Research Report

by Dr. R. Stephen Humphreys

posted on November 10





(lecture presented at the University of Tokyo, 21 October 1997)




SINCE 1960


Dr. R. Stephen Humphreys

University of California at Santa Barbara


I will be speaking primarily about the history of the Islamic Middle East, both medieval and modern, since that is the field I know best. I will include other approaches and disciplines insofar as they throw light on the study of history.

I begin with a few general thoughts. History is not really a discipline. It certainly is not a body of data which has been organized within the framework of a unified theory. In this way it differs from the hard sciences like physics or microbiology, even from economics. Nor is it driven by the concern to build models or test theories like sociology and political science. Rather, history is a frame of mind, a way of thinking about human affairs.

Historical thought is ad hoc, holistic, inclusive, and pluralistic. That is, the study of history allows historians to do just about whatever they want. They need only show that their statements are based on valid evidence. Historians focus on particular sets of events for their own sakes, not as cases which exemplify some overarching theory. At the same time, they try to connect these particular events to the whole universe of human thought and action. Most important, historians aim to describe and explain change; they want to show how one situation gave way to or was transformed into another.

In doing these three things -- describing particular events, connecting one set of events to many others, analyzing the process of change -- historians consciously and unconsciously draw on many theories. But they rarely use theory in a rigorous way. Taken as a whole, the study of history is not governed by any single body of theory; there is no one theory accepted by all historians that says which facts are relevant or irrelevant, or how facts must be connected to one another. In this way, history is inclusive and pluralistic. In principle all facts are relevant. Any given set of facts may be explained and interpreted in many different ways.

When we try to describe and explain the actions and thoughts of some group of people in the past, we may do so in two basic ways. First, we can try to enter into their world. We can focus on trying to reconstruct, as accurately as possible, their actions, words, and institutions in and for themselves. We can write a history which says: "So far as we can determine, this what they said and did." We can also try to understand how they thought about what they were doing. What words did they use to describe themselves to one another? What values did they apply to thoughts and actions; what was good and bad, desirable and undesirable, etc.? This kind of history aims to describe and explain events in terms that would make sense to the actors in those events. It means that we try to leave our world and enter into theirs. As a historians in this mode, we are acting as honest and skilled translators; we translate with as little distortion as possible the words and actions of the past into terms that make sense to our own contemporaries.

A second approach to the past is quite different. It tries to connect the events and persons of the past with the ideas and concerns of our time, or it tried to link several disparate periods into some larger structure. Here we are not acting as simple translators. the people whose history we are describing might not understand a single word we say about them. This kind of history is a search for broad patterns and processes, for a unifying logic in things. Here theory plays a much greater role, because only theory allows us to discover (or create) broad patterns and a unifying logic.

With this introduction, I come to the study of the Middle East and Islam in North America. When I began my graduate studies thirty-three years ago, the first type of historical study was paramount, or so we thought. The main issue was finding and learning to use new sources effectively. In particular, the treasures of Istanbul promised a revolution in our knowledge and understanding of the medieval Islamic world. In fact, studies on medieval Islamic history operated on two planes. One was that of trying to discover what medieval Muslims had actually said and done, what social structures and systems of governance they had devised, how they made sense of their world. For this task, good traditional philology, as devised by great German and French scholars of the 19th century, was the main tool--"the painstaking analysis of difficult texts" (L.Binder). The vast amount of new sources made this seem an unusually rewarding task; every time you opened a manuscript, you were likely to find something genuinely new.

The second plane was very different. This was the plane of the broad interpretive framework within which we applied our philological tools. That framework was, of course, "Orientalism." Said's analysis of Orientalism is overdrawn and misleading in many ways, and purely as piece of intellectual history, Orientalism is a very bad book. But it is also and important one, and it did underline how much we were entrapped within a vision that portrayed Islam and the Middle East as in some way essentially different from "the West." That vision was already being challenged by my teachers in many ways. They insisted that it was possible to write the history of Islam and the Middle East just as we would for any other culture and society. Yet the old vision of an "essential Orient" (in this case the Middle East) still shaped many of the questions we asked.

"Late Orientalism" can be trace through three overlapping generations.
H.A.R. Gibb--Gustave von Grunebaum--Jean Sauvaget
Bernard Lewis--Claude Cahen
Marshall G.S. Hodgson

The members of this influential group were all born between 1895 (Gibb) and 1921 (Hodgson). The first generation developed a kind of final synthesis of classical Orientalism, while trying to connect it with new trends if historical and social thought. Von Grunebaum, who reached his intellectual maturity at the University of Chicago during its glory years (1940s-50s) explored the possibilities for comparative analysis provided by contemporary sociology and anthropology. Sauvaget, an urban survey archaeologist, found greater value material culture that in religious and literary texts. Gibb, the most traditional in his formation and approach, had an uncommon grasp of values, and a deep sense of Islam's integrity and validity as a spiritual tradition.

Lewis and Cahen represent scholars who were trained as historians of Europe, and turned their attention to the Middle East only a bit later on. Both were (and of course Lewis still is) hard-core political and social historians, for whom religion and culture were rather secondary and accidental concerns. When you read Cahen's work, in fact, you could just as well be reading about Russian or Italian history; religious beliefs and cultural expression play no substantial part in his analysis. Lewis is a great believer in a well-defined, durable, though always changing entity called "Islamic civilization." But for him the beliefs and customs of Muslims are just ideology --their way of articulating the things that mark the boundary between them and the outside world.

Hodgson -- another University of Chicago product -- was very slightly the youngest of this group. His critique of Orientalist construction of islamic society was far more searching that Said's, but he understood the value as well as the shortcomings of that tradition. His Venture of Islam (written by 1968, published 1974) was both the last work of Orientalism and and effort to frame a new, more adequate paradigm.

