For Muslims, history is important. The mission of Mohammed and the promulgation of the Qur’an are events in history, and knowledge of them was preserved and disseminated through historical memory and record. In this respect Islam takes the same view as its two predecessors, Judaism and Christianity, and indeed the earliest Islamic narratives bear a generic resemblance to those of the Jews and Christians, in scripture and elsewhere. The Muslims too had their kings and their prophets, their saints and their martyrs, and preserve the memory of their lives and their deaths in history and biography, in tradition and commemoration.

But there is more. Judaism began among a small group of migratory tribes, grew among the inhabitants of a small kingdom, overshadowed and often dominated by mighty neighbors, and achieved its greatest flowering among a people who were conquered, subjugated, and ultimately dispersed. Christianity first appeared as the faith of a small minority in a subject province of the Roman Empire, and remained through its early formative centuries a religion of the downtrodden and the oppressed. Islam, in contrast, triumphed during the lifetime of its founder, who created a state which, under his immediate successors, the caliphs, became a vast empire.

The Christianized Roman Empire strove, albeit with limited success, to preserve the language, the laws, and the institutions of pagan Rome, and even the barbarian conquerors who became its real masters paid at least lip service to the state and church which they had conquered.

The situation in the Islamic caliphate was totally different. The Muslim Arabs, unlike the Western barbarians, brought their own scripture, in their own language, and created their own imperial system and structure. Though much remained of the Roman and Christian past in the former Roman Christian provinces of the Levant and North Africa, it evoked no respect, and conferred no legitimacy. Its survival was, so to speak, surreptitious, and ultimately vestigial. In the new Islamic polity and society, only Islam conferred legitimacy; only Islamic precedent, that is, the Islamic past, could validate government and law.

Even during the lifetime of the Prophet, and much more frequently after his death, pressing problems arose for which the Qur’an provided no explicit answers. At an early date the principle was adopted that the Prophet was divinely guided in all his actions and utterances, and that after his death the divine guidance was given to the Muslim community as a whole. “My community,” the Prophet is cited as saying, “will not agree on an error.” The practice of the Prophet and the decisions made by the early caliphs thus constitute a body of precedents, ranking second only to the Qur’an itself among the authoritative sources of the shari’a, the holy law of Islam. It is designated by a variety of terms, the commonest being sunna, meaning, approximately, the corpus of custom and example left by revered predecessors.

But between the Qur’an and the sunna there was an important difference. The Qur’an was scripture, in the Muslim view literally divine, having been dictated to the Prophet by an angel. A written and authorized text was established at an early date, and, apart from a few minor and insignificant variants, there is no argument about the accuracy or authenticity of the canon. The sunna, on the other hand, though divinely inspired, was human, and therefore subject to error and even fraud. It consists of a multiplicity of traditions, orally transmitted for generations before they were committed to writing. Human memory is always fallible, and the bitter religious, social, and political struggles of the early Islamic period encouraged the distortion or even the fabrication of traditions, designed to support an argument, a faction, or a cause.

From an early date Muslim scholars recognized the danger of false testimony and hence of false doctrine, and developed an elaborate science for criticizing tradition. “Tradition science,” as it was called, differed in many respects from modern historical source criticism, and modern scholarship has often disagreed with the evaluations of tradition scientists about the authenticity and accuracy of ancient narratives. But their careful scrutiny of the chains of transmission, and their meticulous collection and preservation of variants in the transmitted narratives, give to medieval Arabic historiography a professionalism and a sophistication without precedent in antiquity and without parallel in the contemporary medieval West. By comparison, the historiography of Latin Christendom seems poor and meager, and even the more advanced and complex historiography of Greek Christendom still falls short of the historical literature of Islam in volume, variety, and analytical depth. For Sunni Muslims—the Shi’a take a different view—God’s community was the embodiment of God’s design for mankind, and its history, providentially guided, revealed the working out of God’s purpose. An accurate knowledge of history was therefore supremely important, since it could provide authoritative guidance in both the profoundest problems of religion and the most practical matters of law.


