From Algiers to Timbuktu: Multi-Local Research in Colonial History Across the Saharan Divide

Abstract The so-called “Saharan Divide” separating sub-Saharan Africa and North Africa into distinct fields has a long and complicated history. Paradoxically, given its dense historiography, this divide is particularly pronounced in scholarship on the colonial period. This article proposes an approach to researching across this division, centered on research in multiple African archives, to build a “multi-local” understanding of colonial-era trans-Saharan Africa. This approach is illustrated by the story of Algerians who taught in colonial schools in Mauritania and French Soudan, and by the author’s discovery of this story in sites across northwest Africa. This approach can help scholars reconceptualize multi-sited research and reevaluate the Area Studies divisions that continue to structure knowledge of African history.

Everything fell into place at the last minute. With my visa set to expire, I would have to leave Mauritania the next day. The day before, after several weeks in administrative limbo, I had been granted access to the Mauritanian national archives. I was unsure what to expect as I passed from the dusty streets of Nouakchott onto the greener grounds of the presidential palace. The archives are housed in a simple warehouse, out of the way. Upon entering the building, I was pleased to see that the single dossier I had been able to request had been located and set out for me. (Without access to any archival catalogues ahead of my visit, I had relied on Mohamed Said Ould Hamody's encyclopedic guide to Mauritania's scholarly resources to prepare. 1 ) The dossier was enormous, with hundreds of pages piled precariously on the table. Thanks to the generosity of the archivist, who passed me glasses of tea as I scanned the pages, I finished the stack of documents by the end of the day. My camera's memory card was full, and it would take weeks to sort through everything. But as I sat in the archive that day, I was struck by one document in particular: a report, from May 1939, by Boualem Ould Rouis, proposing reforms to the Médersa of Timbuktu. 2 Historians sometimes fantasize about finding the one document that uncovers a whole new history. This report did not exactly do that, but it did connect a lot of dots. I had encountered Ould Rouis's name a few months earlier, in the dusty attic of a high school in Algiers. There, I found a brief manuscript, titled Bilad Shinqit, that he had written in 1953 describing his impressions of Mauritania. 3 Because of that manuscript, I knew that he was one of a number of Algerians who had traveled to French West Africa to teach in médersas, colonial schools that combined Islamic and French curricula and forged new links between North and West Africa through Islamic education. I wanted to learn more about Ould Rouis and others who taught in both Algeria and West Africapeople who embodied this trans-Saharan colonial connection. But in Algeria, as elsewhere, it was difficult to find information about Ould Rouis and his peers. Official archivesthe National Archives and National Library in Algiers, and regional archives in multiple provincesall came up short when I went looking for documentation of the médersas and the medérsiens, as students and graduates of these schools were known. Of the three Algerian médersas, only the Médersa of Algiers has a relatively complete archive preserved within Algeria. 4 I stumbled upon that collection, preserved Africa in this era is striking. Colonial rule altered, but did not eliminate, trans-Saharan connections. Reconstituting a trans-Saharan history of the colonial period, as the example of Ould Rouis and the médersas demonstrates, necessitates a new approach to both historiographical frameworks and archival research to link North and West Africa into a unified space of analysis. As I circulated among archives in Algeria, Mauritania, Senegal, and France, I came to think of this methodological and interpretive approach not only as multi-sited, but as what might be called "multi-local." Trans-Saharan research requires deep engagement with multiple local contexts, multiple historiographies, and multiple approaches to the colonial archive. In what follows, I connect the West African careers of two Algerian teachers, Boualem Ould Rouis and his successor Abderrahmane Nekli, with my own research trajectory. Following in their footsteps illustrates how such a multi-local approach to institutions and individuals can help historians move beyond the colonial geographies that continue to shape the questions we ask and the stories we tell.

Trans-Saharan Archival Research
Although the historiography of Africa has long emphasized long-distance connections both within the African continent and through the worldwide African diaspora, a persistent gap in scholarship remains at the Sahara. This gap, which effaces the Sahara as a site of human history, exists for reasons that are both epistemological and practical. For academics and policymakers, North Africa, grafted onto the Middle East, is often considered in a separate geographical framework from "sub-Saharan" Africa. There are deep, complex epistemic roots to this division, often articulated in racial terms separating Black Africa from the Arab world, or, in more antiquated work, White (North) Africa. 8 The so-called Hamitic Hypothesis lent credence to racial thinking among scholars -Arab Muslim and European alikefor centuries, and allowed the Sahara to be construed as a racial barrier as well as a physical one, even among those who lived there. 9 During the colonial era, French scholarship adopted the terminology of Afrique noire and Afrique blanche; the Area Studies paradigm reinforced this division in academic departments, government agencies, and non-governmental organizations from the midtwentieth century to the present. The so-called "Saharan Divide" appears natural and timeless given these multiple roots.