The history of the modern middle East (19th-20th centuries) was equally in ferment during the 1960s. At that time, the scholarship on the Middle East since 1800 was markedly inferior by any criterion to that available on medieval Islam. Few modernists read Arabic, Persian, or Turkish; those who did were not professional historians but ex-foreign service officers or missionaries. Nor could they compensate for this lack through the use of European archives. Even the British PRO, with a fifty-year rule, allowed scholars access to official papers only up to the beginning of World War I. The French had a 75-year rule. Finally, the field was crippled by the most intense partisanship and nationalism.

In this contest, the work of Bernard Lewis ( a master of all periods) and Albert Hourani came as a revelation. Both scholars were of course masters of the relevant languages, and both knew Middle Eastern society very well. Lewis' Emergence of Modern Turkey (1958) demonstrated that late Ottoman and Turkish history was simply a part of modern history generally -- there was nothing exotic or "Eastern" about it. It presented the same problems and could be analyzed in exactly the same terms as (for example) the emergence of modern Italy. Hourani's Arabic Thought in the Liberal Age (1962) was the first real intellectual history of the 19th and early 20th century Middle East. With immense erudition and perceptiveness, it showed that there was a genuine secularist intellectual tradition in the modern Middle East. Moreover, Hourani writes without apologetics or polemics.

I have said that the 1960s marked the end of a tradition. But these years also opened a new era. By the end of that decade, the study of Middle Eastern history -- both medieval and modern -- had begun to go in a new direction. By the mid-1970s this shift was confirmed by three major changes: 1)the rise of the book-length monograph -- basically the PhD dissertation -- which vastly increased the body of detailed knowledge at our disposal; 2) the growing interest in disciplines, methodologies, and approaches from outside the field of traditional middle Eastern studies (Geertz, Wallerstein); 3)the systematic critique of the existing scholarly tradition -- not only the famous general critique by Edward Said, but also the attack by John Wansbrough, Patricia Crone and Michael Cook on the accepted interpretation of Islam's origins --which opened up many issues previously left unspoken of regarded as settled.

Until about 1970, the article had been more important than the book in North American and British scholarship on the Middle East. Thereafter the book has become the main instrument of scholarly communication. The book-length monograph (usually a revised dissertation) reflected an institutional change in Middle Eastern Studies -- the huge investment in Area Studies by the US government beginning in 1958. In 1972 or so, the Lam Bert report declared this effort mostly a failure, but the Lam Bert report was quickly swamped in the stream of publications written by scholars who had been trained through this investment. Though US government funding has waxed and waned over the last 25 years, it remains today a fundamental element in the training of graduate students in all fields of Middle Eastern Studies.

American graduate students who take a PhD in most disciplines connected with Middle Eastern Studies - History, Political Science, Anthropology, etc. -- must publish a book before they can obtain tenure. Normally they must complete this book within six years or so of their initial academic appointment. Hence the flood of books. Indeed, most American scholars make their early reputation on the basis of this first dissertation-based book. Some of these books are hasty and premature, but many are very solid pieces of scholarship, and a few open up genuinely new directions for research and interpretation. I will discuss some of these later in my talk.

The second change in Middle Eastern Studies during the early 1970s was the wide use of theories, models, paradigms, etc., from outside the field. In the United States, Middle Eastern Studies is usually regarded as not very innovative, not creative of new theory. I think that this charge is misplaced, for several reasons, but I cannot go into the reasons here. In any case, Historical Studies in the United States over the last twenty years have been increasingly theory-driven. In this way, they fit within the second model of historical thought which I discussed at the beginning. These studies use theory in two ways.

First, they use it as a way to organize and explain the factual material which they have gathered. This is in fact the most common way that theory is used in studies on Middle Eastern history. For example, a theory of tribal organization originally developed to explain African societies in Uganda and the Southern Sudan is used to explain the tribal societies of contemporary Arabia and North Africa, and those explanations are applied to Arabia in the time of the Prophet (Evans-Pitchard, segmentary organization -- Gellner -- Donner). I do not oppose this approach. Often our data make no sense by themselves, they do not suggest any convincing way of putting them together. A good theory from whatever source can be invaluable, if used carefully. But one must be very critical. How often has Marxist analysis been applied to societies where it does not fit the evidence, or where there is not enough evidence to see how it might fit?

Second, a few historians have blended elements of general social theory (usually but not always Marxist in inspiration) with historical data to generate "applied theories" which are meant to explain specific historical processes. The most famous case is no doubt the dependency theory of Andre Gunder Frank, which has been very widely used in studies on the Middle East since the 16th century. A more sophisticated version, and one used with even more enthusiasm, has been the world-systems approach of Immanuel Wallerstein -- a blend of Braudel and dependency theory. As it happens, no such theories have been developed from within the field of Middle Eastern Studies. On the contrary, Middle East specialists have been quite content to borrow, adapt, or criticize Dependency or World-Systems approaches. I would say in fact that some version of Wallerstein's world-systems approach is the dominant one in recent studies of the economic and social history of the Middle East since ca.1600.

In very recent years -- no more than a decade -- we have also seen the increasing use of literary-critical or cultural theory, especially Foucault. (The fad for Foucault, who knew nothing about the Middle East, we owe to Edward Said, I think.) Two cases in point would be Timothy Mitchell's Colonising Egypt and Brinkley Messick's The Calligraphic State -- both fine books, but explicitly books about language -- about how language is used to reflect or shape relationships of power. Some writing of this kind, especially the closely related post-colonialist writing (often by Indian expatriates living in the United States and Britain), can be terribly affected and almost unintelligible. But I am an old dog now, perhaps too old to learn new tricks.

