Rivalries between Dogon and Fulani in Central Mali


Last month, a Dogon militia massacred 160 Fulani in central Mali, more than 1000 km from Bamako. Condemned by the international community, this incident led the Malian President to sack two generals and dissolve Dan Na Ambassagou, the group suspected of originating the attack. Amid growing pressure on insecurity in the country, this event also contributed resignation of the Malian government on April 18.

Very little is known about the Dan Na Ambassagou militia, whose name translates as “Hunters who trust in God” in Dogon and who denies the attack. Some commentators have argued that the development of religious extremism in central Mali has fueled the attack. Ethnic militias, they argue, increasingly target Fulani populations, which they accuse of supporting jihadist groups, such as the Macina Liberation Front (FLM) run by Amadou Koufa, now affiliated with the Group to Support Islam and Muslims (GSIM).

The local roots of conflict

The development of religious extremism is obviously not the only cause of violence in Central Mali. Conflicts have many causes, which combine differently in different regions. In what follows, I argue that more attention should be paid to the local factors that can potentially explain why ethnic militias develop in this part of Central Mali. The recent history of Central Mali shows that periods of contraction resulting from the withdrawal of nomadic and sedentary groups are followed by periods of trade or agricultural expansion.

In the Séno-Gondo plain, where the massacre of Ogossagou and Welingara took place last month, the expansion of the Fulani in the 19th century led Dogon farmers to withdraw from their villages on the plain and adopt defensive settlements in the Bandiagara Cliff and on the plateau. In the 20th century, the descendants of these farmers have moved eastwards to fill the void left by the Fulani and have given rise to an agricultural front that they regard as their original lands[1].

The Bandiagara Cliff at Banani. Source: © Walther (1996)


The Fulani expansion

The Dogon arrived in the Bandiagara area in the 15th century and dispersed into relatively autonomous communities that colonized not only the Bandiagara cliff and plateau but also the vast plain of Séno-Gondo, a sandy area east of the cliff that provided fertile ground for cereals, abundant water resources and nutritious wild fruit. Some of the oldest Dogon villages were established in the Séno-Gondo plain, such as Diankabou, whose language (diamsay) is still considered as one the lingua franca of the Bandiagara cliff and Séno plain. This first colonization of Séno-Gondo followed a principle common to many pioneering fronts of West Africa according to which land belongs to the first people to establish an inalienable agreement with the deities of the place.

The Séno-Gondo plain and the village of Nombori. Source: © Walther (2001)


This agricultural expansion was short-lived. In 1818, a Fulani conqueror named Seku Amadu founded the Empire of Massina (Diina) and a new capital, Hamdullahi, located southeast of Mopti. The Massina Empire gradually extended from Segou in the south to Timbuktu in the north. Like many other precolonial political formations, the Massina Empire maintained fuzzy peripheries where pagan peoples were either converted to Islam by force or enslaved. The Fulani used their cavalry to raid the Dogon plateau and the plain of Séno-Gondo, destroying the crops of the farmers and enslaving the local populations. In response, the Dogon built spectacular fortress villages in the Bandiagara cliff, a World Heritage site listed by UNESCO in 1989. Those who remained in the plain of Séno-Gondo became the serfs of the Fulani (rimaïbé). Further north, in the Hombori Mountains, the warriors of Seku Amadu also tried to impose Islam by force on the Dogon populations, who took refuge in perched villages or in the surrounding bush.

Nomadic herders in front of Hombori Tondo, the highest point in Mali (1155 m). Source: © Walther (2006)


The Dogon could hardly count on their military strength to defeat the Fulani. Instead, they allied themselves with El Hadj Umar Tall, a Toucouleur conqueror from Futa Toro in the valley of the Senegal River who conquered the city of Hamdullahi in 1862. El Hadj Umar Tall’s nephew, Tidjiani Tall, settled in Bandiagara where he ensured the fidelity of the Dogon. The French arrived in the region in 1893 and supported the Toucouleur in their fight against the Fulani of Massina. The fall of the Massina Empire and subsequent French colonization allowed the Dogon to resume their expansion in the plain of Séno-Gondo at the expense of the Fulani. The small city of Bankass near Ogossagou and Welingara quickly became a market where Toucouleur families settled.

Colonization and the Dogon expansion

Soon after having created the administrative Cercle de Bandiagara in 1903, the French tried to prevent the Dogon from migrating to the Séno-Gondo plain. The colonial administration feared that the local political chiefs would lose control over the population and no longer be able to collect tax. France, however, never had the means to counter the Dogon migration and merely counted the migrants at their village of origin until 1958, which minimized the magnitude of the migratory movement towards the plain in the colonial demographic statistics.