Of course, the extensive networks that spanned the desert throughout history challenge this framework. The best known were the trans-Saharan trading caravans that linked towns and cities across northwest Africa, transporting gold, salt, and enslaved people across the desert. 10 States and empires, most famously the Almoravids, also spanned the desert, linking the Mediterranean to the Senegal and Niger Rivers. Islam forged other links, from the famous fourteenth-century hajj of Mansa Musa to the Sufi orders, such as the Sanusiyya and the Tijaniyya, that so frustrated colonial powers bent on controlling them. 11 Anthropologists have shown how, in the postcolonial period, Saharan crossings have taken new forms as trade and migration have shifted with new political, economic, and environmental contexts. 12 These trans-Saharan studies, while challenging the notion of the desert as a barrier, often leave a temporal gap: the era of European colonialism in the nineteenth and twentieth centuries.
Writing a trans-Saharan history of the colonial period confronts these epistemological challenges as well as methodological issues related to archival research. As Baz Lecocq argued in an analysis of trans-Saharan historiography, European colonial scholarship on Africa reinforced the "Saharan Divide," prioritizing different academic fields in North Africa (philology, archaeology, Orientalism) and West Africa (anthropology, ethnography, linguistics). 13 Colonial "scholar-administrators," who combined service in the French colonial bureaucracy with academic work justifying their policies and who sometimes found university posts later in their careers, remain 10 Robinson (eds.), La Tijâniyya: Une confrérie musulmane à la conquête de l'Afrique (Paris: Karthala, 2000); and Jean-Louis Triaud, La Légende noire de la Sanûsiyya: Une confrérie musulmane saharienne sous le regard français (1840-1930) (Paris: Éditions de la Maison des sciences de l'homme, 1995). For an overview of this literature, see Scott S. Reese, "Islam in Africa/Africans and Islam," Journal of African History 55-1 (2014), 17-26. 12 Judith Scheele, Smugglers and Saints of the Sahara: Regional Connectivity in the Twentieth Century (New York: Cambridge University Press, 2012); Meike Meerpohl, "Camels as Trading Goods: The Transition from a Beast of Burden to a Commodity in the Trans-Saharan Trade between Chad and Libya," in Ahmida, Ali Abdullatif (ed.), Bridges Across the Sahara: Social, Economic and Cultural Impact of the Trans-Sahara Trade during the 19th and 20th Centuries (Newcastle-upon-Tyne, UK: Cambridge Scholars Publishing, 2009), 157-185. 13 Lecocq,"Distant Shores,[30][31][32][33][34] foundational to their respective fields to this day. 14 The division of Africa at the Sahara is thus built into the foundations of colonial history, although these colonial officials sometimes explicitly undermined this division in their policies and in their personal trajectories. 15 Indeed, following the careers of these individuals is one way that historians have begun to address the trans-Saharan history of French colonial rule. 16 Tracking colonial institutions is another. 17 Adopting this trans-Saharan approach, however, poses practical challenges based on the material conditions of the colonial archive. For all their methodological and theoretical innovations, historians of colonialism tend to follow the archival grain in the sense that they reify colonial borders. These frequently arbitrary lines on the map, which by and large became national borders at independence, are replicated in archival institutions, whether in Africa or in the former colonial metropoles. One salient exception is the National Archives of Senegal, in Dakar, which hold colonial-era records for the nine countries that were once part of French West Africa. Many archival records, though not all, were brought to France during the era of 14 See Christopher Harrison, France and Islam in West Africa (New York: Cambridge University Press, 1988), especially 93-117, and Lecocq, "Distant Shores." Paul Marty, Maurice Delafosse, and François Clozel are the three most famous of these figures, but many administrators also produced scholarly studies of African societies, especially in the late colonial period. Another later figure is Yves Person, whose career as an administrator and as a historian culminated in his massive study Samori: Une révolution dyula (Dakar, Senegal: IFAN, 1968). 15 For example, General Louis Faidherbe stated in the 1850s that Senegal should be "no more than a sub-division of Algeria," governed by the same institutions. decolonization; researchers there and in African archives face ongoing challenges to access materials. 18 The Area Studies paradigm poses practical problems as well, as trans-Saharan scholars train, teach, and research as specialists in sub-Saharan Africa or the Middle East, but very rarely both. Some historians of colonialism are trained in French history, though it is happily becoming less common for histories of Africa to be written solely out of European archives. Scholars of Jewish history have also made significant contributions to the study of the Sahara and trans-Saharan networks. 19 The dispersal of archival collections and the disjointed Area Studies approaches to this region require trans-Saharan researchers to master multiple historiographies and to design complex research itineraries.
Constituting a trans-Saharan archive differs from other sorts of multisited research. In an influential article, Angela Zimmerman has argued for "critical imperial history [as] one of the routes beyond imperial history, toward transnational history." 20 While they emphasize the importance of Africanist scholarship in "provincializing" European theoretical frameworks, Zimmerman's material approach to transnational scholarship retains an emphasis on the relationship between the European metropole and the African colony. In other words, for historians of colonialism, transnational research has involved traveling to a large European imperial archive and to smaller African archives with a focus on a single region or theme. For trans-Saharan researchers, interested not only in the relationship between metropole and colony but between different colonized spaces, this approach is insufficient. Trans-Saharan colonial histories demand a multi-local approach, sensitive to the fragmented colonial archive and the long-term history of interconnection across northwest Africa.