Interestingly, feminist approaches are just beginning to have an impact on Middle Eastern historical studies. At this point, it is hard to call feminism a theory, for there are as many kinds of feminism as there are feminist scholars. For the time being, perhaps we should think of feminism as a frame of mind rather than as a theory. In any case, there is now a substantial literature from a feminist perspective on the 20th-century history of the Middle East. However, there is very little of it on the thirteen centuries of Islamic history which came before. The reason, I think, is that women have left so few traces in our sources for earlier periods, and almost never do women get to speak for themselves in their own words. There is not much but silence for feminist scholars to work with. In a way, silence is very telling evidence, but in the end it does not support positive conclusions. There are at least two major exceptions to this lack of studies, but both are about exceptional cases, Denise Spellberg, Politics, Gender, and the Islamic Past: the Legacy of A'isha bint Abi Bakr , and Leslie Pierce, The Imperial Harem.

Let me return now to the third change that affected the study of Islamic and Middle Eastern history in the mid-1970s, some twenty years ago. I have already alluded many times to Edward Said's Orientalism, which has had an enormous impact on the study of the modern Middle East. It threw into doubt, indeed disrepute, the whole body of studies that had been published during the 1950s and 1960s. Even the best of these (e.g. Bernard Lewis' Emergence of Modern Turkey) had explicitly or tacitly used liberal modernization theory as the basis for their analysis; they had assumed that the contemporary West represented the model of a modern, rational society; as such, the West was the model toward which less-developed societies were evolving and should evolve. Said's book, published at a time of grave self-doubt and self-criticism among Western intellectuals, encouraged a wholesale attack on the whole modernization/Westernization scheme. In an ironic way, it also emboldened the Islamic activists and militants who were then emerging. These could use Said to attack their opponents in the Middle East as slavish "Westernists", who were out of touch with the authentic culture and values of their own countries.

Said's book has had a much more limited impact on the study of medieval Islamic history, if only because medievalists know how distorted his account of classical Western Orientalism really is. But medievalists have had their own shocker from a different direction, in particular the brilliant graduate-student essay of Patricia Crone and Michael Cook, Hagarism. The Making of the Islamic World (1977). Rejected by most as practically insane when it came out in 1976, this book -- and indeed much of Crone's later writing -- has had remarkable staying power. No one, or very few, have accepted the radical rewriting of early Islamic history proposed by Crone in Hagarism, but everyone has had to go back to the sources, examine them with a fresh eye, and rethink from the bottom his approach to the beginnings of Islam. If nothing else, Crone has sown a pervasive sense of uncertainty and agnosticism about the history of this critical period.

At this point, it seems appropriate to examine the fruits of the revolution of Middle Eastern historical studies caused by these three events of the early and mid 1970s. If, as I say, they changed everything, what kind of work is being done now? There are many ways to answer this question, but the most direct approach is to name a few of the most interesting books of the past five years or so and talk about what they have tried to achieve.

I. Studies on the Medieval Period:

Roy Mottahedeh, Loyalty and Leadership in an Early Islamic Society (1981)

Paula Sanders, Ritual, Politics, and the City in Fatimid Cairo (1994)

Denise Spellberg, Politics, Gender and the Islamic Past. The Legacy of A'isha bint Abi Bakr (1994)

Shaun Marmon, Eunuchs and Sacred Boundaries in Islamic Society (1996)

Devin Deweese, Islamization and Native Religion in the Golden Horde (1994)

II. Studies on the Early Modern and Modern Periods:

Timothy Mitchell, Colonising Egypt

Leslie Peirce, The Imperial Harem. Women and Sovereignty in the Ottoman Empire

Beshara Doumani, Nablus

(Note the number of women scholars in this list; it is not an accident.)

There is of course a great deal of work on the Middle East which is based on careless research and whose ideas reflect merely the passing fashions of the moment. But the studies I have just cited will retain their value for many years, even for decades. First of all, they are based on thorough research; their authors match and sometimes surpass the philological and diplomatic skills of their teachers and predecessors. Second, they know how to use theory (gender studies, world systems, symbolic anthropology, etc.) to elucidate their subjects. In every case, I think, these scholars were drawn to their subjects by the intrinsic fascination of those subjects. But as they worked through their materials, they showed uncommon taste and skill in finding those elements of modern theory which could help them explore these materials in a fresh, revealing way. They combine the best of what their traditional masters had to teach them with the new possibilities revealed by the revolution of the 1970s.

Comments on the lecture of Prof. Humphreys


HANEDA Masashi (University of Tokyo)


Let me start my comments with two personal experiences. The first one is that several years ago, one of my articles written originally in French in 1984 was translated into English by a US based Dutch scholar and published in a periodical named Iranian Studies. This is the famous journal of the Society for Iranian Studies in the USA. One of my colleagues rang me one day to congratulate me. He believed that it was my wish to have the article translated into English. But, in fact, it was translated without the consent of the author and without permission from the editor of the French journal in which my article first appeared. To make the matter worse, the translation got a prize called the Vladimir Minorsky Prize. This is a new prize given to a translation of a work written in a language other than English.

The editor of the French journal, famous Iranologue, Jean Aubin was so angry that he sent a public letter of protest to the editor of Iranian Studies, the matter became a scandal. Finally, a third party, Nikki Keddie intervened. She was obliged to write a response to Aubin in the journal and she officially apologized. But, in my opinion, there was a problem with her apology itself. She said she felt sorry because she mistakenly believed that the necessary procedures had already been carried out. Nevertheless, she still insisted on the necessity of the translation from French into English and declared her intention to continue to offer the prize.

This example clearly shows that at least some American researchers don't want to read even French nowadays. They work almost exclusively in English. I am sure that I should not easily generalize this tendency, because I know that several friends of mine in North America use works in French or German. But, I think the hegemony of English language in the Middle Eastern Studies has become more and more clear. This is a dangerous situation for we non-English speaking people. According to today's lecture by Prof. Humphreys, this phenomena may be one of the results of the stream of publications written by scholars who were trained through funds provided by the US government. The flood of dissertations in North America is truly impressive. We have a long list of dissertations concerning the Middle East which are submitted to universities of the North America every year.

