WESTERN BURKINABE POWER association leader Karfa Coulibaly temporarily hung a poster from Burkina Faso’s 2005 presidential campaign in his workroom some eight months after the November election (fig. 1.1). Amid an array of animal horns, bird feathers, and plant matter, the poster of Blaise Compaoré, who, in 2006, had been Burkina Faso’s head of state since October 15, 1987, hints at influential networks and potent knowledge that power association leaders cultivate. Suggesting the relevance of power association activities beyond rural communities, the poster connects Coulibaly visually to a man whom Burkinabe citizens recognized as wielding immense power. Political posters—like animal horns, bird feathers, plant matter, and other materials prevalent in the assemblages and installations that Coulibaly and other specialists create and manage—were difficult to obtain in the country’s rural communities in the early 2000s. According to Coulibaly, Drissa Traoré, a Kono power association leader who lived in a nearby town, procured party posters from his older brother, the national politician and international diplomat Mélégué Traoré.¹ The politician belonged to the Congrès pour la Démocratie et le Progrès, the party of Compaoré. After receiving the posters, Drissa Traoré apparently gave one to Coulibaly, who had studied Kono with the younger Traoré. The poster hanging in Coulibaly’s workroom thus attested to his connection with the Traoré family and his position within a network of prominent individuals. Indeed, power associations foster connections among experts. Through exchanges with other experts as well as observation and experimentation, specialists expand their knowledge of tangible materials and intangible energies. They use the knowledge they develop to effect change in their own lives and in other people’s lives. Power associations’ reach also extends beyond local spheres to national, transnational, and international realms.

I follow Africanist art historian Patrick McNaughton in using the English phrase power association to refer to what Mande-language speakers call jo (sg.; jow, pl.) (McNaughton 1979a, 8–17; 2001, 167–173).² Jow are organizations that promote the exchange of potent knowledge among specialists. For more than a century, European and American observers have written about power associations and their arts in a vast area that spans across western West Africa. With no exact English or French equivalents for jow, authors have used a variety of terms or phrases to approximate its meaning. Authors writing in English have referred to jow as fraternities (see Strother 1993, 178n5), initiation associations (e.g., Brett-Smith 1994; Diamitani 1999; see also McNaughton 1979a), secret associations (e.g., Brett-Smith 1997; Diamitani 1999), secret societies (e.g., Goldwater 1960), secret initiation societies (e.g., Imperato 2009), and societies (e.g., Bangali 2002). Authors writing in French have characterized the organizations as associations socio-religieuses (e.g., Zobel 1996), associations religieuses or pseudo-religieuses (e.g., Delafosse 1912, vol. 3), confréries (e.g., Labouret 1934), cultes (e.g., Binger 1892; Zeltner 1910; Colleyn 2009b), cultes à Satan or cultes des boli (e.g., Henry 1910), cultes de génies (e.g., Delafosse 1912, vol. 3), grandes confréries (e.g., Monteil 1924), sociétés d’initiation (e.g., Zahan 1960; Dieterlen and Cissé 1972), sociétés initiatiques (Colleyn and Jonckers 1983), sociétés religieuses (e.g., Tauxier 1927), and sociétés secrètes (e.g., Le Barbier 1906; Travélé 1929).³

The English and French terms and phrases variously highlight the organizations’ male orientation, religious nature, or secrecy bias, or they suggest some sort of formal or graded initiation processes. Each of these attempts to translate jow into English or French points at some important feature of the organizations. But some of the phrases are laden with negative connotations or perpetuate images of isolated African communities with secret or mysterious religious or initiation practices. In her analysis of organizations in present-day Sierra Leone and Liberia that observers have characterized as “secret societies,” historian Nanina Guyer (2015, 34–56) examines how the phrase is connected to organizations charged with fostering political instability in Europe in the eighteenth and nineteenth centuries. An organization deemed a secret society may raise fear or suspicion.⁴ While secrecy is an important dimension of power associations, neither the phrase secret society nor any of the above-listed glosses in English or French captures the core operations of jow.⁵ Men are usually, but not always, leaders of power association chapters, and being a man constitutes a criterion for viewing certain power association arts. Yet women contribute to the organizations and participate in their performances. The organizations’ leaders are specialists who manage resources in this world and in the otherworld in order to accomplish their objectives, namely, to achieve some difference in their lives or other people’s lives. They seek to safeguard the potent knowledge they cultivate, and they restrict access to it. The specialists also gain greater understanding over time, as they learn from each other and from difficult cases they confront.