Since the seminal work of Ann Laura Stoler and Frederick Cooper, historians of empire have developed sophisticated methods of reading colonial archives as evidence of imperial power and colonized resistance. 21 More recent scholarship has emphasized the complex entanglements of imperial history, as in Mostafa Minawi's and M'hamed Oualdi's studies of Ottoman engagement in North and Central Africa and Sarah Zimmerman's study of globetrotting West African tirailleurs and their conjugal relationships. 22 One result of these trans-imperial and intra-imperial studies is a clearer sense of how imperial rule forged and reshaped long-distance connections. The colonial médersas and the trans-Saharan figures like Ould Rouis who led them are another example of such entangled histories that reveal otherwise overlooked connections. Yet the seemingly straightforward issue of where to find sources for such a complex history remains urgent. The major repositories of French West African colonial history, in Aix-en-Provence and Dakar, left many questions unanswered. Ould Rouis and the médersas appeared in those archives, but haphazardly and obliquely. What brought an Algerian teacher to Mauritania? Even that basic a question had no answer in those archival collections.
Searching for answers in archives across northwest Africa, I came to see that the médersas were colonial institutions that underwent a process of adaptation to their local contexts. To understand Ould Rouis's trajectory, I had to understand the complex Algerian context he came from, the complex contexts of Mauritania and French Soudan he worked in, and the linked histories of education, Islam, and colonization across Algeria and French West Africa. In other words, multiple areas of inquiry inform one another and allow for a larger picture to emerge, one more complete than can be gained from any single vantage point, historiographical tradition, or archival collection. Pursuing these questions in multiple places means crossing borders, both historiographically and physically. It requires confronting the geographical frameworks that structure scholarly inquiry; it requires time and effort to build multiple local networks and to engage with multiple communities and their archival institutions. As an approach to the dispersed colonial archive, multi-local research allows for a colonial history that overcomes colonial boundariesa colonial history that is, in some sense, decolonized.
The French colonization of northwest Africa produced documentation that differs in both quantity and quality from the archive of the periods before it. The colonial bureaucracy was the core producer and preserver of printed archival materials, and its output was so vast that it remains incompletely catalogued or examined by archivists and historians. In the case of the médersas, colonial bureaucrats from the relevant officesthe schools themselves, various education offices, the Service of Muslim Affairsproduced voluminous reports, correspondences, and other ephemera. These documents were inconsistently produced and preserved; administrators occasionally skipped a quarterly report, for example, leaving a gap in the record. So, too, did students produce a voluminous archive of examinations, petitions, even doodles. Their preservation in archives is similarly inconsistent: correspondence is often split among archives; copies of reports are distributed seemingly haphazardly; troves of administrative minutiae appear unexpectedly in archival dossiers devoted to other topics. This vast but diffuse paper trail allows for a relatively complete view of the history of institutions such as the médersas, but achieving that view necessitates research in a wide range of archives in Africa and Europe. In tracking individuals alongside institutions, the research process necessitates multi-local inquiry. Putting the disparate pieces together, while remaining attentive to multiple local contexts, reveals how colonized people navigated imperial spaces and shaped imperial systems. The careers of two Algerians, Boualem Ould Rouis and Abderrahmane Nekli, both of whom leveraged their experiences in Algeria to transform the médersas of Mauritania and French Soudan, illustrate how multi-local research illuminates this trans-Saharan history.

Algerian Teachers in Mauritania and Mali
Médersas were institutions of bicultural education, incorporating both French curricula and traditions of Islamic learning, run by the French administration. 23 Three médersas created in Algeria in 1850 educated young men there until 1951. Amid the conquest of the West African interior in the early twentieth century, seven more médersas were founded in West Africa: one in Senegal (in Saint-Louis), two in the French Soudan, today Mali (in Djenné and Timbuktu), and four in Mauritania (in Boutilimit, Atar, Timbédra, and Kiffa) (see Map 1). Mauritania was an especially popular site for Franco-Muslim schooling for two reasons. The first was the importance of Islamic education to the elite bīdān families of the region; the second was the French racist conception that the bīdān practiced a more authentic form of Islam than Black groups in Mauritania and elsewhere in West Africa. 24 The médersas 23 See Samuel D. Anderson, "Domesticating the Médersa: Franco-Muslim Education and Colonial Rule in Northwest Africa, 1850-1960" (unpublished PhD thesis, University of California, Los Angeles, 2018). As Louis Brenner and others have explained, "médersa" became in the late colonial period a category of bicultural school that was not necessarily under the control of the colonial or post-colonial government. See Brenner, Controlling Knowledge: Religion, Power and Schooling in a West African Muslim Society (Bloomington: Indiana University Press, 2001) for a case study focused on French Soudan/Mali. 24 The term bīdān, meaning "white" in Arabic, is a racial category with a long history in the southwestern Sahara and indicates a dominant position vis-à-vis those categorized as Black (known by terms such as sūdān, hratīn, and bellah in multiple contexts). See Catherine Taine-Cheikh, "La Mauritanie en noir et blanc. Petite promenande linguistique en hassâniyya," Revue du monde musulman et de la Méditerranée were thus a trans-Saharan institution: originally developed in a Maghribi context, they underwent a series of adaptations in the Sahara and Sahel, thereby creating a system of schools united in their basic structure and purpose but with varying institutional traitsdifferent student populations, differing curricula, wide-ranging relationships with other branches of the colonial administration and with the African communities within which they operated.