History was important—that is to say, Muslim history. The history of non-Muslim states and communities, which did not accept God’s final revelation and did not obey God’s law, offered no such guidance, and possessed no such value. Muslim historians therefore paid scant attention to non-Muslim history, whether of their neighbors in Christian Europe and elsewhere, or of their own Christian, Zoroastrian, and pagan ancestors. Some knowledge of the immediately pre-Islamic past was retained, and included in the corpus of historic knowledge. Of the great civilizations of the ancient Middle East, all that Muslim historiography knew was what was related in the Qur’an, with some additional explanatory matter, mostly derived from Jewish and Christian informants. The rest of the ancient past was forgotten and, often literally, buried beneath the ground. Perhaps even more remarkable was the almost total disregard of Europe which characterizes Islamic writings in Arabic, Persian, and Turkish from the beginning of historiography until the nineteenth century, when the advancing power of Europe demanded, and finally received, attention.

For a long time, European Christian historians showed a similar lack of interest in non-Christian history. They were of course very much concerned with the advance of Islam, which had wrested the eastern and southern shores of the Mediterranean from Christendom, and which for a thousand years threatened Europe itself from both ends, the Arabs in the West, the Turks and Tatars in the East. But the study of Islam in medieval Europe, such as it was, was religious. It was pursued principally in monasteries, and had as its primary aim to protect Christians from conversion to Islam—a very real danger at the time—and, ultimately, to launch a Christian counter-mission for the conversion of the Muslims to Christianity.

By the end of the Middle Ages, it was becoming increasingly clear that the first task was no longer necessary, the second impossible. But if the Islamic religion was no longer seen as a serious rival, Islamic power, now embodied in the rising empire of the Turks, remained a major threat. European writers, their concerns aroused by the Turkish conquests, their curiosity whetted by the adventures in ideas of the Renaissance, their methods sharpened by the new learning, devoted increasing attention to Turkish history, and produced a considerable body of writing on the subject. By the beginning of the seventeenth century, an English clergyman, Richard Knolles, was able to publish an immense volume on the history of the Turks, based on a wide range of continental European writings, some of them translations from Turkish originals.1

But all this was what we would nowadays call modern and contemporary history, with some minimal necessary background. This literature paid little attention to classical Islam, and it had no place in the universities, or in the research and teaching carried out under their auspices. The universities of the time were not interested in modern languages like Turkish, any more than they were interested in French or German; they were not interested in the history of Turkey, any more than in the history of Germany or France. The only history worthy of the attention of academic scholars was ancient history, based on classical and scriptural sources, and it was to this aspect of Islam that, in due course, scholars in the European universities began to devote their attention.

The first chair of Arabic in France was established by King Francis I at the Collège de France in 1529; the first chairs of Arabic in England at Cambridge and Oxford in 1633 and 1636. Others appeared at about the same time in the German states and, notably, at the University of Leiden in Holland, which from the seventeenth century onward became a major center of Arabic and Islamic studies. Those engaged in these and similar studies came to be known as Orientalists, a term coined on the analogy of the Hellenists and Latinists whose intellectual discipline and rigorous philological method they adapted to the study, and later to the edition, translation, and annotation, of Oriental texts. These texts were principally in Hebrew, Syriac, and Arabic, and the third led naturally and inevitably to the study of scriptural and then classical Islamic history—that is, the life of the Prophet and the history of the medieval caliphate.

Like the Jewish scholars who helped the first Christian Hebraists, so also Arab Christian scholars from the Middle East contributed significantly to the early development of Arabic scholarship in Europe. These early modern scholars performed a task of the greatest importance in discovering, editing, analyzing, and interpreting the original sources for early Islamic history. In doing so they laid the foundations of virtually all modern scholarship in this field, in the Arab lands as well as in the West. At first they published most of their work in Latin, still at that time the common language of European scholarship. But soon, especially but not exclusively in Protestant countries, they began to use their own languages. By the nineteenth, still more the twentieth century, the serious student of classical Islamic history required a wide range of linguistic equipment, giving him access not only to the indispensable original sources, but also to scholarly monographs in all the major and several of the minor languages of Europe.


Arabic and Islamic studies presented—to some degree still present—special difficulties, as compared with Latin and Greek, or even with medieval European studies. The quantity of surviving texts was enormous, but in the early days of Orientalism they were almost all still in manuscript, scattered through countless libraries, many of them difficult or impossible of access. Very few were printed, and the small production of published texts was for long due largely to the efforts of Western scholars. Printed editions of classical texts did not appear in Middle Eastern countries until the mid-nineteenth century; critical editions not until the twentieth. Because of their small numbers, and the magnitude of their tasks, the Arabists of Europe fell far behind the Hellenists and Latinists.