During this second colonization of the Séno-Gondo plain, the Dogon reclaimed the lands their clans possessed before the Fulani conquest of the 19th century. Numerous families left what was known at the time as the Old Dogon Country to settle in the New Country east of the Bandiagara cliff. As the maps of Jean Gallais show, each of the old villages developed a colonization corridor roughly perpendicular to the cliff. The villagers of the Togo clan of Kani, for example, established Kani-Kombolé at the cliff’s foot before settling a few kilometers away from their village of origin in the direction of Bankass.

Some parts of the plain were more favorable than others to agriculture. In the north, for example, the sand of the erg of Séno-Mango was too deep to allow access to the water table and agriculture only occupied the cultivable land near the cliff. In some other parts of the plain, as in the Bankass region, the agricultural expansion was also encouraged by a handful of fortified Dogon villages that had resisted to the Fulani.

New and old village of Téli, Bandiagara Cliff. Source: © Walther (2001)


Agricultural rivalries today

The independence of Mali in 1960 did not challenge the colonization of the Séno-Gondo plain by the Dogon. While the Bandiagara plateau and the cliff became known as a tourist destination, the plain transformed into an agricultural front where competition for land and water between farmers and herders was increasingly fierce. Weakened by the Toucouleur, the Fulani of the Séno-Gondo plain were divided by internal rivalries and never managed to contain the demographic and economic expansion of the Dogon. As a result, they progressively lost most of their cattle and turned to agriculture or small livestock that could be raised around the villages. Politically, the Fulani could also not match the political support that the Dogon found at all levels of the state hierarchy.

In recent decades, the Dogon migration from the cliff towards the Séno-Gondo plain has intensified due to demographic growth. New crops have replaced the pastures of the Fulani, whose way of life is now threatened by a lack of investment in the pastoral sector and recurring droughts. The presence of the Malian state in the region has also diminished as insecurity in the center of the country increases. This explosive context encourages ethnic militias to capitalize on the fear of religious extremism to promote their local objectives, such as that the agricultural resources of the Séno-Gondo plain.

The current militarization observed in Mali is unlikely to provide a lasting response to these conflicts, whose roots lie in the power relations between local communities. In the Séno-Gondo, as elsewhere in Mali, the beginning of a political solution involves the reestablishment of state and community institutions capable of guaranteeing that farmers and pastoralists have equal access to land resources, education, and political representation.​


[1] More than 40 years after its publication, Jean Gallais’ Pasteurs et Paysans du Gourma (CNRS, 1975) remains an indispensable reference on the spatial organization of the Inner Niger Delta and its peripheries. More recent studies include Véronique Petit’s Migrations et Société Dogon (L’Harmattan, 1998), Jean-Christophe Huet’s Villages Perchés des Dogons du Mali (L’Harmattan, 1994) and Bénédicte Thibaud’s Enjeux conflictuels autour de la ressource. Le cas de la réserve des éléphants du Gourma (Sahel malien) (Les Cahiers d’Outre-Mer, 2011).

Women and trade networks in West Africa


Women play a crucial role in the West African food economy. Yet, their trade networks remain often unknown. Our new study contributes to fill this gap by providing the first comprehensive analysis of the structure of West African social networks in which women play a particularly central role.

Our report published by the OECD last week identifies both the socio-economic barriers that restrict the opportunities of women in the rice supply chain, and the constraints that affect the governance networks aimed at promoting women’s entrepreneurship in the region.

The study shows that the general structure of business relations within the rice network imposes an unequal division of labor based on gender. The most prosperous actors in the rice network are those who have established numerous ties within and beyond their community. Women earn far less and are less central in the rice trade network than men.

Our work also shows that the governance network supporting women entrepreneurs in West Africa has a relatively dense center composed of international organisations, some West African countries and western partners that are at the heart of a wide range of initiatives. However, very few well-connected organisations can play an intermediary role at the regional level and the exchange of information between them is fragmented. We argue that the fragmentation of the governance network contributes to weak co-ordination of development policies designed to help women entrepreneurs in the region.

The study was part of the “Cities and Borders” project funded by the OECD from 2017-2018 which I coordinated with my colleague Leonardo A. Villalón at the University of Florida, Marie Trémolières at the OECD, Leena Hoffmann at Chatham House, and Lawali Dambo and Moustapha Koné at the University of Niamey, Niger.


Border cities in West Africa


The OECD just released a series of four papers on West African border cities. These papers are part of the “Cities and Borders” project that my colleague Leonardo A. Villalón and I coordinated from 2017-2018 in the region.

The papers provide a systematic analysis of the role West African border cities play in the process of regional integration. Based on a multidimensional mapping of 18 countries, they analyze the local dynamics that have developed in urban areas, the impact distance has on national cohesion and the impact territorial divisions have at the international level.