Algerian teachers were central to this process of adaptation in West Africa; their actions show the power of colonized individuals to shape colonial institutions. Ould Rouis and Nekli exemplify the ways in which Algerians were able to act as intermediaries for the colonial government in Mauritania and elsewhere. 25 Their position vis-à-vis other colonial intermediaries endowed them with more power, and their experiences were more unusual. It was rare for colonized individuals to cross colonial borders with positions of authority or leadership (other than the rare example of Caribbean-born administrators such as Félix Éboué). More commonly, Africans who moved across the empire did so as soldiers, traders, or in other positions that left fewer individual traces in the archive. By virtue of their powerful colonial careers, Nekli and Ould Rouis appear in multiple archives; it is only by retracing their footsteps that their impact becomes fully visible.
Ould Rouis was not the first Algerian to come to Mauritania, but he was among the most influential. He arrived to direct the médersa in Boutilimit, in the southern Trarza region, in 1929. He replaced another Algerian, Mekki el-Djenidi, who had a series of disputes with local leaders that resulted in the school's temporary displacement to the nearby town of Mederdra. The school's interim director was Bokar Ba, a Fulani Mauritanian who, reversing the trajectory of the Algerians, had studied at the médersas in Saint-Louis and Algiers. He was by all accounts a more competent administrator than el-Djenidi, but he was disqualified from a permanent post directing the school because of his race. Black educators could teach at the West African médersas, but they were not allowed to serve as their directors. 26 Ba's successes helped pave the way for Ould Rouis, who was credited with the school's recovered reputation and subsequent prosperity. Within a year of his arrival in Mauritania, Ould Rouis was praised for "the confidence and respect he inspired in his students and their parents," by the Lieutenant-Governor of the colony. 27 His tenure in Boutilimit saw the médersa grow into the largest colonial school in Mauritania. Only under his direction did the school begin to enroll the long-coveted sons of the bīdān elite of the region. Whether or not he was fully responsible, Ould Rouis was credited with the transformative success of the Boutilimit médersa.
He became an important collaborator of the colonial administration of Mauritania. In 1935, when the French decided to open another médersa in the northern Adrar region, Ould Rouis became the main architect of the project. He was sent as an emissary to the region's elite to convince them of the médersa's value. His efforts paid off. He consulted especially carefully with the Reguibat, a nomadic confederation who ranged from the Adrar north to Guelmim, in southern Morocco, and who posed the most consistent threat to Spanish rule in Rio de Oro (what is today Western Sahara). When, upon Ould Rouis's urging, some of the Reguibat began to enroll their sons at the new médersa, it was seen as a major accomplishment for the French and a testament to Ould Rouis's personal influence. 28 The new médersa was built in Atar, a town that hosted the largest French military post in the Adrar region. It was chosen over Chinguetti (Shinqit), a nearby town with a longstanding reputation for Islamic learning. 29 The choice was doubtless made so that the colonial authorities could more effectively surveil the school and its students; Ould Rouis and others also argued that Chinguetti was in decline and would not be able to attract students. 30 The school's final organization followed Ould Rouis's recommendations very closely. The physical structure replicated that of the Médersa of Boutilimit with a small compound, including a four-classroom school building, a small house for the director, and two dormitories for the students. 31 The site was a shady palm grove a short distance from both the historic center of Atar and the new French military post. Ould Rouis provided a list of materials, from classroom furniture to kitchen utensils, which also matched those used at Boutilimit. He listed a set of ambitious goals that he thought the new médersa would accomplish, including "vanquishing prejudice and inspiring the confidence" of the bīdān elite; "awakening a sense of humanity and a sort of French consciousness" in the students, and rivaling, at least locally, the prestige of well-known centers of Islamic learning including the Zaytuna mosqueuniversity in Tunis and al-Azhar in Cairo. 32 In short, Ould Rouis and the French administration sought to supplant the scholarly history of Chinguetti through the new médersa in Atar, itself a carbon copy of the médersa in Boutilimit.
Though the loftiest of these goals went unmet as the French blocked students from traveling to study in North Africa through the end of the colonial period, Ould Rouis's reputation remained sterling through the end of his time in Mauritania. He was named the inaugural director of the Médersa of Atar, where his tenure was "entirely satisfactory," in the words of one education inspector. 33 In selecting Ould Rouis for that post, and in following his other recommendations so carefully, the governor of Mauritania went against the advice of the preeminent Orientalist Louis Massignon, who had provided the names of several other candidates, none of whom had experience with either Franco-Muslim education or with Mauritania. Rouis's experience and renown made him a powerful figure in the colonial administration in Mauritania for a decade. Few others, especially few Algerians, could claim a similar impact.
