The Ethnographic State

ONE

France and the Sociology of Islam, 1798-1890

Writing in the New York Review of Books in 1971Clifford Geertz argued that the old explanations of North African society were no longer valid, if they ever had been. When he arrived in Morocco in 1965, with the classics of French colonial ethnography as his guide, Geertz looked immediately for the primordial groupings that he had been led to believe structured social relations at every level. Yet on closer examination, he discovered, the concepts of tribe, saintly lineage, sufi tariqa, and even the extended family, tended to dissolve before his very eyes. The French colonial literature seemed suddenly suspect; all now was dyadic ties. What had changed of course was not just North Africa. The observer had also changed. As a consequence of the shifting patterns of world politics in the postwar era, as well as changes in intellectual fashions, Morocco was suddenly of interest to British and American social scientists. With these sea changes it was perhaps to be expected that the image of North African society in Western scholarship would be temporarily out of focus. But so too, in a way, was the French tradition of the sociology of Islam. How and why the latter was transformed is the subject of this chapter.

In Search of the Sociology of Islam

The French tradition of the empirical study of Muslim societies began in 1798 with the Napoleonic expedition to Egypt and the twenty-three volumes of the Description de l'Égypte. A second formative moment was the work of the first generation of French researchers in Algeria (1830-48). Thereafter the tradition emerged largely within these early parameters, its major phases corresponding to the shifting patterns of French colonialism. By the outbreak of the Algerian war in 1954, it had become a mummified version of its former self, and in its evident inability to explain the outbreak of the war, or its raison d'être,collapsed of its own weight. Somehow an intellectual tradition that had begun with aspirations of bringing the fruits of the French Revolution to the lands of Islam had become instead an apologist for empire, a disseminator of racist stereotypes, and a producer of irrelevant folklore. It is no accident that the existence of this tradition dates from the beginnings of French imperialism in the Middle East in 1798 to its bloody and convulsive end in 1962 with the independence of French Algeria.

But can there be said to be a French tradition of the sociology of Islam? The very terms of such a formulation are full of traps for the unwary. Indeed, at the risk of sounding perverse, I am tempted to observe that the sociology of Islam does not exist, and that it has until recently been preeminently a French tradition. Part of the problem, part of the misconceived problematique of the field, is the assumption that Islam is an appropriate field for sociological inquiry. The question of the relationship between the study of Muslim peoples and the discipline of orientalism, never an easy or straightforward one, is thereby posed. Then there is the fact that most of the French sociology of Islam was produced by individuals who were not sociologists at all. Rather they were colonial native affairs officers, civilian amateurs, and orientalists. Academic sociologists did not concern themselves seriously with such questions until the twentieth century. Thus we will have to consider the relationships between those whom we are calling sociologists (using the word in its broad nineteenth-century sense), and intellectual circles in metropolitan France. A third objection can be raised to the title of this chapter. It presumes that something called a tradition of the sociology of Islam exists. What does it mean to speak of a tradition of the sociology of Islam? Before going on to discuss in greater detail the legacy of French sociology of Islamic societies (which, as we will see, is principally a sociology of North Africa), it is necessary to consider these questions further.

To speak of a tradition of French sociology of Islam involves an intellectual exercise the implications of which we should be aware. To retrospectively construct a tradition is in some sense already to validate it, to accord it legitimacy, to assert its existence. A tradition in this sense, what Foucault calls a discours (or Bourdieu a doxa),is invested with a kind of mana that is highly interested. In this sense, a tradition is a politically structured discourse, whose function is to dominate, control, and or orient our understanding. As we are no longer contemporaries of the period in which this discourse flourished, there is no reason for us to accept its central assumptions. Because we are not innocent consumers of its product (colonial sociology), but are forced to have reference to it insofar as we are interested in North African society, it is crucially important that we situate its practitioners against the political and intellectual background in which they flourished, and in the context of their central assumptions about that society.

