Collaboration and Empire in the Middle East and North Africa: Introduction and Response1

In 1801, a great rumpus erupted over the untoward behavior of the British resident to the Court of Hyderabad, James Achilles Kirkpatrick. Rumors in Calcutta claimed that Kirkpatrick appeared in public dressed in Indian Muslim garb, stained his hands red with henna in the manner of a Mughal nobleman, and composed Urdu poetry. Worse still, Kirkpatrick had somehow established a liaison dangereuse with a fourteen-year-old girl from one of Hyderabad's Sharifian families (i.e., of the Prophet Muhammad's lineage). The girl in question, Khair-un Nissa, had been raised in a harem and kept in strictest purdah; yet, the indefatigable Scotsman had managed to impregnate the Sayyida. Still more rumors maintained that Kirkpatrick had embraced Islam and married the girl. British officials feared that Kirkpatrick could no longer be counted upon to furnish accurate information about Hyderabadi politics. Marriage and conversion would transform him into a double agent, at best; or at worst, someone now on the "other" side; or so British officials reasoned.

Cases such as Kirkpatrick's were not uncommon in the late eighteenth and early nineteenth centuries in India. By revisiting India Office and other records, both C. A. Bayly in his Empire and Information and William Dalrymple in White Mughals have chronicled a number of instances of "going native," leading to a greater frequency of mixed marriages between British men and local women, and a larger number of mixed race children, than previously thought.² Needless to say, until very recently, stories about the crossing of cultural and/or sexual boundaries and, therefore, political frontiers—had been largely excised from the historical narrative, and above all from the collective memories of the concerned British families and officials.

Compared to the Gulf native agents, Kirkpatrick's cultural trajectory went in the opposite direction; yet, the Scotsman's case speaks directly to several critical points made by James Onley in his paper, "British Agents in Arabia and Persia, c. 1758 to 1958," as well as to our panel which revisits Ronald Robinson's notion of collaboration advanced nearly three decades ago.³ First, by examining Britain's presence in the Gulf from the perspective of local informants and civil servants, Onley undermines the conventional narrative about the nature of British rule.The reliance upon indigenous merchants, whose stories have gone largely untold, was much more prevalent and fundamental to Britain's rule than was formerly acknowledged—in the same way that more British officials became integrated into the social and sexual structures of the countries they worked in than has been previously admitted. Thus, Onley's research represents a welcome corrective to the work of older scholars of Gulf history such as J. B. Kelly.⁴

Onley's second argument is that British residents, ships, and merchants operating in the Gulf had gradually shoved their way into an exceedingly complex, long-standing, and well-established system of trade, political relations, and information-gathering stretching from southwest Asia through the Indian Ocean. To succeed at all, the British had to play by rules developed over centuries by Persian, Arab, and South Asian players who operated in a commercial world economy underwritten by Islamic laws and customs and linked by networks of exceedingly diverse merchant communities. Thus, the handful of British residents established in the Gulf had to, in a sense, "go native"--if not culturally, as James Kirkpatrick had done, then at least in terms of engaging local merchants to transcribe the social field of communication—which deeply informed politics and commerce. The "information order" of the Gulf could only be understood by merchant-dynasties, such as the Safar family, acting as purveyors of intelligence between Gulf shaykhdoms and British officials; the relationship was one of mutual benefit based upon intersecting interests, although subject to inherent tensions, even dangers.

Professor Onley's fine analysis reveals the bonuses as well as the handicaps entailed in relying upon indigenous clans to provide critical services, particularly once again, in the realm of information-gathering. The native agent closely resembled his counterpart, the dubash [translator or informant], in his indispensable role, and also recalls the dragoman—that polyglot intermediary needed for the conduct of business between the Ottoman Porte and Europe in the same period. Indeed, by leaning upon Gulf trading families, the Crown was spared the expense of maintaining large numbers of Metropole officials who might desire to bring along families (which always complicated empire as lived on the ground). Of course, the opposite "threat" was that residencies too sparsely peopled by British expatriates would produce James Kirkpatricks. Onley's research is part of a larger scholarly effort not only to re-interpret modern European empires but also to reconfigure histories of the Afro-Eurasian ecumene, particularly the regions making up the Mediterranean-Indian Ocean system.⁵ Recent works by Jerry Brotton, Nabil Matar, and Patricia Risso, to name only a few, argue that inter-dependence, interaction, and exchange characterized relations between the Islamic and European worlds as much as did conflict or violence.⁶

My questions for Onley are as follows. First, I wonder if the merchant-brokers did not also function as double agents; in other words, they not only accumulated news and intelligence for their British patrons but also surely furnished local Gulf shaykhs and others with valuable information regarding what the Europeans were up to. Indeed, would these native agents have been able to conserve their pivotal middle-men position in Gulf society had they not furnished something to their other patrons--Gulf elites? In any case, to paraphrase Bayly, fields of social communication by definition have always been dense, flexible, and multi-directional.