At the end of his tenure in Mauritania, Ould Rouis's influence spread yet further. In 1939, Ould Rouis accepted a position at the Médersa of Constantine. On his way back to Algeria, traveling overland, he stopped for several weeks in Timbuktu to evaluate the médersa in that city. His status and experience in Mauritania proved influential among administrators in the French Soudan as well.
The Médersa of Timbuktu was the oldest continuously operating Franco-Muslim school in West Africa. It had been founded in 1911, part of a wave of médersas that opened in Djenné and Saint-Louis, and, slightly later, in Boutilimit. The médersas in Djenné and Saint-Louis closed promptly, for reasons that were closely tied to the institutionalization of Islam noir ("Black Islam") in French administrative policy and practice across West Africa. 34 In Timbuktu, as in Boutilimit, the médersa stayed open because it catered to "white" nomadic elite families. Whereas Ould Rouis's leadership in Mauritania led to the growth and expansion of médersa education, in Timbuktu the médersa languished from the 1910s to the 1930s. Until 1934, it was under the direction of an eccentric Frenchman, Auguste Dupuis, also known as Yakouba, who had come to Timbuktu as a missionary. He soon left the church, achieving some degree of celebrity in France as someone who had "gone native" in the colonies. 35 His leadership at the médersa did little to inspire confidence, either from French administrators or the Timbuktu elite whom he was meant to cultivate as allies. Reports by colonial inspectors criticized him as "old, incapable, and otherwise useless" with only a "superficial" knowledge of Arabic. 36 His successors, a French-trained Soudanese teacher named Bouillagui Fadiga and a French teacher named Vermande, were both tasked with reforming the médersa but met with little success. Fadiga adopted a heavyhanded approach to disciplining students and staff and emphasized the French curriculum; his students and the elites of Timbuktu promptly rejected him. 37 Vermande (his first name went unrecorded) echoed the 34  It is worth noting that Brenner cites sources here from the National Archives of Mali, which I was unable to visit but which I did not see copies of elsewhere. It is likely that Bamako and Timbuktu, like Nouakchott, hold rich archival materials that are unavailable in Dakar, Aix-en-Provence, and elsewhere. disdain of French educators in Djenné and Saint-Louis for Franco-Muslim schooling. He recommended the closure of the médersa, writing that "Timbuktu is not a holy city" and, as such, embracing Islamic education was a strategic mistake. 38 Unlike the médersas he had led in Mauritania, the médersa Ould Rouis encountered in Timbuktu was clearly not a thriving institution.
Indeed, when he arrived in Timbuktu, Ould Rouis was dismayed by the state of the school. Unlike Vermande, however, he laid out a comprehensive plan for reforms instead of recommending its closure. These recommendations drew explicitly and directly on the Mauritanian médersas he knew so well in Boutilimit and Atara comparison that earlier leaders, without such expansive experience, had not made. Indeed, his proposals for the Médersa of Timbuktu mirrored almost exactly his plans for the Médersa of Atar, drawn up four years before and put into successful practice afterward. He recommended that the school move from its site in the center of Timbuktu's old city to a new compound on its outskirts. He argued that students should be housed in dormitories on siteas in Atar and Boutilimitrather than with local host families. Throughout, he cited the positive response to his ideas in Mauritania as evidence that they would succeed in Timbuktu as well. 39 Ould Rouis criticized the faculty of the school as mediocre, especially in their knowledge of Arabic, which he saw as the core of a médersa education. In recommended changes to the teaching staff, he again sought to connect the Soudanese and Mauritanian institutions directly. He proposed hiring two Mauritanians to teach in Timbuktu as a way to guarantee the Arabic fluency of the staff. One, Mohamedou Ould Abdillah, was a jurist, poet, grammarian, and the professor of Qur'anic studies at the Médersa of Atar; Ould Rouis proposed him as the professor of Arabic at Timbuktu. The other, Beddi (also recorded as Badadi) Ould Mounir, was still a student at the Médersa of Atar, originally from Chinguetti. Ould Rouis recommended hiring him as a moniteur, or a teaching assistant, for the Arabic classes in Timbuktu. 40 By the next year, Ould Mounir was indeed employed in that capacity at the Timbuktu médersa. 41 The médersas developed closer institutional and personal connections, through Ould Rouis, across colonial borders.