Historically the study of Islamic subjects in France followed one of two paths: either the discipline of orientalism, a linguistically defined field based on the critical study of texts written in oriental languages, or the less rigorous tradition of colonial studies of Muslim societies, the product of amateurs and enthusiasts, expatriate sojourners in an exotic world they sought to understand. French orientalists dominated the study of Islamic subjects in the metropole. From the time of Silvestre de Sacy (1757-1838), who can in many ways be said to have single-handedly invented the field, orientalists came to occupy positions of institutional power within the Parisian academic establishment: in the department of oriental manuscripts of the Bibliothèque nationale, the École des langues orientales, the Collège de France, and the Académie des inscriptions et belles lettres of the Institut de France. Although mastery of Islamic languages made orientalists useful in supporting French colonial adventures in North Africa, the fundamental impetus of the discipline moved in other directions. Resolutely hostile to the study of contemporary subjects and devoted to the formalist study of classical texts, orientalists looked with deep suspicion on those who studied the living languages of the Middle East. Engaged in a field that they regarded as the elaboration of "the historical science of the human spirit" (Renan), they refused to sully their intellectually noble calling with less worthy intellectual pursuits. Not until the end of the nineteenth century, with the introduction of comparative linguistics and the introduction of the spoken oriental languages to the curriculum of the École des langues orientales vivantes, did the discipline undergo important changes that were eventually to lead to its renewal.

What orientalism contributed to the study of Islamic societies was the concept of Islamic civilization. This had its advantages, as well as its liabilities. On the one hand, it compelled students of colonial North Africa to recognize the historical past of North African Muslims and their place in the wider world of Middle Eastern Islamic civilization. Had the concept of Islamic civilization been taken seriously, it might have inhibited the tendency to conceive of the Maghreb as an island cut off from its Mashriqi point of reference. On the other hand, civilizational studies accredited essentialist notions about the nature of Islamic history, even while it systematically devalued the study of the more recent periods of North African history. For an E. F. Gautier, the murky periods (les siècles obscurs) included everything after the Arab conquest of North Africa. In this way civilizational thinking supported the dichotomous and racist categories of the colonial world.

Nineteenth-century French orientalists also shared with the students of Islamic societies the romantic quest for the exotic, for dépaysment, perhaps ultimately for the autonomous individual. Ethnographers, orientalists, painters, and writers-people like Léon Roches, Eugène Daumas, Émile Masqueray, Eugène Delacroix, Gustave Flaubert, and Charles de Foucauld-were propelled from their secure moorings in metropolitan bourgeois society into an encounter with the other, the Arab, the Oriental. One of the most striking things about this imperialism of the spirit is the watchful impersonality that characterizes the closest personal encounters, a kind of voyeurism, an emotional ethnography of the other. Even in its most intimate and personal form, that of sexual relations, one finds a strange, distanced watchfulness. The heroic couplings of Gide at Biskra, no less than Flaubert's affair with Kuchuk, or those of Isabelle Eberhardt, or indeed any number of French lieutenants of the Arab Bureaux have this quality. Edward Said, who is our best guide to this encounter with the Orient, is therefore correct to insist on both its sexual nature, and its paradoxical abstraction of the other. He is also right to point to the manner in which this encounter sprang from a desire for power and domination, which is the very opposite of mutuality. This romanticism was most potent in the first half of the century, but it continued to be an important current of thought and feeling until well into our present century. The psychosexual dynamics of the encounter with oriental peoples are of central importance to any effort to understand the complex origins and nature of French orientalism and colonial ethnography.

Compelling as it is in certain respects, Said's analysis of orientalism is unsurprising. After all, what would one expect the colonial sociology of Islam to be, if not colonialist? Said's discussion of the Description is limited to a brief examination of its Preface historique, and fails to engage the essays in the État moderne that discuss contemporary Egyptian society of the period. In this way, as historian David Prochaska remarks, Said "reproduces the textual, philological, and linguistic tendencies which constitute the basis of the orientalist enterprise he is so critical of otherwise!" We might agree with Said that orientalism was a discursive system that created the Orient as a field of power in which France might intervene. But if our intention is to achieve a historical understanding of French colonial forms of knowledge, such a diagnosis is insufficient. French power relative to the rest of the world was neither evenly deployed nor always the same, and French forms of knowledge and understandings of colonial peoples were always changing as well in relationship to the power dynamic, but also in response to changes in France itself. Overtheorized and inadequately historicized, Said's Orientalism provides neither a history nor a sociology of the authors of the Description de l'Égypte.

