Century, second edn. Harlow: Pearson Longman, 2004.

Lapidus, Ira M. A History of Islamic Societies, second edn. Cambridge University Press, 2002.

The New Cambridge History of Islam, 6 vols. Cambridge University Press, 2010, vol. I : The Formation of the Islamic World, Sixth to Eleventh Centuries, ed. Chase F. Robinson; vol. II: The Western Islamic World, Eleventh to Eighteenth Centuries, ed. Maribel Fierro; vol. III: The Eastern Islamic World, Eleventh to Eighteenth Centuries, ed. David O. Morgan and Antony Reid.

2. Regional historical surveys

Asher, Catherine B. and Cynthia Talbot. India before Europe. Cambridge University Press, 2006.

Berend, Nora. At the Gate of Christendom: Jews, Muslims, and “Pagans” in Medieval Hungary, c. 1000 - c. 1300. Cambridge University Press, 2001.

The Cambridge History of Africa, 8 vols. Cambridge University Press, 1975-84; vol. ii: From c. 500 Bc to ad 1050, ed. J. D. Fage; vol. ιιι: From c. 1050 to c. 1500, ed. Roland Oliver.

The Cambridge History of Egypt, 2 vols. Cambridge University Press, 1998; vol. I : Islamic Egypt, 640-1517, ed. Carl F. Petry.

The Cambridge History of Iran, 7 vols. Cambridge University Press, 1968-91; vol. ιv: The

Periodfrom the Arab Invasion to the Saljuqs, ed. R. N. Frye; vol. v: The Saljuq and Mongol Periods, ed. J. A. Boyle; vol. vι: The Timurid and Safavid Periods, ed. Peter Jackson and Lawrence Lockhart.

Golden, Peter B. An Introduction to the History of the Turkic Peoples. Wiesbaden: O. Harrassowitz, 1992.

Holt, Peter. The Age of the Crusades: The Near Eastfrom the Eleventh Century to 1516. London and New York: Longman, 1986.

Imber, Colin. The Ottoman Empire 1300-1481. Istanbul: The Isis Press, 1990.

The Ottoman Empire, 1300-1650: The Structure of Power. Basingstoke and New York: Palgrave Macmillan, 2002.

Kennedy, Hugh. Muslim Spain and Portugal: A Political History of al-Andalus. London and New York: Longman, 1996.

Levtzion, Nehemia. Ancient Ghana and Mali. London: Methuen, 1973.

Lewis, Bernard. The Middle East: A Brief History of the Last 2,000 years. New York, NY: Scribner, 1995.

Morgan, David. Medieval Persia. London and New York: Longman, 1988. Soucek, Svat. A History of Inner Asia. Cambridge University Press, 2000.

3. Aspects of Islamic culture

Alam, Muzaffar. The Languages of Political Islam: India 1200-1800. University of Chicago Press, 2004.

Berkey, Jonathan P. The Formation of Islam: Religion and Society in the Near East, 600-1800. Cambridge University Press, 2003.

Blair, Sheila S. and Jonathan M. Bloom. The Art and Architecture of Islam, 1250-1800. New Haven and London: Yale University Press, 1994.

The Cambridge History of Arabic Literature, 6 vols. Cambridge University Press, 1983­2006.

Crone, Patricia. Medieval Islamic Political Thought. Edinburgh University Press, 2004. Ettinghausen, Richard, Oleg Grabar, and Marilyn Jenkins-Medina. Islamic Art and

Architecture, 650-1250, second edn. New Haven and London: Yale University Press, 2001.

Grunebaum, Gustave E. von. Medieval Islam: A Study in Cultural Orientation, second edn. University of Chicago Press, 1953.

Hillenbrand, Robert. Islamic Architecture: Form, Function and Meaning. Edinburgh Univer­sity Press, 1994.

Karamustafa, Ahmet T. Sufism: The Formative Period. Edinburgh University Press, 2007. The New Cambridge History of Islam, 6 vols. Cambridge University Press, 2010, vol. ιv:

Islamic Cultures and Societies to the End of the Eighteenth Century, ed. Robert Irwin. Robinson, Chase F. Islamic Historiography. Cambridge University Press, 2003. Rosenthal, Franz. A History of Muslim Historiography, second edn. Leiden: Brill, 1968. Rypka, Jan. History of Iranian Literature. Dordrecht: D. Reidel, 1968.

Ullmann, Manfred. Islamic Medicine. Edinburgh University Press, 1978.

Walther, Wiebke. Women in Islam. Princeton and New York: Marcus Wiener, 1993.

4. Primary sources in translation

Frye, Richard N., trans. Ibn Fadlan’s Journey to Russia. Princeton, NJ: Markus Wiener, 2005.

Hopkins, J. F. P. and N. Levtzion, ed. and trans. Corpus of Early Arabic Sources for West African History. Cambridge University Press, 1981.

Lewis, Bernard, ed. and trans. Islam from the Prophet Muhammad to the Capture of Constan- Unople, 2 vols. New York and Oxford: Oxford University Press, 1987.