Ould Rouis drew a further connection through the idiom of race. In Timbuktu, as in Mauritania, administrators had long been frustrated that so many of the médersas' students came from the local Black populations, especially those of lower social statusthe hratīn in Mauritania, the bellah and Songhay in Soudan. In their view, the médersas were meant to transform the existing social elite, called white, or bīdān, in both places, into allies and agents of the colonial state. But the elites were often wary, and lower classes 38  more eager to seek advantage through a colonial education. A 1935 report on the Médersa of Timbuktu lists the students as follows: out of 105 students, 93 were Black (Songhay, Fula, or bellah). The remaining students included 8 "Cherifs," probably denoting residents claiming Moroccan ancestry, dating back to the 1591 Moroccan invasion of the Songhay Empire, as well as 3 Berabich (an Arab nomadic clan) and 1 Kel Arawan (a Tuareg group). 42 In other words, only 12 percent of the school's population came from the Arab-Amazigh (Berber) elite that the administration targeted; even fewer came from within the politically crucial target demographic of the sons of local chiefs. In his 1939 report, Ould Rouis dismissed the administration's concerns and proposed a simple solution. The students would be separated into two groups, bīdān and sūdān, such that the "racial question" would not be an issue. He also suggested an answer to the question of how to increase the numbers of elite students: the administration should simply "accept only the sons of noble families." 43 Finally, Ould Rouis recommended that the position of médersa director be reserved exclusively for an Algerian graduate of the Algerian médersas. Europeans, he argued, were incapable of fully understanding the complex social and political context of Timbuktu and its hinterland. Only an Algerian would be able to transform the médersa by following Ould Rouis's prescriptions, all taken explicitly from Ould Rouis's Mauritanian experience in Boutilimit and Atar. They suggest the complex, trans-Saharan logic at work in the West African médersa system, which drew on, and reinforced, longstanding racial idioms that emphasized Arab authority, especially in religious matters. The personal success that Ould Rouis had found in Mauritania seemed natural to him and naturally applied to Timbuktu; ensuring that other Algerians followed in his footsteps seemed to be a sure way to continue making the médersas into useful instruments for the colonial administration.
Upon concluding his study in Timbuktu, Ould Rouis traveled to Gao, where he turned north and returned to Algeria by crossing the Sahara. He taught at the Médersa of Constantine for several years; he ended his career as a professor at the Médersa of Algiers, the largest and most important of the médersas. In 1951, the Algerian médersas were transformed into secular high schools, and in Algiers that school continues to operate as the Lycée Amara Rachid, named after a martyr of the independence struggle.
The archives of the médersa are preserved in the lycée's attic, and they are not formally classified or incorporated into the Algerian national archives system. I was able to access them after gaining authorizations from both the Ministry of Education and the school's director. By and large, even more than the other documentation preserved elsewhere, this médersa's archive is predominantly composed of quotidian bureaucratic sources: lists of attendance records, for example, dating back to the late nineteenth century. Nestled among these prosaic pages, though, are more unusual documents. One of them is a description of Mauritania, written in Arabic by Ould Rouis. It is, to my knowledge, the only account of the region by a North African in the employ of the colonial administration. There is no clear explanation for why he composed the manuscript, though it is kept in a folder alongside other texts by médersa professors and students, usually treating topics related to Islamic law, presumably written for pedagogical purposes.
The text is dated 9 April 1953, indicating it was written fourteen years after Ould Rouis left Mauritania, and two years after the transformation of the Algiers médersa into a lycée. The account is descriptive, even ethnographic, in nature. He characterizes the inhabitants of Bilad Shinqit as "healthy, strong, and smart," despite the scant resources they have access to in the desert. 44 He devotes much of the text to describing bīdān women, praising the way that girls are educated alongside boys, for example. He recounts the way the bīdān equate fatness with beauty and wealth: because of the region's poverty, "fat people are the most beautiful thing eyes can see there." 45 He describes the way women dress, walk, and clean and perfume themselves in the desert; he contrasts the hair texture and hairstyles of Shinqiti and Sudanese (in other words, bīdān and sūdān) women. Finally, he describes how marriage and divorce are conducted in bīdān society. The text concludes, "though the Shinqiti woman enjoys some freedom, she does not have a happy family or a stable marriage.…divorce is her destiny." 46 It is sprinkled with references to the Qur'an and to Arabic poetry, both classical and modern. The document demonstrates Ould Rouis's intimate familiarity with the society and culture of the southwestern Sahara.
By itself, this text is remarkable but difficult to interpret. It does not explicitly describe Ould Rouis's experience teaching in the Mauritanian médersas. In fact, Ould Rouis does not insert himself at all into the text, so when I encountered it in the attic in Algiers, I did not know what to make of it. Within the context of his life and work in Algeria, and within the context of the school's archive, his text seemed incongruous. It was only after reading his report from Timbuktu, in Nouakchott, that the disparate puzzle pieces of his career began to fit together. On their own, each of these two documentsand the many other, less detailed sources scattered in different archivesseemed incomplete, even unremarkable. Taken together, however, they revealed his central role in the development of Franco-Muslim schooling in Mauritania and Mali and demonstrated how the dispersal of archival materials in lesser-known repositories obscured his importance.