If our goal is to develop a sociology of these sociologists, the work of Pierre Bourdieu is indispensable. He provides some useful concepts that can help us understand the ways in which to understand the evolution of orientalist knowledge (especially the fields of ethnology/sociology that are our primary concern here). If knowledge is produced in particular institutional contexts, Bourdieu suggests, it arises in specific political contexts as well. In order to better understand the ways in which the sociology of Islam developed, it is important to situate it in both its political and its scientific fields. If we perform this operation we must recognize that while the sociology of Islamic societies had an ambiguous relationship to French orientalism, it was frankly marginal to intellectual currents in France, an insignificant back eddy to the unrushing stream of French science. This is all the more true of its relationship to the emergent discipline of sociology. The major figures of French sociology, from Auguste Comte and Frédéric Le Play to Émile Durkheim, Marcel Mauss, and Claude Levi-Strauss were little interested in the Maghreb, or Islamic societies generally. Nor were their disciples. French ethnography of Islamic societies developed in a kind of intellectual ghetto, clearly subordinate to the metropolitan world, a little tradition against the greater French tradition. This explains why despite a vast ethnological literature, no significant contribution to general sociological theory can be found in French studies on North Africa.

It is in the examination of the relationship of the sociology of Islam to the French political field that we encounter the reasons for its distinctive shape. Although it was marginal to metropolitan science, the sociology of North Africa was much closer to the French political arena. Often, it was dominated (to its detriment) by political questions. Thus the study of the nature of the system of landholding in rural Algeria was from the outset a highly charged political question for French settlers, and the literature on arch land reflects this fact. The bitter struggle for the control of research on Morocco at the beginning of the twentieth century is another reflection of this as we shall see in chapter 4. All of this is to say that as a colonial social science, the sociology of North Africa was decisively shaped by the political and intellectual contexts in which it existed. This drastically weakened its achievements from the beginning, skewing its central paradigms, suggesting false questions, and inhibiting the asking of important ones. The French tradition of the sociology of Islam was thus from the outset significantly inflected by its colonial context.

Already by the 1850s, the central assumptions of French views about the nature of North African society can be observed in the work of such writers as Eugène Daumas, Louis Pein, Charles Richard, and Ismail Urbain. These assumptions included the supposedly anarchic state of precolonial society, the essentially negative and obscurantist role of Islam in North African society, the innate fanaticism of Islam, and the division of the society into dichotomous and mutually exclusive groupings: Arab and Berber, nomad and sedentary, rural and urban. From this it was but a small step to conclude that North Africans were congenitally incapable of independence. It goes without saying that these authors accepted the French civilizing mission and French rule as legitimate. The stereotypes of the colonial gospel provide the connecting thread between the liberal humanitarianism of the Saint-Simonians, the vae victis of the settlers, and the positivist social Darwinism of late nineteenth-century ethnologists. That such a set of assumptions had direct political relevance is evident. That it also helped to condition the questions that were asked and not asked, and the answers that were given, is no less true and demonstrable.

Finally, if we consider the characteristic emphasis of French ethnologues on certain topics (and not others), we may assert that we are dealing with a tradition. It is the gaps and silences in the French ethnographic record, as much as its preferred topics, that outline its distinctive shape. The two topics that most interested students of Algerian society were the principal manifestations of Islam in the countryside (the ulama, and other urban religious institutions more generally, were of little concern), and the structure and role of the tribes in Algerian society. Most of the ethnological literature on Algeria focuses on these topics, including the manners and customs variety of reportage, but also a great many more ambitious treatises. To the two major themes of study contributed by the Algerian literature, the ethnology of Morocco added two more: the study of cities, and the study of the relationship between the tribes (notably the "Lords of the Atlas") and the state administration (the makhzan). This last topic was the central motif in Robert Montagne's work on the Berbers of the central High Atlas. But it also informs many lesser efforts. The French sociology of the Maghreb remained fixated on these questions until the eve of decolonization.

At the same time some important questions were never raised, and this helps to define the tradition further. There were no studies of contemporary social change, for example (Robert Montagne's work on the bidonvilles of Casablanca dates from only 1951), and the political economy of both the precolonial and the colonial Maghreb was neglected. Other important but absent topics were segmentary lineage structures (a British monopoly), the ulama (what some have called ulamology is a postcolonial phenomenon), the Islamic legal system as it actually operated (as opposed to how orientalists said it should), the roots of resistance (despite numerous insurrections, the topic surprisingly failed to attract scholarly interest), and finally, economic ties between city and country (French sociologists tended to presume economic autarchy). According to the French colonial gospel North Africa lacked a history, as a result of which it was condemned to repeat the same meaningless gestures. A veritable island, the Maghreb was viewed as having no cultural and political links to the outside world. France was the sole source of progress; resistance would bring only disaster.