There may come a time when we have to write in English whether we like it or not. How shall we cope with this reality? Before answering this question, let me talk about my second experience.


This happened to me very recently. When I was editing an English newsletter for our project this summer, our American translator complained quite often about the difficulty of the translation. In fact, most parts of the English newsletter were written originally in Japanese and then translated into English. The difficulty of the translation was partly because of the poor Japanese of the authors. I know this is true because I couldn't easily understand what they wanted to say even when I read the original Japanese. But, in order to give credit where credit is due to the authors, I have to stress that thinking and logic in Japanese are quite different than in English. If we translate word for word from Japanese into English, the result might often be not understandable to English speaking people. If we want to make ourselves understood well in English, it would be better to write in English from the beginning. That means we have to change our original way of thinking and use the logic of English from the beginning.

It is a great disadvantage for us Japanese researchers to have to learn how to write and speak in English in addition to Middle Eastern languages. We must learn not only the English language, but also American culture. I know we have a lot of difficulties with this. To be honest, I couldn't have made comments on your lecture without having consulted your paper before.

But, at the same time, I would like to point out our unique position in the world of Middle Eastern Studies. We know both Middle Eastern and Western cultures well and can understand them with much more objectivity than Western people. I believe there remains room for approaches, methods and views of researchers whose culture is different from Western or Middle Eastern cultures. In this regard, I am very happy when I read reviews on our Islamic Urban Studies. Reviewers often point out the existence of a "Japanese view" in our books. That is a point we wanted to bring out and they understood it well! I am sure that as long as we keep this attitude in our studies, our future will not always look so bleak.


Concerning the hegemony of English, I would like to ask Prof. Humphreys what you think of the French and German academic traditions. You mentioned the names of Jean Sauvaget and Claude Cahen as members of the earlier generation. So, in your opinion, it was after the sixties that European and American scholarship separated from each other. Before then, both were in the same framework. If I am correct in assuming that this is what you think, what do you think of the great tradition of French and German scholarship after the sixties?


One more remark. You said that by the mid-seventies, the study of Middle Eastern History experienced three major changes. But, I think you could have added the fourth one, not less important than the others. That is the increase of scholars who are originally from Middle Eastern countries and live in the West. Before the sixties, there were few Middle Eastern scholars who lived in the West and wrote in Western languages. But, after seventies, the number of such researchers has increased considerably in the United States as well as in Europe. This is particularly true in the field of Iranian Studies. This has caused a drastic change in the character of Middle Eastern Studies. Middle Eastern Studies are no longer permitted to be simply a way of viewing others. At least, for the scholars of Middle Eastern History who are from the area, the subject of their studies is their own history and not the history of others. How do you evaluate this new tendency?


KATO Hiroshi (Hitotsubashi University)


My response today to the lecture of Professor Humphreys has two points. The first is on the mutual understanding between the historians in the United States and the historians in the Middle Eastern countries. And the second is on the definition of modernity in the historiography on the Middle East in the United States.

(1) Dr. Humphreys defined orientalism as an interpretive framework within which we apply our philological tools. In my view, however, orientalism is an overview on the "other" in intellectual thought . And as long as orientalism is defined as such, we cannot completely avoid this concept in our studies of the history, culture and society of the "other"; that is, regions and countries other than our own.

We must, however, minimize the gap between the views of the foreign and native researchers. Historical observations and descriptions by foreign historians must be compared with those of native researchers, in this context the Middle Eastern academics, describing their own history and identifying themselves within its context. For this, we foreign historians must exchange ideas with our Middle Eastern counterparts to attain mutual understanding.

From Dr. Humphreys' lecture today, we have learned of some academic trends in the historiography of the Middle East in the United States. But in addition, I would like to know how Dr. Humphreys estimates recent historiographical trends among the Middle Eastern historians, and how he thinks an understanding may be achieved between historians in the United States and those in the Middle East.

(2) In his lecture, Dr. Humphreys commented that Dr. Edward Said's book Orientalism has had a much more limited impact on the study of medieval Islamic history than on studies of modern history, if only because medievalists know how distorted Said's account of classical Western orientalism really is. According to Dr. Humphreys, Orientalism is actually a polemical critique, pitching Foucault's literary-critical or cultural theory against the liberal modernization theory or the modernization/westernization scheme. The modernization/westernization scheme has been until recent years the most commonly accepted and most popular opinion in the historiography of the Middle East.

In the opinion of Dr.Humphreys, it is obvious that Said's Orientalism provokes far more heated disputes in the field of modern historiography than in the medieval. My concern, however, is not the controversies provoked by Said's Orientalism, but that after the publication of Said's Orientalism, one remarkable trend appeared in the historiography of the Middle East, not only in the United States but also in Japan. This is the concept of revisional historiography. One typical work of this kind is Dr.T. Michell's Colonizing Egypt, to which Dr.Humphreys referred in the lecture.

What, then, do the revisionists want to change in their writing of history? The answer is the traditional method based on the liberal modernization theory, both Marxist and non-Marxist. This had been the main trend in the historiography of the Middle East until the mid- 1970's, and this is what Said's Orientalism attacked, as Dr. Humphreys points out. In revisionist works, the emphasis is placed on the analysis of culture rather than of social structures as viewed by the subject under study. And rather than methodological fact-finding in the primary source-materials, the emphasis is placed on the interpretation of text and discourse. Works by revisionists are at times classified as post-colonialist, and sometimes as post-modern, as Dr. Humphreys points out.

Then, here is my question. What is modernity in revisional historiography on the Middle East, if it can also acceptably be referred to as both post-colonialist and post-modern? In my opinion, modernity is a concept deeply connected with two modern systems: politics and economics-- that is, the nation-state and capitalism. In this context, I have been interested in two particular controversies concerning modernity in the historiography of modern Egypt.