Just as one document, or one archive, does not reveal the full picture of this history so, too, does a focus on one individual obscure larger patterns. Ould Rouis forged a path that would be followed by other Algerians in the final decades of the colonial period. Abderrahmane Nekli followed most closely in Ould Rouis's footsteps. Nekli left an archival trail that is less extensive than Ould Rouis's; I was, however, able to interview two of his children in Nouakchott in 2016. Their memories allow for a more complete picture to emerge as their family's history across northwest Africa emphasizes again the importance of multi-local research in this region. Nekli was born in 1914 to a scholarly family in Kabylia. He studied at the Médersa of Constantine, and at the Superior Division of the Médersa of Algiers. He first arrived in Mauritania in 1939, posted originally at the new Médersa of Kiffa, in the southern Assaba region, and spent most of the next several decades in West Africa.
Before long, Nekli moved to Atar, where he succeeded Ould Rouis as the director of the médersa. He expanded on his predecessor's work, integrating himself into the local society and enhancing the school's prestige. Soon after his arrival in Atar, Nekli married Maryem Mint Habot, the daughter of his colleague at the médersa and a member of one of the region's most prominent families, the Ahel Habot. This marriage between an outsider and the Habot family would have been unthinkable in the earlier social world of the Adrar, but it had the effect of bolstering the médersa's reputation, linking Nekli and the school to local networks of Islamic authority. Nekli befriended the region's French administrators and visiting European researchers, forging especially close ties with scholars from the University of Neuchâtel in Switzerland. According to his children, Nekli felt closely connected with the local community: he would share the family's modest resources by providing food to hungry families in Atar and Chinguetti, especially in times of scarcity. 47 Through these acts, Nekli wove himself, and the médersa he led, into the region's social and cultural fabric.
One of Nekli's most influential contacts was a colleague at the Médersa of Atar, the Mauritanian historian Mokhtar Ould Hamidun. Originally from the southern Trarza region and a member of a prominent scholarly lineage, Ould Hamidun was hired to teach Arabic at the médersa in 1943. 48 Together, Ould Hamidun and Nekli oversaw a grand period in the history of the Médersa of Atar, when it gained wider acceptance by the local community. Even Marebbi Rebbo, a son of Shaykh Ma' al-Aynayn, the great resister of French and Spanish incursions into the Sahara, allowed two of his nephews to study at the médersa during this time. 49  teacher. 50 In 1949, Ould Hamidun moved on to a post at IFAN, the French West African Research Institute in Saint-Louis, where he joined luminaries including Amadou Hampaté Bâ, Théodore Monod, and Jean Suret-Canale and where he began publishing encyclopedic works on the history and culture of the peoples of the Sahara. 51 Like Ould Hamidun and Ould Rouis, Nekli developed a keen understanding of Saharan social life and used that knowledge to develop the médersa into an important educational institution in the region.
Nekli also followed Ould Rouis's footsteps to Timbuktu. Around the time that Ould Hamidun departed for Saint-Louis, Nekli was appointed to the post of director of the Timbuktu médersa. He would remain in the French Soudan for many years. In Timbuktu, he oversaw the transition of the médersa from its Franco-Muslim curriculum to Franco-Arabic schooling (in other words, the removal of the Islamic curriculum and the increase in Arabic language training), which was the fate of all the médersas in West Africa and in Algeria in the late 1940s and early 1950s. 52 Nekli was based in Timbuktu through the 1950s and 1960s, when, while continuing to teach in colonial schools, he became involved in the Algerian war for independence. He helped to coordinate the efforts of the FLN (Algerian National Liberation Front) and the GPRA (Provisional Government of the Algerian Republic) in the diplomatic and military arenas in West Africa. After Algeria won its independence in 1962, he remained in West Africa, serving as the Algerian ambassador to Mali, Mauritania, Niger, and Burkina Faso. He represented Algeria at the creation of the Organization of African Unity in Addis Ababa in 1963 and helped to organize the Pan-African Cultural Festival in Algiers in 1969. A pioneering force in Algeria's African diplomacy, he retired to Algiers, where he died in 1990. 53 Nekli and Ould Rouis are emblematic figures in the history of Franco-Muslim education in the late colonial period. Their complex statusas Algerian Muslims educated in both French and Islamic traditionsmade them embodiments of the hyphenated Franco-Muslim culture that the médersas were meant to foster in Algeria and West Africa. They were mediating figures in the Saharan societies where they lived and worked: colonial agents who earned the respect of local communities, and colonized people who were seen as experts by French officers. Through their efforts, the médersas of Boutilimit and Atar, and to a lesser extent Timbuktu, became important educational institutions in colonial Mauritania and Mali, the training ground of a new elite. Nekli's political activity demonstrates that these men were not always loyal pawns of the colonial state. Rather, the position afforded by employment at the médersas enabled these Algerian teachers to work in new ways to achieve distinct political and personal goals. Among groups that the French administrationand more recent scholarship considered fundamentally distinct from one another, these individuals developed affinities and relationships that shaped the course of local and regional politics, societies, and cultures. Invisible from the major archives of Dakar and Aix-en-Provence, their impact is clear through multi-local research in smaller archival collections.