The French tradition of the sociology of Islam went through four major phases, which may be described as the Egyptian phase (1798-1828), the Algerian phase (1830-70), the West African phase (1880s-1910), and the Moroccan phase (1900-1930). To these four some might add a fifth phase for the Syrian and Lebanese mandates. Of these, by far the most important was the Algerian phase. Inevitably, just as the Indian colonial experience tended to shape the perceptions of British orientalists and anthropologists, so the Algerian experience marked the French image of the world of Islam. Most of the leading French scholars of Islamic societies were formed in Algeria, and even those who were not tended to define themselves negatively with reference to the work of the French Algerian experts. One could do an interesting study, for example, of the efforts of French Africanists to free themselves from the influence of the Algerian paradigm. As we will see, the development of the sociology of Morocco was powerfully influenced by the Algerian model. It was through an Algerian lens that Frenchmen viewed other Islamic societies.

Yet the Egyptian experience did not count for nothing. For one thing, as Edward Said has argued in Orientalism, the authors of the Description created a textual Orient in their own image, where the Islamic Orient "would appear as a category denoting the orientalist's power and not the Islamic people as humans or their history as history." The European relationship with the Orient inseparably linked the desire for knowledge of the other and the desire for domination over the other. As a result of the Napoleonic expedition, orientalism as a discourse on the Islamic other was born. The Description de l'Égypte gave birth to a number of subsequent textual children-later large-scale French projects aimed at mapping out the geographic, geological, scientific, and ethnographic terrain on which French imperialism would operate in Greece, Algeria, and Morocco. They include the five-volume Expédition scientifique de Morée (1832-36), the thirty-seven-volume Exploration scientifique de l'Algérie (1844-67), and the publications of the Mission scientifique du Maroc (1903-25) that are the focus of our attention in this book.

The French Tradition of the Sociology of Islam

A distinguishing feature of at least some of the essays on contemporary Egyptian society in the Description de l'Égypte is the care they take to provide precise social documentation. Nothing in French colonial ethnography on Algeria, French West Africa, or Morocco comes close to the effort at statistical documentation of the material life of ordinary Egyptians that is found in the Egyptian essays. Nowhere else is there a major effort to map a particular city street by street, including workshops, mosques, baths, and other public buildings. On the other hand, the authors of the Description show little interest in popular Islam as a major theme (especially Sufism) or in orientalist production on the Maghreb. Consequently the Napoleonic forces were shocked when they encountered popular resistance from Sufi groups. Why these differences? Who in fact were the authors of the essays on contemporary Egypt in the Description? Mostly, they were recent graduates of the École polytechnique, a central revolutionary institution. Disciplined, scientifically trained observers, they were indoctrinated to believe that national regeneration depended on their putting their training to practical use. The effort to gather social statistics and to map human populations in the Description project derives from the same Enlightenment instincts as the desire to inventory the natural resources and appropriate the ancient history of Egypt. Here we encounter a more complex and interesting genealogy than that provided by Said.

The thirty-seven volumes of the Exploration scientifique de l'Algérie (1844-67) edited by Edmond Pelissier de Reynaud represent a conscious imitation of this early fruit of French colonial science. Drawn from the same Enlightenment spirit that had informed the Description de l'Égypte project, the Exploration scientifique sought to compile a thorough inventory of Algerian society. As we'll see later on, a similar inspiration lay behind the construction of the Moroccan colonial archive in the period 1890-1925. It is also intriguing to observe how rapidly the process went. In all three of these cases as well, the job was done within the span of a single generation. The colonial administrators and officials who came afterward for the most part confined themselves to further elaborations of the central themes laid down at this time.

In retrospect, examined in historical perspective, the French tradition of the sociology of Islam consisted of three main strands, the complex patterning of whose interactions over a century and a half constituted the tradition. These three strands represent the contributions of three groups: the Arab Bureaux, the civilian amateurs, and the academics. Attached to real social forces with real interests and perceptions of the society, these three groups are of primary importance in understanding not only the contributions to the intellectual tradition of colonial sociology, but also much of the dynamics of French colonial politics. Inevitably, of course, it is important to recognize that there was considerable overlap in the contributions of the groups.