Second, one wonders if the imposition in the 1890s of an additional bureaucratic layer of British political officers upon the older system of native informants paradoxically impaired access to information, or at least restricted the flow of certain kinds of information previously available in relative abundance. This pattern is seen in late nineteenth-century Algeria when the growth of civilian bureaucracy, along with the dismantling of the Bureaux Arabes which had depended upon indigenous collaborators, ironically decreased French knowledge of native society. Finally, what impact did rapidly accelerating changes in transportation have upon this culture of surveillance and intelligence gathering? If the London to Bombay mail had been reduced to one month's time by 1845, then what did the 1869 opening of the Suez Canal mean for British residents in the Gulf trying to accumulate and, more importantly, interpret the ever greater volume of news disseminated by growing numbers of travelers, pilgrims, and traders?

Moroccan soldiers' service to France is the topic of Professor Gershovich's paper. This case would seem to be much closer to the model of collaboration under consideration here, whereas the Gulf agents appear to fit more closely the notion of "accommodation" as first enunciated by Peter von Sivers.⁷ Gershovich's research points to a significant difference between French colonialism as practiced in Algeria, where military conscription was established in 1907-1908, and Morocco, where compulsory military service for native men was never attempted. The Algerian experience had served as a counter-model, a template for what not to do in achieving imperial governance. Resistance by Muslims to forced conscription in Algeria, for example, during World War I, had proven too costly; the huge losses of French nationals in the Algerian colonial army had caused an enormous outcry in the Metropole throughout the nineteenth century. In contrast, Moroccan soldiers were contracted as paid "volunteers" or "perfect mercenaries" for combat in Europe or in other French possessions world-wide. For subduing Moroccan lands and peoples, native troops from specific tribal groups were also recruited. Thus, as hired guns from the most disadvantaged regions, they conveniently deflected two sources of opposition to conscription—one in the Metropole, another from native society. What forces impelled Middle Atlas Berbers to fight under the tricolore? As was true for the Gulf agents, economic motives were fundamental--but there were essential differences as well. The Berber recruits constituted what might be termed "subsistence collaborators," while the native agents were mainly well-heeled merchants doing what they had always done best in the Gulf region and beyond.

I have three questions for the second paper. The first returns to my observation regarding Algeria as imperial model and counter-model for much of French colonial policies elsewhere. In the early conquest period, French military officialdom imitated their Turkish predecessors by hiring tribal groups, such as the Zawawa or Zouaves, which had traditionally served the Turks before them. Thus, I wonder about the recruitment of Middle Atlas Berbers by the French army. Did this practice evolve from an earlier military tradition followed by Moroccan sultans? My second query again evokes Bayly, who noted that in the period before modern nationalism, native agents or informants serving European masters cannot necessarily be categorized as "traitors." "In fact, it was from the descendants of the informants that many of the future nationalists would be drawn."⁸ Is it possible that the recruits from the Middle Atlas did not see themselves as necessarily Moroccans when recruited, and thus, should not be interpreted as waging warfare against "fellow Moroccans"? Directly related to this, my third question is: how does contemporary Moroccan society regard veterans of the French army, barely subsisting on their military pensions from Paris--as former colonial collaborators, as less than loyal citizens?⁹ And did some of the demobilized soldiers in independent Morocco find employment in the sector of state-organized coercion, such as the gendarmerie, the rural police, or as prison guards? In any case, as was true of Onley's study, Gershovich's research raises questions about the staying power, or universality, of the Robinson theory of collaboration since the "European inputs"--military power--in this case were overwhelmingly Moroccan.

In our third paper, collaboration also comes to mean something akin to von Sivers's formulation of accommodation used in his study of nineteenth-century Algerian resistance movements. James McDougall traces one among many possible paths of accommodation—the appropriation of French educational and association forms—which gave birth to the culturally hybrid "Progress Club." While this Club might be seen as "marginal," in fact, it occupied a central geographic location adjacent to the Place du Gouvernement in Algiers, a highly visible, symbolic site. The Club also represented what one wing of colonial and Metropole thinkers, a minority to be sure, had long advocated as the very essence of the civilizing mission: a fusion between modern Islam and French universal principles. As director of the Algiers Progress Club, Shaykh 'Uqbi interpreted his mandate of "auto-emancipation" as furthering France's civilizing project through educational innovation as well as sustained intellectual contacts with "enlightened" members of Algerian society. Here, he was referring to Indigènophile Europeans, as opposed to the much more numerous Indigènophobes or Islamophobes. Fusing a social welfare bureau, the Khayriyya, reminiscent of the Sociétés de Prévoyance in France, with theolder madrasa shabiba [youth school], the Shaykh had reconstituted - and in the colonial regime's most evocative imperial space—the ancient sufi complex or zawiya, offering multiple communal, social, and instructional services. What could be more subversive than a liberal, "multicultural," and apparently apolitical Muslim Algerian? Assimilation and the civilizing mission threatened to become a reality. Small wonder that the Office of Indigenous Affairs probably had the mufti of Algiers assassinated in 1936, and then implicated Sidi 'Uqbi in this nefarious act. As elsewhere in French North Africa by this period, what colonial governments in Tunis and Algiers and colonial lobbies in Paris wanted most were rejectionist, militant nationalists--not integrationists like Shaykh 'Uqbi.