One of them is the controversy provoked by Dr. P. Gran's Islamic Roots of Capitalism: Egypt, 1860-1840 published in 1979. Prof. Gran insisted in this book that in the latter half of 18th century Egypt, native elements were germinating into what could have developed into capitalism. These were, he continues, suppressed in the process of modernization under the reign of Muhammad Ali. As for the opinion of Prof. Gran, Dr. Samir Amin strongly saw it as a polemical vision criticizing the traditional liberal modernization theory also attacked in Said's Orientalism. The late Dr. G. Baer and Dr. F. De Jong, on the other hand, severely criticized Gran's views on the level of fact-finding.

As for the nation-state system, we cannot but mention Dr. Daniel Crecelius's book The Roots of Modern Egypt, published in 1981. In his book, Prof. Crecelius returned to the original formation of the Egyptian nation-state during Ali Bek Kabir's period in the latter half of the 18th century, and not to the reign of Muhammad Ali in the 19th century. Prof. E. Toledano, however, in his book State and Society in Mid-Nineteenth Century Egypt (1990), argues that even in the middle of the 19th century, Egypt did not shape itself as a nation-state, but was merely one of the Ottoman provinces, characterized by the Turkish-Egyptian political culture. My last question to Prof. Humphreys is what his understanding is of the above two controversies on modernity as seen in the United States.


(Lecture at the University of Kyoto, 29 October 1997)




R. Stephen Humphreys

University of California at Santa Barbara


It has long been argued that Islamic society developed first and most fully in the urban centers of the Middle East and North Africa. (By an "Islamic society," I mean a set of social relationships, patterns of behavior, ethical norms, and forms of cultural expression which are defined and legitimized on the basis of Islamic concepts and values, specifically the shari`a. Non-Muslims may do or say much the same sorts of things, but their societies are not "Islamic" because they do not define what they do in terms of Islam.) These urban centers may or may not have played such a privileged role in reality; that point is strongly contested nowadays. But the cities of medieval Muslim lands will always have a privileged place in our thinking about Islamic societies, because we are infinitely better informed about the cities than we are about the rural areas where most Muslims lived before modern times. (This is not true of every society: we know a great deal about rural France in the 12th century, or rural Japan under the Tokugawa clan.)

It is thus not surprising that cities have attracted much of the best work devoted to the social history of medieval Middle East. However, this work has been compromised by two major shortcomings. First, many scholars have written about something called "the Islamic city," basing their conclusions either on very narrow fields of data or (just as bad) blending together data from many different times and places. Second, even when they have focused on a single entity, they have tended to overlook problems of change and development.

As I pointed out in my lecture at the University of Tokyo, these traditional shortcomings began to fade during the 1970s. This was partly due to the attack on Orientalism, which made everyone suspicious of even the most sophisticated and highly regarded older works. Partly it was due to the influx of theory from the social sciences and later on literary theory. But most of all it was due to the flood of new sources, especially the Ottoman archives, an umbrella term which covers shari`a-court records of many kinds, tax censuses, fiscal and political reports sent to Istanbul, etc. Beginning with Andre Raymond's masterpiece on 18th-century Cairo, our knowledge of cities in the Ottoman period has been transformed. In addition, there have been very close surveys of monuments, street plans, etc., by both foreign (especially German) and indigenous scholars. (It should be said here that we know far less about Iran than about the old Ottoman lands; Iran never had a "source revolution," and since the other revolution in 1979 it has been extremely difficult to carry out surveys on the spot.) All of this has encouraged us to go back and rethink our ideas about the Muslim towns of earlier eras, although for these (especially before 1400) the evidence is far more scattered and difficult to interpret. For that reason, our ideas about Middle Eastern cities before the Mongol invasions still reflect much of the old Orientalist construction of "the Islamic city."

In this context, the work which my students and I have been doing has a special significance. The goal of the project is to reconstruct, in as many dimensions as possible, the evolution of society in Damascus and Aleppo between the Arab Muslim conquest (635 CE) and the Mongol invasion (1260). A study of this kind, focused on two cities over a period of six centuries, could transform the field of urban history in the medieval Islamic world. First of all, it would address both of the problems identified above, by avoiding diffuseness and disparity of data, and by compelling us to take change and development as a major dimension of our inquiry. Moreover, it would provide a solidly documented and highly detailed model, which we could use both to control the statements of comparative essays and to generate problems and strategies applicable to other cities.

On both substantive and methodological grounds, Damascus and Aleppo are very promising subjects for such an inquiry:

(1) Within the world of medieval Islam, Damascus and Aleppo were -- or at least they ultimately became -- major urban complexes. More precisely, both places moved from a rather peripheral role ca. 750 A.D. to the status of metropolitan centers by ca. 1250 -- if not in terms of population (for neither could ever have exceeded 75,000 inhabitants), then at least in terms of cultural prestige, commercial importance, and political stature. The process through which such a transformation came about is clearly an extremely important one, and few other cities in the Islamic world present such good prospects for reconstructing this process.

(2) The two cities are geographically fairly close together and have historically been of more or less the same size. They have had a great deal of contact on all levels, and at times they have been linked together within a common political framework (though never very comfortably). Even so, each seems to have had a distinctive cultural, economic, social, and political profile. A few examples will suffice. [a] Aleppo had an important Shi'ite community while the Muslim population of Damascus was solidly Sunni. [b] Aleppo was always a frontier city, facing Byzantines, Turkoman migrants, the Mongols, etc., while the security of Damascus -- an interior city -- was more commonly threatened by inter-dynastic quarrels within Islam. [c] Aleppo ultimately became a major link in the international trade between the Mediterranean on one side and Iran and eastern Anatolia on the other. Damascus exported its own products through the ports of Tyre and Acre, but more significant was the north-south road linking Syria to Egypt and the pilgrimage centers Of Mecca and Medina. From a methodological perspective, then, we are here in a position to compare two places that really are comparable.