At the local level, the study of demographic and morphological changes identifies the effects density has on border cities. It shows that since 1950, the growth of border cities has almost always been greater than that of other cities in the region. This rapid growth has been especially visible within 50 km of national borders, where the most dynamic markets are located. Our work also confirms that border cities emerged without a concerted development plan and remain very dependent on each other.

At the national level, we studies the impact of distance on health services and formal businesses. Our work shows that the potential of harmonizing the health policies of the different countries has been largely untapped. The mapping of formal businesses specializing in certain sectors of strategic importance to regional integration shows that most are located in political and/or economic capitals where decisions are made concerning customs policies as well as import/re‑export strategies, and where major transportation and communications infrastructure is located. The lack of public investment in health services as well as roadway and education infrastructure in border regions can potentially pose major problems for national cohesion.

At the international level, we looked at the effects of territorial divisions using an accessibility model. The results show that the population base for border cities could be 14% larger if border crossings did not impact the flow of goods and people and 12 to 50% larger without roadside checkpoints. Between Benin, Niger and Nigeria, an analysis of the condition of the road network shows that the combined population base of eight border cities would increase by one third if there were no waiting at the borders. An exhaustive list of adjacent border posts in place or planned by national governments or regional organisations throughout sub-Saharan Africa further shows that trade facilitation runs up against the special interests of public servants and private-sector actors making a living from regional integration frictions. In West Africa in particular, few states are now able to benefit from the newly built border post structures in the region, most of which are not operational.

An Atlas of African Cities


West Africa is urbanizing at one of the fastest rates in the world. Between 2015 and 2040, the population of the region will double and cities will absorb most of this demographic growth.

To better understand urban growth in West Africa, the UF Sahel Research Group will contribute towards the flagship report “Atlas on Cities in Africa” published by the OECD Sahel and West Africa Club in 2020.

The main objective of the Atlas is to analyze urban transformations and integrate their economic, social and territorial impacts into public policies and the development agenda.

This project builds on the Atlas of the Sahara-Sahel (2014), on our recent work on West African border cities and on the Africapolis database on urbanization in Africa recently released by the OECD.

Foreign interventions and networks of violence in the Sahel

africacheering-1031743-300x200 (1)

From 2019-2020, I will be coordinating the new research program on political insecurity of the OECD in collaboration with my colleagues Steve Radil and David Russell.

In this new project, our first goal is to develop reliable data on the spatial patterns and social networks of violent extremist organizations in the Sahara-Sahel from the late 1990s to today and at the smallest possible geographic and temporal scale for analysis.

We’ll look at three conflicts in which the relationships between the belligerents are often characterized by complex sets of alliances and conflicts that change over time: the civil war in Mali, the Boko Haram insurrection around Lake Chad, and the civil war in Libya.

We’ll then examine the effect of foreign interventions on the networks of alliances or conflict between violent extremist organizations in the region. we are particularly interested in testing whether foreign interventions lead armed groups to align with the side they believe has the highest chance of winning the conflict or whether they reinforce internal divisions within groups.

We also want to study how should foreign powers choose between competing groups in conflict. Which intervention strategy is more likely to promote cooperation between warring parties while also reducing conflict?

AAG 2019 – Space and social networks


Call for papers: American Association of Geographers Annual Meeting, Washington D.C., 3-7 April, 2019

Session title: Space and social networks

Organizers: Steven Radil, University of Idaho; Olivier Walther, University of Florida

Session description: The spatial metaphor of the network along with its accompanying abstractions, such as flow, movement, and connectivity, have been central themes throughout the relational turn in human geography. However, to date networks in geography have been primarily explored either through actor-network theory or assemblage thinking, both of which embrace the network metaphor without specifically and formally interrogating networks themselves. The session seeks to problematize the treatment of networks in geography by exploring the largely underutilized literature on social networks as an alternative to other relational frameworks. Our session invites papers that explore the conceptual connections between core geographic concepts, such as place, distance, scale, territory, and power, and social networks. We are particularly interested in papers that open new directions for geographers that are interested in more than the metaphor of the network.

We envision a session that explores themes such as:

  • Geographic approaches to social network analysis
  • Networks of places and/or places as networks
  • Spatial proximity and networks
  • Cross-border networks
  • Geographic variation in social networks
  • Networks and scale
  • Power and networks

We invite papers on these or related topics. Please send proposed titles and abstracts (250 words or less) and/or expressions of interest to both Steven Radil (sradil (at) uidaho.edu) and Olivier Walther (owalther (at) ufl.edu) no later than 31 October, 2018.

Why so many rebel groups in Syria?

ooIn Syria, hundreds of factions operate under dozens of separate organizational command structures. Rather than coalescing into a unified rebel front, rebel groups continue to compete for power, thus failing to develop structures of governance and political authority that cut across factional divides, enclaves and provincial boundaries.