Conclusion
The lives and careers of Boualem Ould Rouis and Abderrahmane Nekli show how trans-Saharan connections transformed under colonial rule. Trading caravans no longer served as the dominant mode of connection across the Saharan space. Rather, French colonial rule from the Mediterranean to the Bight of Benin shaped new modes of connection among Africans in this region. The médersas, with their Franco-Muslim character first developed in Algeria, show how colonial institutions grew across northwest Africa. Through these schools, Algerian teachers like Ould Rouis and Nekli became agents for the colonial administration, chosen for their ability to navigate complex social worlds alien to the French. Ould Rouis's service in particular illustrates how Algerian expertise shaped colonial institutions like the médersas and how it crossed colonial borders, in his case from Algeria to Mauritania and thence to Mali. Nekli's career likewise demonstrates the impact of Algerians in fostering Franco-Muslim education in West Africa, and it, too, indicates the extent to which individual agency and political action could operate outside of colonial frameworks, even for colonized intermediaries. In short, both men's careers show how accommodation and resistance operated in tandem, in individual lives and through individual agency, and how these forces were built into the structure of colonial institutions.
The intricacies of these lives are impossible to understand from a single vantage point, and impossible to reconstruct through a single archive or method. Only through trans-Saharan research, retracing their movements through multiple archival collections, can historians piece together their trajectories, and their significance. Moreover, following these individuals through their institutionsin this case, the médersasonly raises the stakes of multi-local research because institutional archives are so unevenly distributed throughout the region. In the story laid out above, the crucial source was Ould Rouis's report on the Médersa of Timbuktu. This proved to me that he drew explicitly on his experience as an Algerian in Mauritania while reforming colonial education in French Soudan. That document is stored only, to my knowledge, in the National archives of Mauritania, a repository that is far off the beaten path for historians of Algeria, French Soudan, or colonial schooling in general. And yet it explicitly and directly links these histories together.
As citations elsewhere in this article show, however, no single, unexpected archival find can illuminate this history or its significance. Piecing together the careers of Ould Rouis and Nekli required making linkages from the Mauritanian archives in Nouakchott to the archives of the Lycée Amara Rachid in Algiers, the Senegalese national archives in Dakar, and multiple collections in France, including the colonial archives in Aix-en-Provence and, surprisingly, the diplomatic archives divided between Nantes and Paris. 54 (I was unable to make use of archives in Mali, which doubtless hold yet more relevant documentation.) I was lucky to encounter, on a subsequent visit to Nouakchott, two of Nekli's children, who generously spent hours speaking with me about their father's life and their family's peripatetic history. Together, the many scraps of information gathered in these places illuminated the larger narrative of these men's lives, their contributions to the médersas, and the larger-still trans-Saharan history of Franco-Muslim education.
In a recent contribution to the American Historical Review, the Africanist historian Emily Callaci reflects on acknowledgments as a place where gender relations and privilege in the academy are laid bare, as scholars credit their families for the sacrifices they make for the sake of scholarship. 55 Footnotes might be another place to look for evidence of intersections between the intellectual and the personal. I was only able to embark on the extended research program this project required due to the confluence of several factors. With my American passport, I was able to visit multiple countries multiple times, obtaining visas and research clearances while others, with different citizenship or less lucky timing, might not have. As a white man, I could navigate these foreign societies and institutions with greater ease than others without the racial and gender privilege I hold. As a young scholar with relatively few family obligations, I was able to devote significant time to traveling for research. I was also lucky to receive external research funding, which made these travels possible in a way that my state university institution would never have been able to fund on its own. This is not to say that these benefits came without other, more hidden costs: it was still logistically and emotionally difficult to uproot my life for an extended period. As a gay man, 54 I initially went to the diplomatic archives looking for material on Franco-Muslim education in the protectorates of Morocco and Tunisia. I was surprised to find significant material related to the West African colonies, in particular from the late colonial periodmaterial that was not held in Aix-en-Provence or in Dakar. 55 Emily Callaci, "On Acknowledgments," The American Historical Review 125-1 (February 2020), 126-131. though I knew my passport protected me, I was ill at ease in countries where homosexuality remains a criminal offense (and, in Mauritania, a capital crime). At present, of course, the ongoing coronavirus pandemic makes such a research agenda all the more complicated to imagine for any scholar.
And yet, researching across the Saharan divide has made it clear how the Area Studies paradigm distorts our understanding of the history of northwest Africa. Equally rooted in the Maghrib and in West Africa, the médersas are one concrete example of trans-Saharan connections in the colonial era. Ould Rouis, Nekli, and their colleagues embodied the ways in which Islam and colonialism intersected to create new connections between places and people usually considered in separate scholarly frames. It is only through multilocal research, across the imagined division of the Sahara, that such histories can be recovered. Doing so requires a new, logistically complex approach to the diffuse colonial-era archive. But the rewards of such a method are many, chief among them a fundamental reconceptualization of conventional regional frameworks to enable a more comprehensive vision of this influential period in African history.