The project as I have presented it so far is obviously too vast, even amorphous, to be pursued effectively. The first step must be to define a set of specific problems or lines of inquiry, so that data can be gathered and analyzed in a purposeful manner. It seems logical to begin by pointing out certain widely accepted characteristics of medieval towns in the Islamic Middle East. These towns were never legally constituted entities ("municipalities"); in law, they were merely relatively dense aggregations of people, and had no formal institutions of government apart from the few officials provided by the sultan or provincial governor to collect taxes and maintain a minimum of public order. Moreover, the states ruling these cities were usually "alien" or "foreign" in some sense; they were dynasties -- would-be empires -- created by various military adventurers with no ties to the populations they ruled. On the other hand, most urban centers had a very strong sense of their own social and cultural identity. Their inhabitants were far from being passive subjects of military despots, and understood and asserted vigorously their own economic and political interests. If they were not dealt with intelligently, they rapidly became ungovernable. In short, "city" and "dynasty" each had to come to terms with the other. Dynasties needed cities as administrative and economic centers; the cities. Dynasties needed cities as administrative and economic centers; the cities in turn required the dynasties for protection and public order, since they had neither institutions for governing themselves, nor any desire to do so.

In a milieu of this kind, a widely attested group whom we call the urban notables -- those men whose wealth and prestige made them the "natural" intermediaries between dynasties and ordinary subjects -- acquire a special value for historical analysis. Unfortunately, it is extremely difficult to decide what characteristics conferred notability on a man, or -- from another perspective -- who belonged to this group and who did not. The very fact that notables are usually identified in our texts only by status terms -- a`yan, wujuh al-nas, or mashayikh -- suggests the problem confronting us. Occasionally notables held public office, as judges or clerks, but such office was more a badge of their standing than a cause of it. They seem normally to have been men of considerable wealth, the sources of which are quite obscure. (Modern historians normally attribute it to "commerce and landholding," without being able to assign much concrete content to these words.) Likewise, many of them enjoyed a reputation for piety and religious learning, but only a few were recognized first and foremost as scholars ('ulama'). At certain times, especially between the mid-10th and mid-12th centuries, when the dynasties of the central Islamic lands were often very weak, certain notables were the chiefs of armed factions or militia bands (ahdath), though clearly they could not always control their own "followers." Sometimes the notables seem to act like a unified class, sometimes they are ridden by factional, bloody strife.

In the end, in fact, it is frustrating and unproductive to think of the notables as a clear-cut body of men -- a sharp-edged group whose membership can be consistently defined according to objective criteria of class, ethnicity, career patterns, residence, etc. It may well make better sense to think of notables as men who acted in a certain way and in a certain milieu. From this perspective, notables would be men who enjoyed a substantial degree of political influence or power, but who used it to pursue local rather than dynastic or imperial goals.

Following this approach, we might best describe the notables by saying that they were those "men of standing" whom the princes and governors relied upon to placate or mobilize the urban populace, and whom ordinary subjects asked to present their needs and grievances to the rulers. This simple definition needs further elaboration, obviously: we must assume that a man would be able to intervene effectively in certain situations, on behalf of certain clients, but not in others. For example, the imam of a neighborhood mosque might be able to gain the ear of a well-placed notary, but could not expect to achieve anything through an audience with the chief qadi or the city prefect. In other words, notability is a function of context, not a stable fact. The role of the notables, like so many of the fundamental social and cultural institutions of medieval Islam, was informal to an extreme, a matter of consensus, of constantly shifting personal alliances, rather than of legally fixed procedures and institutions. To paraphrase Fergus Millar's well-known dictum about the Roman Emperor, "Notables are what notables do." My own discussion in fact shows how frustrating the problem really is: the notables could act only because they were men of standing, but if that is the case, how did they acquire and retain this standing?

What lines of inquiry are best suited to the description and analysis of this group? Among many possible choices, two have proved to be especially promising:

1) By careful attention to such things as lineages, marriage ties, and teacher-disciple links, we can identify the leading families within each city, the social networks which clustered around those families, and perhaps even the alliances and rivalries among the various networks. This line of analysis will surely do much to reveal the basic principles of social action and cohesion in medieval Syria. There are few examples of this kind of work for medieval Islamic history, no doubt because it requires such an extraordinary investment of time in a field of study which is very thinly populated. A provisional but interesting and useful essay in this direction is Richard Bulliet, The Patricians of Nishapur (1972). In addition -- and obviously closer to our subject -- there is a good chapter on the leading scholarly/notable families of twelfth-century Damascus in an unpublished dissertation by Joan Gilbert (UC-Berkeley, 1975).

It should be admitted that network analysis is very difficult to do properly, because our sources are such that we can only construct rather broad sketches of the original networks. Many of the original actors cannot be identified, nor do we know the multitude of everyday transactions which constitute the actual functioning of a network. Finally, we cannot really get at the process by which any given network emerged and dissolved. (For those who favor sociologese, we can reconstruct "personal networks" fairly well, but not the more interesting "systemic networks.") In spite of their shortcomings, however, these sketches can be a useful heuristic device for understanding how people coalesced into groups and pursued their common interests in medieval Damascus. Moreover, they can help us identify persons and families who in fact enjoyed much influence and prestige, but who are quite invisible in the court chronicles which have heretofore supplied most of our information.