McDougall's research offers an "inverse" or "inverted" history to Robinson's theory regarding native mediators and nationalism; it also raises questions about Robinson's assertion that a large European settler presence "strangled native politics." Sidi 'Uqbi's repeated insistence upon the apolitical nature of his educational activities suggests that he might have quietly believed otherwise—Shaykh Ben Badis and the Reformed Ulama also continually proclaimed the non-political nature of their religious schools. By the inter-war period, the classroom and schoolyard in the Maghreb and elsewhere had become major battlegrounds for competing social ideologies and political platforms. For many settlers, modern pedagogy disseminated to the colonized was regarded as an inherently political act endangering empire. As early as 1912, La Tunisie Française, a right-wing settler newspaper, observed that:"à quoi bon alors avoir construit cette école [for native children] et dépensé 150.000 francs environ qui auraient été plus utiles ailleurs" ["what was the good of building this school (for native children) and spending 150,000 francs which could have been better used elsewhere"]. Other issues of this same paper often opined that "... les pires ennemis de la France étaient les jeunes bourgeois tunisiens éduqués dans les écoles françaises ou franco-arabes" ["France's worst enemies come from the ranks of young, bourgeois Tunisians who have been educated in French schools or Franco-Arab schools"].¹⁰

Moreover, McDougall's study demonstrates that—below the surface of the monolithic categories structuring the narrative of Algerian history until now--a wide-ranging number of adjustments to, and accommodations with, the colonial regime existed and indeed flourished right up to the end. Thus, his work makes us rethink the seductive thesis first advanced by Jacques Berque in 1962 of Islam as a "refuge" for the beleaguered Algerian society.¹¹ Along these lines, 'Sidi 'Uqbi could be conceived of as the Farhat Abbas of one wing or tendency of the Islamic establishments—by no means monolithic; or perhaps he was a sort of James Kirkpatrick, crossing over, or at least muddying the increasingly rigid boundaries of imperial identities and hierarchies.

My questions for the third paper arise from McDougall's mention of local police informers—the guys who probably ratted on Sidi 'Uqbi when he gave public lectures at the Cercle du Progrès. Here, we seem to be in the shady company of the collaborator as understood by scholars working in the field of social network analysis: "...recruited from the community on which they prey, only to be banished from it without being admitted into the society of their political masters."¹² Surely, the shaykh knew that informants were present in his audience when he gave his public lectures, thus, the use of what has been termed the "double discourse." Sidi 'Uqbi brings to mind Shaykh Muhammad and his daughter, Lalla Zaynab, of the Rahmaniyya Sufi Center at al-Hamil, both of whom praised France's just rule when communicating with French officers, even as they labored to safeguard Algerian religious and moral identity.¹³ In any case, declaring one's position to be apolitical represents a political stance and statement in and of itself.

Returning to the panel's theme, Robinson's theory of collaboration as first elucidated exactly thirty years ago, these papers demonstrate the enormous diversity that collaboration—or accommodation, or just getting along, or doing what one had always done—assumed in the British and French Empires: from providing information, to serving under the colonial flag, to attempting to carve out an autonomous space in the realm of association, culture, and education. McDougall, in particular, has raised a number of questions regarding the applicability of Robinson's theory to a settler colony such as Algeria—which is ironic since Algeria is so frequently taken as the quintessential colonial state. How then should we assess Robinson three decades later, particularly in the light of the recent flood—almost a glut—of research on empire? On the negative side, our papers demonstrate clearly that Robinson's formulation has a rather low tolerance for ambiguity or nuance and thus ignores the in-between spaces of colonial encounters—precisely the sites where many significant transactions took place. On the positive side, by tracing his intellectual genealogy and subsequent scholarship inspired by his theories, we may find that Robinson's emphasis upon indigenous actors in empire opened the way for theories of agency.

On a more general level, as examples of the latest and best scholarship on empire, these papers help dispel a pernicious, if enduring, historical myth. This myth is informed by the notion of an age-old Christian-Muslim frontier stretching from the Mediterranean and Red Sea-Gulf region across the Indian Ocean, a frontier allegedly shaped by implacable hatreds and impervious to all, or most, exchanges, with the exception of warfare and violence. Thus, our papers represent a timely antidote to Bernard Lewis's polemical writings which for decades have sought to discredit Muslim societies and states by portraying them as innately hostile to Europe, to things European, to Western civilization, particularly from the early modern and modern periods on.¹⁴ In its most recent incarnation, this myth has been transformed into, and pedaled as, the "Clash of Civilizations" thesis so fashionable in certain American imperial circles at present as the U. S.'s informal global empire gives way to formal empire.