The phrase power associations reflects another effort to translate the term jow. Rather than use an English phrase for the Mande-language term, I could insist on using jow throughout my text. However, I would still need to explain to English readers what jow are, and thus I would still need to attempt to translate the concept. My use of an English phrase as an approximate equivalent points at the fact that my writing about jow and their arts depends on my own understandings of terms, concepts, events, and interpretations. My aim is to explain to readers why I select certain terms or how I arrive at certain conclusions. As I see it, the phrase power associations does not carry with it the same negative historical connotations that phrases like secret societies or initiation associations carry with them. It also highlights two concerns of jow that I have found are central to the institutions.

The term power in the phrase power association refers to a complex concept that broadly means the capacity to effect significant change, either positive or negative (see also Arens and Karp 1989). When Arnold Rubin (1974) considered the concept in an analysis of accumulative arts from the African continent, he asserted that such works reflect individuals’ efforts to display status or manage powerful energies. Rubin identified boliw (pl.; boli, sg.), the Mande-language term for accumulative objects that power association leaders create and manage, as distinctive examples of their makers’ studied configurations of potent matter and energies. Indeed, power association leaders are and have long been specialists who concentrate energy in objects as they attempt to harness worldly and otherworldly resources to achieve their goals. The works that the organizations’ leaders create and manage are thus power objects. And in a way that recalls Rubin’s understandings of power and accumulation, McNaughton uses the phrase power associations to refer to institutions known in Mande languages as jow.

The term association reflects the fact that each organization consists of a group of experts who share certain foundational knowledge. Power association members pursue additional knowledge as they work to help themselves, their clients, and their communities. Their expertise, activities, and networks distinguish power association members from other people who seek similar knowledge on their own or who never try to acquire such knowledge. When McNaughton (2001, 167) characterizes power associations as “institutions that wield power by concentrating it into material objects and launching it, under control, out into the world,” he emphasizes the abilities of the organizations’ members to manage potent energies as well as the importance of power objects to the associations’ activities. Comprising some of the most important tools that specialists use to do their work, the objects that power association leaders construct and manage reflect their investments in the knowledge, skills, and materials necessary to meet their own needs and those of their constituents.

Specialists and other community members recognize different types of power objects. Burkinabe art historian Boureima Tiékoroni Diamitani observed a general distinction between speaking and nonspeaking power objects. He explains that “the speaking power objects ‘speak’ directly to people through a masquerade, trances, and divination,” and he recognizes Komo as “the most fearful and dangerous of all the speaking power objects” (Diamitani 2004, 434; see also Diamitani 1999, 60–62; cf. Jespers 1995). Power association leaders and other community members I interviewed similarly referred to power objects that speak, including Komo, Kono, and Wara objects, and others that do not speak.⁶ A specialist’s design for one power object may further distinguish it from other power objects he or one of his colleagues constructs.

Nyama is the Mande-language term for the “activating energy” that power association leaders and other specialists learn how to wield (McNaughton 1979a, 23). As I prepared to conduct research in western Burkina Faso, I frequently encountered the term in publications.⁷ I thought the term or a Senufo-language equivalent would emerge repeatedly during my informal conversations and formal interviews with power association leaders and other inhabitants of the area.⁸ The exchanges proceeded differently. Most often I heard people mention nyama or an equivalent term when they talked about the potentially dangerous energy released when an animal was killed.⁹ They noted that people who hunt wild animals take precautions to avoid illness or death provoked by the nyama of an animal released at the moment of death. The late Missa Coulibaly, a prominent hunter in the town of Noumoussoba, specified that the potential for a wild animal’s nyama to cause harm to a hunter or someone close to him promotes respect among hunters.¹⁰ Coulibaly explained that a hunter can never know the extent of another hunter’s knowledge and may one day need to call on a colleague for help counteracting the devastating nyama of a particular animal.