The relative backwardness of Orientalist scholarship gave rise to, and was in turn aggravated by, the lack of certain basic research tools which before long were taken for granted in other fields of study. The medieval Arab scholars had left a magnificent legacy of studies on the grammar and vocabulary of their language, but the task of transforming this into modern philology and lexicography took many generations. The medieval Islamic states and societies left vast hordes of coins and innumerable inscriptions, but the science of Islamic numismatics and epigraphy, though begun in the sixteenth century, did not achieve major results until the twentieth, when it also benefitted from contributions by scholars in the Muslim lands.

Despite all these difficulties, substantial progress was made. Already by the late seventeenth century, when a French scholar called Barthélemi d’Herbelot published a multivolume alphabetical dictionary of Oriental civilizations,2 he was able to draw on a substantial body of printed scholarly work. A century later, when Edward Gibbon was writing the Arab and Turkish chapters of his Decline and Fall of the Roman Empire, sufficient material was available to him, both translated sources and recent studies, for him to perform this task with his customary flair and distinction.

But there were still many gaps in the equipment of the Arabist. He did not have—we still do not have—a historical dictionary of classical Arabic, or even a satisfactory dictionary of classical Arabic usage. The preparation of a comprehensive Arabic dictionary is now underway, for the first time, under the auspices of the Arabic Language Academy in Cairo, with the cooperation of many scholars in Arab and other lands. When this work is completed, it will be an immense boon to scholarship concerned with all branches of Arabic and Islamic studies.

Another deficiency, particularly felt by students, was the lack of the kind of bibliographical and methodological handbooks available in most other fields of historical and philological studies. Muslim scholars—Arabs, Persians, Turks, and others—had been writing bibliographical works since the high Middle Ages, but this literature, like that of the classical lexicographers, had not been reshaped in accordance with modern bibliographical standards and practices.

A first Orientalist bibliography was published in Leipzig in 1846.3 It contained a classified list of published texts in Arabic, Persian, and Turkish, most of them, as was inevitable at that date, printed in Europe. An Italian work,4 published in 1916, provided invaluable guidance to several generations of scholars and students; it formed the basis of a more recent work in English.5 These provide detailed information about reference works, sources, libraries and archives, and centers of study; they offer, however, no methodological guidance, and no critical examination of either the problems discussed or the solutions propounded. Gustav Pfannmüller’s handbook of literature on Islam, published in 1923, does offer such discussions, but only of Western literature, and not of the Arabic sources, with which he was apparently not acquainted.6 It is still useful as a guide to the early European travel and scholarly literature.

The first who attempted to do for students of Islamic history what scholars like Ernst Bernheim and Louis Halphen did for historians of the West, the one in his handbook of historical method, the other in his initiation to the study of the Middle Ages,7 was a French Orientalist called Jean Sauvaget, whose introduction to medieval Islamic history was published in Paris in 1943.8 Conditions in wartime Paris were not favorable for the preparation of full and up-to-date bibliographies, and the bibliographical part of Sauvaget’s book is now badly out of date. It remains, however, invaluable for its methodological chapters, in which he enumerated the various categories of source material and the ways of approaching them, and for his incisive and often pungent comments on previous work in the field, sometimes of a severity nowadays usually reserved for political and ideological opponents, and not, as previously, for flawed scholarship. A revised edition was prepared by Claude Cahen and published in 1961, in which Cahen added a great deal of bibliographical information, but removed most of Sauvaget’s sharper critical notes. An English translation of this edition was published by the University of California Press in 1965.9 A new and entirely recast French text was published by M. Cahen in 1982.10 There are a number of other bibliographical works,11 but basically the sequence of works by Sauvaget, Sauvaget-Cahen, and Cahen alone have continued to provide the most basic and comprehensive guidance available. There has hitherto been no other work of comparable scope, depth, and value.