2) In both Damascus and Aleppo, a remarkable series of texts describing the cities and their monuments were composed between the twelfth and early sixteenth centuries. These allow us to reconstruct quite precisely when the essential organs of Islamic life (mosques, convents, etc.) were built and to identify the persons who founded and maintained them. It is reasonable to suppose that those who supported these institutions -- both men and women -- made up an important segment of the sociopolitical elite in each city. I have been able to validate this hypothesis through an exhaustive catalogue of construction and architectural patronage in thirteenth-century Damascus which I compiled over a period of some years.

Studies based on the analysis of networks and patronage are of course only feasible insofar as we possess appropriate sources and at least the rudiments of a scholarly literature. In this regard we are quite well off, at least in terms of the modest expectations appropriate to medieval Islamic history.

To begin with, we now have a fairly secure outline of the political history of medieval Syria. Thus at least one dimension of social-political action -- i.e., state formation -- has been explored in some depth. For the first Islamic century, only the battles and the quarrels within the Umayyad court have been adequately studied. The two centuries between 750 and 950 remain extremely obscure, though a recent dissertation by Michael Bonner (Princeton, 1987) on the Syro-Anatolian frontier should clarify certain issues. After 950, however, the picture brightens. The extremely complex period from 950 to 1075 has been examined in important works by Marius Canard, Ramzi Bikhazi, Suhayl Zakkar, and Thierry Bianquis. The latter's new book (1986) on the Fatimids in Syria is in fact one of the most impressive political studies we have for any period in medieval Islam. The twelfth century is covered by Nikita Elisseeff's three volumes on Nur al-Din (1967), and my From Saladin to the Mongols (1977) analyzes Syrian politics in the first half of the 13th century.

Second, the urban archaeology of Aleppo and Damascus has been studied with considerable care, though much remains to be done. I need not go into detail on this issue due to the meticulous review of the literature in Western languages, Arabic, and Japanese by Professor Miura. I can add three important items to his careful review: a PhD dissertation on the citadel of Damascus by Dr. Paul Chevedden (UCLA, 1986), which strongly revises earlier conclusions about its development; a survey of the architectural patronage of Nur al-Din by Dr. Yasser Tabbaa, also a PhD dissertation (NYU, 1982); and most recently, an impressive book by Dr. Tabbaa on the citadel of Aleppo, Piety and Power in Ayyubid Aleppo (Penn State Univ. Press, 1997).

Note that the literature on the monuments of Aleppo and Damascus focuses strongly on the 12th century and later. Very few monuments from the period 750-1100 are extant, and there has been no serious work on these centuries. Our textual sources have little to tell us, and only excavations (as opposed to surface surveys) will give us useful archaeological data. Since both cities are densely inhabited, excavations are not likely to happen.

The key to the project lies in two vast biographical compilations, the Ta'rikh madinat Dimashq of Ibn `Asakir (d. 1176) and the Bughyat al-talab fi ta'rikh Halab of Kamal al-Din ibn al-`Adim (d. 1262). Ibn `Asakir's work includes some 9000 names, including 800 or so women in a separate volume; only about one-quarter of Ibn al-`Adim's compilation survives, but it has some 2500 names. Obviously these were works on the largest scale. Both works have now been published in some form -- I can go into details if you wish later on -- and are far more usable than they were ten years ago when I began this project. Both works have also been studied: Ibn `Asakir in a PhD dissertation by my former student James Lindsay (Univ. of Wisconsin, 1994), and Ibn al-`Adim in a monograph by David Morray (Leiden: E. J. Brill, 1993). Much remains to be done on both, but at least we have a serious beginning.

Each of these two vast compilations is arranged in alphabetical order, according to the given name (ism) of the subject. Each of them tries to include a biographical entry on every person of importance associated with either city, from antiquity down to the author's own day. In addition, both give citations from a host of carefully identified sources, many of which have disappeared in their original form. The two works are by no means identical. Ibn `Asakir is a rigorous scholar of hadith, and almost all his entries conform to the strict methodology of `ilm al- rijal -- hosts of isnads, meticulous citations of a few hadiths for each biography, long lists of teachers and students. He gives us very little narrative material, and he strongly favors men of religion. Dr. Lindsay correctly points out that his purpose was to demonstrate that Damascus (and Syria generally) was a sacred territory, blessed by God with prophets and holy men. Ibn al-`Adim has a broader scope, and includes narrative material on the events of people's lives when he has access to it. But of course most of his entries are about men of religion as well.

All of this suggests that an effort to build a systematic data base and subject it to rigorous quantitative analysis will pay very poor dividends. In Ibn `Asakir and Ibn al-`Adim, we do not have a cross-section of the populations of Damascus and Aleppo, but rather a highly biased selection. We are best off defining rather small sub-groups or cross-sections, and then subjecting these to a close textual analysis. To go back to the problem of notables, we know that many of the men of religion and bureaucrats must have been notables. But on the basis of the data in Ibn `Asakir and Ibn al-`Adim we seldom know which men were notables, or why, or what they did.

Now, in fact we have discovered and collected the data for a body of notables for Damascus and Aleppo during the Ayyubid period (ca. 1180-1260). However, we did this by cataloging and sorting all available information on the patronage of architecture. For some reason, people who paid for religious monuments to be constructed (whether princes, amirs, `ulama', or bureaucrats) tend to leave some traces of their activity in other realms. These patrons of architecture -- many of them at least -- can often be situated quite precisely within the political and social life of the city. Since we have a fair amount of information on the patronage of religious architecture under the Seljukids and Zengids, it will be possible to achieve more or less the same thing for the 12th century. But before 1100 information is extremely sketchy. In the present state of our sources it is unlikely that we will ever know in any detail how Damascus and Aleppo evolved during the 300 years between the fall of the Umayyads and the coming of the Turks.