The Arab Bureaux, which began in 1844 with the appointment of Eugène Daumas (1802-71) as head of a Direction of Arab Affairs, contributed the most important strand. From these administrators, or Robinsons galonnés, as Jacques Berque has called them, came a disproportionate share of the canonical works on Algerian society, customs, and religion. Closely involved with the life of the tribes, the native affairs officers of the Arab Bureaux knew from direct observation and experience what the civilians and the academics seldom grasped: the attractiveness of Muslim society, its endless capacities for resistance, its subtleties as well as its vulnerabilities. The social and political consciousness of the officers of the Arab Bureaux, already stimulated by its propinquity to the lives of those they administrated, was also enhanced by virtue of the fact that many of them were graduates of the prestigious École polytechnique. A number were also Saint-Simonians, members of a utopian socialist movement. Their revolutionary commitments and sensitivity to cultural difference gave the analysis of this group a depth lacking in the work of other French observers. At the same time, their participation in the military conquest failed to prevent at least some of them from holding racist attitudes toward Muslim Algerians.

Both the strengths and the weaknesses of the Arab Bureaux as a colonial archive of knowledge about Algerian society derived from its alliance with the tribes. The culture of tribal society, no less than its "moral topography," was first limned by the men of the Arab Bureaux. The model of tribal structure first worked out by Eugène Daumas, Thomas Pein, Charles Richard, and Thomas Ismail Urbain was based on the genealogical structure of the patriarchal family. Techniques of oral investigation, including careful cross-questioning of local informants, made possible Daumas's study of the tribes of the Sahara and Ernest Carette's of those of Kabylia. The application of these methods was later to produce such works as Émile Laoust's Mots et choses berbères and Auguste Mouliéras's study of the Rif.

Another early topic of study was popular Islam, especially the role of the Sufi turuq. These were widely believed to be deeply involved in Algerian resistance to the French presence. Charles Richard's study of the popular religious roots of the Algerian resistance effort, Étude sur l'insurrection du Dahra (1845-46), sought to explore the history and content of popular Islamic millenarianism. While Richard, like many other French observers of the period, viewed Islam through the lens of the religious struggles of the French Revolution, he also saw Muslim popular religious beliefs as a source of spiritual strength and political resistance. François de Neveu's Les Khouan was the first inquiry based on responses to an administrative questionnaire that sought to systematize information on Algerian membership in Sufi brotherhoods. The political role of the marabouts (as the leaders of rural saintly lineages were called) was a related topic that stimulated much research. Here the major work was Louis Rinn's Marabouts et Khouans. However, following the suppression of the last major resistance movement in 1871, interest in the political power of Islam essentially disappeared. Thereafter, studies of popular religion focused on rural marabouts, saintly lineages, and sufi orders, generally viewed through the lens of an exoticizing orientalism.

The rule in general appears to be the following: the farther from the blood and thunder of military conquest, the less relevant and less reliable the ethnography. Why should this have been the case? A number of reasons come to mind. The first generation had the advantage of being first, and of everything thus being new, fresh, and interesting. Moreover, insofar as many of the early observers were military men or directly tied to the colonial enterprise, they had a vested interest in understanding the society in all its specificity, and in being able to distinguish its chief components. Put simply, lives might depend on it. Later on, as bureaucratic routine took over, officers had only to update the reports of their predecessors. They became increasingly overwhelmed by paperwork, and spent less time in the marketplace or on horseback frequenting their charges. As time went on, the military advantage shifted decisively toward the Europeans. The ability to deploy powerful new military technologies rendered moot errors in political analysis. Confronted with the machine gun, tribal forces, however brave and resourceful, stood no chance. For these reasons, therefore, the early periods were formative.

Accordingly, if we were to plot on a map the successive thrusts of French military advance in North Africa, we would discover a close correspondence with the advance of knowledge about the society. By the end of the Second Empire the ethnographic inventory of Algeria was virtually completed; only the Sahara remained to be explored. For Tunisia, we have good ethnographies only for the southern regions, as only there was a tribal military administration on the Algerian model established. The historical development of French knowledge about Morocco adhered to much the same outline: first the populations of the coastal districts and the central plain, then the cities of the interior, and only later on, the tribes of the Rif, Middle and High Atlas, the Sous, and the Saharan steppe. In Morocco, however, the populations of the Berber areas and the Saharan zones were for the most part the monopoly of the military administrators of the Arab Bureaux, while French civilian researchers (many of them academics) worked on the cities and the plains.