Despite the fact that inter-rebel violence is a common phenomenon, the mechanisms causing fragmentation and infighting within rebel movements remain poorly understood. In a new working paper posted today on SSRN, my former graduate student Patrick Steen Pedersen, who now works at the Royal Danish Defence College, and I discuss what has caused the Syrian rebel movement to fragment from 2011 to 2017.

Building on Bakke et al.’s study who suggests that divisions within rebel movements can be explained by examining how numerous, institutionalized and powerful are the organizations within a conflict, we divide the conflict into four different phases and discuss the development of various patterns of fragmentation throughout the conflict.

Our study reveals that the causal mechanisms of rebel fragmentation in Syria are both endogenous and exogenous. The circumscribed state of the pre-war political opposition, the absence of independent institutions and the state’s co-optation of Syria’s civil society organizations severely thwarted the way in which the widespread discontent with the Syrian regime could manifest itself at the onset of the conflict.

Exogenous factors also played a key role, particularly the lack of donor coordination, intra-regional competition and reorientation of patron policies.

Applying Social Network Analysis to terrorist financing


Illicit financial networks are, by their very nature, difficult to detect and, therefore, difficult to study. Much of the information on individuals and their activities is either classified or unknown. Nonetheless, tracking how terrorists raise, move, store and use money is fundamental to deter and discourage terrorist networks.

In a new chapter published in the Handbook of Criminal and Terrorism Financing Law, my colleague Christian Leuprecht and I posit network science as a method to improve our understanding of the way terrorists, criminals and their facilitators exploit the global marketplace

Drawing on evidence from select Hezbollah and Al-Shabaab financing networks, we argue that the application of Social Network Analysis (SNA) to the study of terrorist financing and money laundering advances the current state of knowledge in this notoriously difficult-to-study field.

Our study shows a network’s structure matters because it dictates the flow of resources and information. The Hezbollah and Al-Shabaab networks display the characteristics associated with fundraising networks: hub structure, brokers with high betweenness centrality and low degree centrality, international linkages, no intent to commit domestic attacks, and remittances to the home country.

Knowing that fundraising networks will conform to this structure gives policy makers and law enforcement important tactics to deter and discourage their activities. Fundraising networks are vulnerable at the hub, but resilient against traditional counter-terror measures that target hierarchies. They tend to compensate for the relative vulnerability of their structure by relying on strong ties with pre-existing acquaintances.

Instead of operating as hierarchical organizations, with orders flowing from a figure at the head down through the network, this article reinforces the view that terror networks should be conceived for what they are and how they work, and not solely according to their formal structure.

While the political party and semi-governmental organization of Hezbollah in Lebanon may follow a more hierarchical organizational structure, illicit networks supported by Hezbollah in North America are able to maintain their secretive and stealthy nature precisely by adopting a more informal and flexible structure.

Regional integration in Africa


More than 50 years have passed since the Organisation of African Unity (OAU) was created in Addis Ababa on 25 May 1963. Yet, despite the many charters signed, the numerous meetings called, and the countless strategic frameworks drawn up on paper, regional integration has led to disappointing outcomes.

In the newly published Routledge Handbook of African Development edited by Tony Binns, Kenneth Lynch and Etienne Nel, I discuss the many factors that explain why progress made towards effective regional integration has been slow on the continent.

The chapter focuses particularly on the gap between regional integration as an institutional project (also known as regionalism) and regional integration as an everyday reality (regionalisation).

Generally speaking, regional integration in Africa has achieved the highest objectives when institutions have targeted very specific areas, such as the environment, have been heavily supported by external donors and member states, and have brought together countries that share a similar currency, such as the CFA franc.

Progress toward regional integration in Africa has been slow for several reasons. In a patrimonial system that nurtures inter-personal relations, many countries have few incentives to effectively engage in deeper institutional integration with their neighbors.

Regional integration also has a disappointing record due to the large number of organisations that exist with similar or competing purposes. Nearly all African states belong to multiple regional groupings, which lead to high co-ordination costs, competition between policies, and confusion among international donors.

Finally, regional integration has been greatly affected by political crises and conflicts, as in the Great Lakes region or the Horn.

Mapping rice trade networks


Women play a key role in West Africa’s food economy. Yet, the functioning of their trade networks is still relatively unknown. To better understand how women do business in the region and what obstacles they face, the OECD Sahel and West Africa Club asked us to map the network that connects the actors involved in the production, transport or sale of locally-produced rice between Benin, Niger and Nigeria.

Our data were collected in the Dendi border region by our colleagues Dr. Lawali Dambo, Dr. Moustapha Koné and their team at the University of Niamey. It took more than three months to interview 800 producers, traders and wholesalers in three countries. Two more months were needed to structure the data and produce what is probably one of the largest supply chain networks ever mapped in West Africa! In the next few months, we’ll analyze the data in order to highlight gender disparities and border effects.