Although it is true that most of the information in Ibn `Asakir and Ibn al`Adim deals with `ulama', and that this information sticks to a narrow formula, we can learn quite a bit about some aspects of these men's lives. If we cannot say just how they functioned as notables, a`yan, we can say a fair amount about who they were and where they came from. After all, we know their names and genealogies, we usually know where they were born and grew up, we know where and with whom they studied, we know where and when they died. In this vein, I will share with you some preliminary results on the `ulama' of Damascus during the century of Fatimid rule in Central Syria, ca. 970-1070 C.E. We developed a small sample -- 113 men whose death-dates are well-attested, drawn from four name-groups (Ahmad, Ibrahim, Hasan, Husayn). We searched these four names because two seem religiously neutral -- any Muslim can have them -- and two might favor the otherwise elusive Shi`ites.

Of the 113, 36 were explicitly identified as long-term residents of Damascus: 8 were natives of the city, 11 were immigrants, 17 were of uncertain regional origin. There were in addition, 33 men identified as mere visitors to the city -- men who passed through on the hajj or to exchange hadith with the scholars of Damascus. That leaves 44 persons whose residence is uncertain; we do not know how long they stayed in Damascus.

How many of the 36 permanent residents of Damascus might have been notables -- regular, effective intermediaries between the government and local society? We decided that a crucial element in making a man a notable was religious standing. A reputation for piety and learning guaranteed access to the governor's chamber, though of course it did not guarantee success. To this end, we separated out a group which I call the "religious establishment" -- all those men who held formal religious office such as qadi, imam, muhtasib, wakil bayt al-mal, or who were registered witnesses (shuhud). We found 18 members of the religious establishment. Of these 18, perhaps 9 might have been notables. By way of comparison, only 1 person out of the 18 not from the religious establishment seems to be a possible notable.

On what grounds do I make the assertion that these nine (or maybe ten) men were notables? Remember that notability depends on what a man does as well as who he is. Without adequate narrative material, we cannot establish that any one of these men actually was a notable. At this point, all we can do is say who is likely to have been a notable -- who could have played that role if chose to do so and the circumstances were right.

We have in our gang of ten 5 qadis (including three sharifs), 3 mosque officials, 1 muhtasib, and 1 ordinary scholar. The qadis are the most obvious potential notables, obviously. But qadis and muhtasibs are of course regime officials, appointed by the ruler and serving at his pleasure. Hence it is fair to ask how effectively they could play the delicate intermediary of the notable. As long as they hold office, they are bound to the regime and required to carry out its policies. And if they leave office, do they still retain that precious commodity of access to the governor's chamber? The question is still open.

Interestingly, of our group of ten, only three were certainly natives of Damascus. Three were immigrants, and four others we cannot be sure about. Apart from that, four were Syrian by family origin, and the four others had family origins elsewhere -- two in Iran, one in Spain. Only one is clearly from an old Damascene family.

The result of this analysis is a bit revisionist. The local social, economic, and cultural elite from whom the notables supposedly emerged was not really a native elite at all. Some substantial portion of this elite came from families who were new to the city, or at least had resided there for only a generation or two. We may have to give up the idea that the notables represented the deep-rooted, indigenous families of the city. Perhaps that should not be surprising. The crucial element of notability is prestige and access to the ruler. This prestige and access could be achieved quite readily by newcomers of the right kind. Obviously there would be tensions and conflict between the "old families" and the "interlopers" -- but those tensions are what makes the notables interesting. At any one time, there were only a few available slots for notables, and families or individuals had to compete for those slots.

Let us move to a different subject, and ask about the contacts of Damascus with the outside world. Scholarly thinking about Fatimid Syria has been guided by two plausible but unexamined notions:

1) that Syria generally, and Damascus in particular, was a provincial place after the fall of the Umayyads, a kind of stagnant backwater of Islamic life;

2) that the Fatimid conquest of Palestine and Central Syria intensified the isolation of Damascus from the main currents of Sunni thought and practice.

Our data suggest a very different picture. First, it is incontestable that Damascus was after 750 CE a provincial rather than an imperial center, but it was far from moribund. It remained the most important town in a very important province. Second, Damascus was in no way isolated from the religious and intellectual currents of the age. Recall some of the figures who lived in Syria during the 10th and 11th century: Farabi, Mutanabbi, Abu'l-`Ala' al-Ma`arri, Muqaddasi. It is not Baghdad or even Nishapur, but it is hardly an intellectual desert.

We can develop this point further by looking at the place of Iran in our data. Men of Iranian background were not a terribly important element among our 36 long-term residents in Damascus. Only 5 immigrants to Damascus (and 2 more to the Lebanese coast) came from Iran. However, Iran had a far more important role among those who were just passing through Damascus. (This is admittedly an artifact of the sources, to some degree. Still, it is significant that Egypt, North Africa, and Spain produced no visitors of sufficient importance to show up in the Syrian biographical tradition.)

Of 33 visitors, 20 were from Iran, 6 more from Iraq. Several of these were men of great stature, like the ra'is of Nishapur or Abu Bakr al-Khwarizmi, regarded as one of the greatest hadith-scholars of the age.

To examine Iran's impact on Damascus from a different angle, our sample of 113 had 11 Sufis (10 percent of the total). Of these 8 were of Iranian origin. Five were visitors, and two took up residence in Tyre. The impact of these men is apparent from the epithets that Ibn `Asakir's sources attach to them: shaykh al-Sham fi waqtihi, tawus al-fuqara' ahad al-rahhalin fi talab al-`ilm. We can thus argue that Sufism in Fatimid Syria had its roots in Iran, and specifically in Khurasan and Transoxiana. Sufism clearly found respect and acceptance among many of the Syrian `ulama' in this period (11th cen. CE), but it was planted and nurtured by Iranian immigrants and visitors to that country.

This has been a superficial review of a project which is only in its early stages. Nevertheless, I hope that I have said enough to give you an idea of what we are trying to do. I also hope that you will see the interest and value of Aleppo and Damascus for our understanding of the early Middle Ages in the Islamic world.