A second group, the civilian amateurs, made contributions to the elaboration of the sociology of Algeria. Men like Camille Sabatier and Henri Duveyrier (1840-92), however, possessed neither the motivation for research nor direct access to the Muslim populations. Therefore their intellectual contribution to the ethnography of Algeria was the weakest of the three groups of researchers. However, if their contribution is assessed in political terms, it emerges as fundamental. Even the few civilians who could speak with real authority on Algerian matters tended primarily to advance views of Muslim society that reflected the interests of French settlers. Indeed, it is this that makes them important. The bitter and unrelenting hostility of French settlers to the Arab Bureaux and the latter's support of the great feudal Arab chiefs provided one of the leitmotifs of French settler politics, and seriously influenced the course of Algerian colonial history. For individuals like Dr. August Warnier, the men of the Arab Bureaux were the real enemies of France in Algeria. Lording it over the tribes, they lived the life of pashas; their corruption was legendary, and so was their duplicity. The aristocratic origins and royalist sympathies of many of the officers were repugnant to the settlers, who loudly proclaimed their ardent republicanism. Although there can be no doubt that the settlers greatly exaggerated, there was some substance to their charges. Many of the officers were of aristocratic origin; some were royalists. Others, Saint-Simonians for the most part, were sympathetic with many features of Algerian society, and adopted a paternalistic attitude toward the Muslim populations. Some conducted social experiments on the tribes entrusted to their administration. The main reason for the hostility of the settlers, however, was the land question ( in this respect Algerian history resembles that of the American West). The Arab Bureaux constituted the chief obstacle to the extension of colonization, civilian rule, and settler domination. So long as a region was under military government, settlers had a difficult time acquiring land there.

To be sure, there were instances of dialogue between these two strands of the French tradition. One such was the collaborative work of Adolphe Hanoteau and Aristide Letourneux, La Kabylie et les coutumes kabyles (1872-73). (Hanoteau was chef de bureau at Fort National, and Letourneux was an Algiers lawyer.) It was Letourneux who, with his barrister's eye for system and order, processed what had been local oral customary practiceinto a kind of Berber administrative handbook (the qanun-s), a useful instrument for combating the supposed nefarious influence of the sharia.The influence of La Kabylie et les coutumes kabyles extended even to Morocco, where it helped to inspire the Berber policy of Lyautey and his successors. Later the same politique des races was applied in Syria and Lebanon with regard to the Druse and Alawi minorities during the mandate period, with equally unfortunate results. In general it would seem that the chief result of efforts at collaboration between military men and civilians was to accentuate the tendency already inherent in the tradition of the Arab Bureaux to turn local customary practices into fixed and rigid principles.

The intensification of the debate between settler interests and the Arab Bureaux, the chief protectors of the Muslim populations, led to the increased politicization of French ethnography. The decree of 1863 (a misguided effort at halting speculation in tribal lands) provoked a new interest in Arab society for a time. However, following the collapse of the Second Empire after the Prussian invasion in 1870 and suppression of the Moqrani rebellion (1870-71), French policy perhaps inevitably took on a sharply antimilitary coloration. The ensuing settler backlash led to the dismantling of the Arab Bureaux (which were permitted to continue only in the Saharan region), and the enactment of a series of punitive regulations known collectively as the code de l'indigénat. The years that followed, 1870-1900, were disastrous for Algerian Muslims. And they were equally disastrous for the sociology of Algeria. From a quasi-autonomous intellectual by-product of the Arab Bureaux, the ethnography of Algeria became increasingly dominated by the discourse of French colonial politics. No longer a serious threat, Muslims did not have to be taken seriously. As a result, there was little incentive to study them.

The Algerian Colonial Gospel

This development found direct expression in the crystallization of the Algerian colonial gospel, a patterned set of stereotypes about the nature of Muslim Algerian society, especially the so-called Kabyle myth, a set of stereotypes on the supposed differences between Arabs and Berbers. Elements of the Kabyle myth can be traced in the writings of Abbé Raynal and precolonial French travelers. What is new about its post-1871 manifestations is the effort, clearly influenced by the racialism of late nineteenth-century social theories, to construct a systematic policy on the basis of these differences. The Berberophilia of the Kabyle myth became fully elaborated in the post-1871 period. The Berber-speaking Kabyle people were believed to be potentially assimilable to French civilization by virtue of the supposed democratic nature of their society, their superficial Islamicization, and the higher status of Kabyle women. Some officers, like Baron Aucapitaine (1832-67), jumped to the conclusion, "In one hundred years, the Kabyles will be French!" For a time a policy of cultural assimilation of the Kabyles was attempted by the French government, but given the uneven results it was soon abandoned (though a few schools were permitted to continue). The most extreme among the settler theoreticians of Berber separatism was Camille Sabatier, whose intellectual nullity was matched only by the extremes of his race-conscious thought, and the credulity of the French Algerian public, whose clamorous support gave Sabatier an influence his ideas were far from deserving. Sabatier observed:

The unknown Lycurgus who dictated Kabyle qanuns [customary law]was not of the family of Mohamed and Moses but of that of Montesquieu and Condorcet. Still more than the skull of these Kabyle mountaineers, this work bears the seal of our race!