To these however we must now add a new book by Professor R. Stephen Humphreys, already known for his studies on Syria and Egypt in the later Middle Ages. The first edition of this important work was first published apparently for limited circulation by the Bibliotheca Islamica of Minneapolis in 1988. It has now been reissued, in a revised and updated version, by the Princeton University Press. Professor Humphreys offers something less, and very much more, than the works described above. Only the first two chapters—one a critical, classified bibliography of works of reference, the other an analytical survey of the sources—provide the kind of bibliographical and technical guidance given by Cahen and Sauvaget, whose work they usefully supplement, but do not replace.

The remaining ten chapters deal with a series of problems that have particularly preoccupied students of Islamic history in recent years. They include such topics as the reassessment of early historical tradition on the beginnings of the Islamic state; the mid-eighth-century revolution that overthrew the Umayyads and installed the Abbasids as caliphs, thus transferring the center of the Arab-Islamic empire from Syria to Iraq, and transforming both its social base and political structure; the role and status of the ‘ulama‘, the professional men of religion, in medieval Islamic society who were at once jurists and theologians, though not, in the Christian sense a priesthood; and urban topography and society with special reference to late medieval Damascus.

Several chapters deal with problems arising from classical Islamic historiography in both Arabic and Persian, and with the complex double problem of the importance of Islamic law in Islamic history and the use of Islamic legal texts by the historian. The first Orientalist students of Islam, impressed on the one hand by the central role of law in Islamic society and religion, on the other by the mighty structure of Islamic jurisprudence, tended to treat the normative statements of the jurists as descriptions of real events and usages. In this, they were both preceded and followed by historians within the indigenous tradition. A later generation of scholars, becoming aware of the often wide disparities between the principles of Islam as prescribed in the law books and the practice of Muslims as revealed in documents, went to the opposite extreme, and dismissed the whole juristic literature as a system of ideals unrelated to reality—“as if it embodied only the vain dreams of aged ‘ulama’ cloistered in their mosques, and did not hope to mold the affairs of everyday life.” As Professor Humphreys remarks, “until very recently historians of Islam have been extremely reluctant to regard law and jurisprudence as a significant source for social and economic life. For anyone familiar with the importance of law in the study of Roman or English history, such behavior must seem simply obtuse.”

Professor Humphreys traces the stages in the evolution of scholarly thought and method on this matter, and offers guidance on the place of law in Islamic history, the difference between Muslim and Western perceptions of the nature of law, the range of Muslim legal texts and the studies based on them, and the manner in which these texts can and should be used, for both legal and general history.

The two final chapters deal with two non-elite groups—in a brief but interesting argument, Professor Humphreys gives his reasons for avoiding the term “minorities,” preferring to speak of groups “systematically excluded by law or custom from the enjoyment of all the benefits which that society has to confer.” Islamic law recognizes three categories of legal inferiors: slaves, women, and unbelievers. Human nature, in defiance of Islamic law and morality, created other categories of inferiors, both racial and social. Professor Humphreys has chosen two groups of inferiors, one de jure and one de facto—the non-Muslims and the peasantry, whom he describes as “the voiceless classes of Islamic society.” He examines the important work of A.K.S. Lambton and I.P. Petrushevsky on agrarian relationships in Iran, some studies on the fiscal and land-tenure systems of the Arab Mediterranean countries, and more briefly, of the Ottoman Empire, which for the most part falls outside the chronological range of the book. The chapter ends with a short review of more recent work on the condition of peasants in the Middle East, and a discussion of how such studies can help a better understanding of the past.

For the non-Muslims, there is extensive documentation, both from the Islamic side and from the non-Muslim communities themselves, and a very considerable body of scholarship. The discussion of the role and status of the dhimmis, the tolerated non-Muslim, in Muslim society and under the Muslim state has probably produced more nonsense than any other topic in Muslim history, ranging from bitter complaints of Hitler-style oppression at one extreme, to myths of an interfaith utopia at the other. Professor Humphreys restores some balance to this question, and also reviews the important matter of conversion to Islam both by individuals and by groups—when it occurred, in what circumstances, and with what response from coreligionists in both the abandoned and the adopted religions.

Much less is known, and much less has been written, about the rural population, and, more generally, about what went on in the countryside. Contrary to popular mythology in the West, medieval Islamic civilization was overwhelmingly urban; its historiography, its literature, and its laws discuss urban problems and reflect urban conditions. Not until Ottoman times do we have archives, from which it is possible to study in detail the day-to-day life of the peasantry; not until very recent times do we find much literature depicting the life of the peasants—even less, peasant literature. Professor Humphreys’s chapter on the countryside is therefore of necessity limited to such matters as the physical setting, technology, land use, and land tenure, and in a brief final section, “Agriculture and the Social Order.”