The persistence of the Kabyle myth was one of the most enduring aspects of the French sociology of Islam. It, together with other elements of the Algerian colonial gospel, exercised a particularly unfortunate impact on French writings on Algeria, particularly in the period leading up to the First World War.

French academics, the third group that contributed to the elaboration of the sociology of Islam, were involved in the study of Algeria from the time of the conquest. But it was not until after 1870 that they emerged as a distinct group. Their new prominence was directly tied to the emergence of the social sciences in France and the reorganization of French higher education after the fall of the Second Empire. And it was Émile Masqueray who, more than any other, endowed the academic study of Algerian society with prestige and legitimacy. As the secretary of Victor Cousin and a brilliant young graduate of the École normale supérieur, Masqueray was well connected in governmental circles and destined to have an important career. His Formation des cités chez les populations sédentaires de l'Algérie (1886) shows the influence of the Berberophilia of the time-as well as that of Fustel de Coulanges (whose Ancient City was then enjoying a major vogue). Masqueray was at the center of the intellectual currents of his time, rather than on the fringes, like the other Algerian academics. Thus he was able to transcend, even if only partially, the crippling effects of the politicization of colonial sociology, and was entrusted with founding the École d'Alger (from which the Université d'Alger would eventually emerge). Masqueray's doctoral thesis differed sharply from the Algerian tradition in its bold application of the comparative method, and its rigorous attention to the verification of evidence and the testing of hypotheses. It is no surprise that he had no imitators.

Toward the end of the nineteenth century the focus of the French sociology of Islam began to shift away from Algeria. The Muslim societies of French West Africa were the next to attract the attention of French researchers. The study of what would later be called Islam noir began in earnest in 1880s, when French native affairs specialists, both military and civilian, turned their attention to the role of the Sufi turuq in the West African Sudanic belt from the Senegal to Timbuktu. Since most French West African specialists had previously served in Algeria, they brought with them a concern with the political role of the Sufi brotherhoods and the marabouts. Thus began an abiding concern with West African popular Islam.

In the superheated atmosphere of racism and chauvinism of fin de siècle France, French officials in Algeria became obsessed with the fear of Pan-Islamic conspiracies, particularly that of the Sanusiya brotherhood (this despite the fact that there was little evidence to support their concern). Unfortunately, this popular psychosis was not limited to colonial Algeria. A preoccupation with the Sanusiya also marked French writing on Islam noir in the pre-1914 period. If we look a little closer it is difficult not to see that French fears of popular Islamic uprisings derived from anxieties generated by the church/state struggle in France of the period (then in a phase of heightened tension). Concerns about Algeria are also reflected in the effort after 1900 to develop a native policy for Afrique occidental française (A. O. F.) aimed at blocking the extension of Islamic law and fostering the spread of the French language. Here we see the influence of the Kabyle myth, which had prompted French efforts in Kabylia to prevent Arabization and Islamization by building schools and law courts employing the French civil code or Kabyle qanun law.

No discussion of French studies of West African Islam would be complete without mention of the role of Paul Marty (1882-1938). Born in Algeria, Marty served in A. O. F., Morocco, and Tunisia in the course of his career. One of the last of the military ethnographers formed in the Algerian native affairs tradition, Marty's writings on Islam noir constitute an important (and sometimes the only) source on West African Islam in the period 1913-25. His sympathy for the plight of Muslim elites caught in the undertow of history, and his conviction that French colonialism was the vector of progress, made him a sensitive observer of African society. Marty had more than a dozen monographs to his credit, most notably "Les mourides d'Amadou Bamba." His extraordinary productivity (based on the dissemination of questionnaires to local officials throughout A. O. F.) stands alone in the annals of the French sociology of Islam.

French Knowledge of Morocco, 1880-1900

In 1900, on the eve of the diplomatic crisis known as the Moroccan question, France possessed little accurate ethnographic knowledge of Morocco other than of the coastal cities and the tribes along the Tangier-Fez "road of the ambassadors." A case could be made that British studies of Morocco at the time were superior. The absence of documentation is surprising, given the political stakes. Yet apart from a few classic works, French policy circles had remarkably little reliable information on Moroccan society. As late as 1900, British studies of Morocco were still clearly superior. Using the pseudonym Eugène Aubin, Eugène Descos described his predicament in 1901 as he sat down to write Le Maroc d'aujourd'hui:

I naturally took advantage of my long hours of solitude to read most of the works published on Morocco. Aside from historical facts, I scarcely found anything to exploit; for nothing exists, to my knowledge, no book, written in any language, which lays out, for those who might be interested, the mechanism of Moroccan life and the Moroccan government.

Most of what the French knew about Morocco derived from the published and unpublished reports and maps prepared by the members of the French military mission. From 1884 until 1912, French military instructors served the Moroccan government and were charged with training the Moroccan army, much as French instructors had earlier helped train a modern Ottoman army. In practice, the chief purpose of the military mission was spying, which in this case meant gathering the political gossip of the court. Commissioned reports (both published and unpublished) focused more on military and political intelligence than on the main features of Moroccan society. The Berber-speaking groups of the Rif and Atlas Mountains-some 40 percent of the total population-remained virtually unstudied. Topographic maps prepared by the military mission were often of high quality and played a significant role in informing French military planners about the terrain on which they might one day be required to act. Both the official reports and maps were prepared for a restricted audience primarily composed of staff officers charged with preparing and updating contingency plans with regard to Morocco. They did not circulate outside the Ministry of War and played little role in shaping French policy toward Morocco at this time.

Before 1900, France had two additional sources of information about Morocco: confidential reports prepared by the diplomatic corps, which did not circulate outside of Quai d'Orsay circles, and studies by individuals (of which there was a great paucity). However, looming over all these sources is Charles de Foucauld's Reconnaissance au Maroc: Journal de route. Incontestably the most important work on Moroccan society published before 1900, it went through numerous editions and served as the vade mecum of several generations of French Morocco hands. At a time when few Europeans dared travel by themselves in the countryside without adopting Moroccan garb, Foucauld decided to travel in disguise, surprisingly opting to dress as a Moroccan Jew and staying only among Jews. In this way he was able to pass almost unnoticed. The publication of Reconnaissance au Maroc immediately propelled its author to prominence in colonial circles.

The book itself became an instant classic of the ethnography of Morocco. However, it was an ethnography with a difference, for despite a few brief descriptions of social rituals or of the operation of social institutions, it contains few passages of any length that can be described as ethnographic. Instead, its primary focus is to provide as meticulous a description of the topography of the terrain traversed as possible. In this sense, it is the author's "cartographic eye" (the emphasis being on cartography more than ethnography) that is its most characteristic feature. As a result, subsequent travelers were able precisely to identify each point mentioned and at every moment to locate themselves in the Moroccan landscape. The French military relied heavily on Reconnaissance au Maroc in the conquest of Morocco in the first part of the twentieth century, and carried the work everywhere it went.

Foucauld's social observations of Berber-speaking groups in Morocco broke in important ways with the verities of the Kabyle myth. Reconnaissance au Maroc contains few of the racist strictures about the alleged differences between Berbers and Arabs that tainted French ethnographic production in this period. Such distinctions, Foucauld believed, should be understood as linguistic rather than racial in character. Why, we might ask, was Reconnaissance au Maroc able to transcend the powerful biases of its time? We might begin by locating the book and its author in the French political and intellectual fields. At the time of its publication, Foucauld himself was unknown. While his aristocratic background gave him a certain cachet, he was marginal to the debates that roiled the French political field. Despite the fact that Reconnaissance au Maroc was explicitly written to facilitate the French conquest of Morocco, it does not conform to many of the then current stereotypes about Moroccan society or Islam. Morocco was not on the active political agenda in either Algeria or Paris in the 1880s, and thus Reconnaissance au Maroc was able to transcend the deforming lens of the political field, at least to a degree. Indeed, the success of the project reflects its dual marginality to both the French intellectual and political fields.

The French government was not the only supplier of information about Morocco, however. The Algerian colonial government (the Gouvernement général de l'Algérie) also had a strong interest in "doing Morocco." It was the principal source of proven experts on North African Muslim societies. The next chapter addresses the Algerian contribution to the ethnography of Morocco.