Each of the chapters follows more or less the same pattern. First, a problem is identified and stated in broad terms, and then is studied in one or two examples in a carefully defined time and place. Professor Humphreys then goes on to examine the sources: their variety, their evidence, and their deficiencies, and to look critically at both earlier and recent scholarship. Young scholars will be especially grateful for his frequent suggestions on work still to be done.

A book of this kind, in the nature of things, cannot be comprehensive, either in the range of topics that it covers, or in the scholarly literature reviewed for each of the topics included. Every specialist reader will no doubt note and regret some omissions under both headings. The treatment of legally established inferiors is limited to non-Muslims, and does not extend to the other at least equally important groups, slaves and women. The treatment of the lower social classes is limited to the countryside, and says little about the urban working class or about the social fringes described in C.E. Bosworth’s fascinating volumes on the vagabonds and beggars and other “groups outside the pale of normal society” in medieval Islam.12 Bureaucracy, warfare, and sects might also have deserved separate treatment in a study of a society in which bureaucrats, soldiers, and religious dissidents played so important a part. There are also a few bibliographical omissions, even in the topics that are covered in the book. Thus, for example, in the treatment of the tenth-century Buyids, a dynasty of Shiite Iranians, whose rise to power marks a turning point in medieval Islamic history, mention is made of only one of Joel L. Kraemer’s two volumes on the cultural and intellectual world of that time—a major contribution to the intellectual history of medieval Islam.13

Two omissions may be noted in the chapter dealing with the recent critical study of the beginnings of Islam and the first Islamic polity. It might have been useful to mention the very radical criticism of the biography of the Prophet, the Qur’an, and the early Caliphate propounded by a group of Soviet scholars in the late 1920s and early 1930s.14 Most of them were inspired primarily by polemical and ideological motives, sometimes reminiscent of those of the religious controversies of the Middle Ages. Much of their work was published under the auspices of a body known as “The League of the Militant Godless,” and formed part of a three-pronged attack on the three religions surviving in the USSR. These writers did however anticipate some of the points, or at least some of the questions, raised by more recent scholars in the West, including such highly sensitive matters as the historicity of Mohammed, and the actual manner and period of the composition of the Qur’an.

Soviet scholarship is not the only omission in this chapter; the Muslims are missing too. The critical study of Jewish and Christian scripture and sacred history has almost entirely been the work of Jewish and Christian scholars. The critical study of early Islam has overwhelmingly been undertaken by non-Muslims. For that reason, special interest attaches to the small number of recent Muslim scholars, writing in Arabic, who have contributed to this discussion, and have raised questions of a kind now commonly accepted among contemporary theologians and Biblical scholars in the West, but not yet heard in Muslim debate. Perhaps wisely in these dangerous times, Professor Humphreys has refrained from naming them; for the same reasons I shall not name them here.

All these are comments, not reproaches. Professor Humphreys’s book, as it stands, will henceforth be indispensable to teachers of Middle Eastern history, invaluable to serious students in the field, and extremely useful to scholars whose fields of study abut on medieval Middle Eastern Islam, either in time: Ancient Middle East, Modern Middle East—or in space: Byzantium, Mediterranean Europe, Africa, India.

But there is another, and larger, sense in which Professor Humphreys has rendered a service. At this time, when both the study and the neglect of other civilizations are condemned—by different critics, sometimes even by the same critics—Professor Humphreys’s review of a great tradition of scholarship provides an answer to both. With the deep knowledge and understanding of the “insider,” but at the same time with that unsparing critical analysis that is the hallmark of the Western scholarly tradition at its best, Professor Humphreys shows, through his chosen examples, how Western scholars approach and study the history of another civilization, the principles and methods that inspire them, the mistakes they can make, and the results they can achieve, notably through the discovery and elucidation of the documents and monuments that remain. And, at the same time—and in this too, he is an authentic representative of the best scholarship, but by no means the whole of Western scholarship—he has a deep and sympathetic understanding of both the concerns and the achievements of the civilization that is studied.

This Issue

December 5, 1991