I listened for references to nyama without insisting people talk about it, as I did not want to lead people into expected answers. Occasionally, I heard the term nyama or a Senufo-language equivalent in statements about activities other than hunting. An older hunter and town official in Dakoro, Tayirigué Ouattara, defined the concept as the image that rests in someone’s head after the person has seen a horrible accident.¹¹ An elder sculptor-blacksmith in Sokouraba, Bè Coulibaly, said during an early conversation that nyama referred to garbage.¹² A younger hunter in Kangala, Aly Traoré, explained that the fast music hunters’ musicians play during a performance creates a hot situation, one filled with nyama.¹³ And Komo leader Yaya Bangali referred to the nyama that increases during a power association performance, even though the only definition of the concept he offered during a previous interview was that of the force in an animal that lingers after it is killed.¹⁴ The different explanations I heard led me to conclude that the concept of nyama exists in western Burkina Faso, even if people did not talk about it in my earliest conversations with them, as I had anticipated they would.¹⁵ Only later did specialists in western Burkina Faso offer descriptions of nyama that coincided more closely with the scholarship on the term I had previously read.¹⁶

The prominence of power-packed objects in power association leaders’ work has led some observers to refer to the organizations as fétiches, or “fetishes” (e.g., Henry 1910). The characterization focuses attention on power objects rather than on the experts who create and work with the assemblages. It also disregards the individual agency, specialized knowledge, and interpersonal exchange among experts necessary for the production and maintenance of any object. Power association leaders I interviewed consistently told me that no object realizes change on its own. They said that the effectiveness of an object depends on the knowledge and skill of the person who makes or attempts to use it to attain a particular goal.¹⁷ A power object is thus a tool a specialist uses, rather than an object that acts on its own in the world. Writers have at times questioned the utility of referring to power objects in different parts of the world as “fetishes,” in part to recognize the agency of the objects’ makers.¹⁸ Attention to objects’ makers coincides with understandings of power objects that experts in western Burkina Faso conveyed to me.

Museum curators, other scholars, and private collectors in Europe and North America who, in the twentieth century, endeavored to classify power association objects that entered European and American collections identified power association arts as specifically Bamana or Mande, thereby implying that the objects’ makers, patrons, and audiences belonged to a bounded Bamana or Mande cultural or ethnic group. Curators, other scholars, and private collectors also attributed the objects to different organizations, including Ciwara, Komo, and Kono, and they regarded the organizations as Bamana or Mande. Despite the attention that art connoisseurs and other scholars have paid to the Bamana and Mande labels or to certain related phenomena, observers have recognized a range of power associations across a vast and diverse area that today spans from Guinea to Burkina Faso and from Côte d’Ivoire to Senegal.

An inventory of all the power associations in this area in the past or in the present remains impossible to construct. As McNaughton reports, “In a hundred years of publications, authors’ lists [of power associations] are wonderfully divergent” (2001, 172). Orthographies for the names also vary. The variety suggests that for distinct reasons at different moments, certain communities have invested in or divested of particular power association chapters. In addition, various observers may have had unequal access to information or may have understood their findings in disparate ways. A complete inventory of all power associations in a single area or across a broad region through all moments in time may be impossible to construct. When I began my research in western Burkina Faso in the early 2000s, the most prominent power associations that people in Sokouraba and nearby towns supported included Komo and Kono as well as Nama, Nya, and Wara, and several chapters of the same association at times appeared within a single town. Hunters’ associations, or dozo tonw, constituted another important power association in the area at the time. Dahaba Ouattara has reported that Komo, Kono, Nama, Nya, Wara, and hunters’ associations have continued to operate in the area.¹⁹

In this chapter, I draw on colonial publications as well as unpublished archival information to demonstrate that there is an historical basis for understanding power associations as organizations of culturally, ethnically, linguistically, and religiously diverse specialists dedicated to the development of powerful knowledge to cause change. This finding disrupts the notion that describing the organizations as specifically Bamana or Mande captures their essence. It also leads to the recognition that a history of power associations and their arts must acknowledge local-level continuities, changes, and exchanges as well as colonial and postcolonial experiences connected to the organizations, their activities, and our knowledge. Even if scholars cannot recover details of local histories, we can allow for the possibility that a lack of documentation does not mean an absence of specific histories. We can thus signal in our writing the importance of such histories, even if we cannot trace them at this time. And we can move beyond the idea that an effort to describe precolonial, “pure” African cultural phenomena based on information gathered in colonial or postcolonial contexts constitutes the writing of histories (see also Vansina 1984). Rather, we must recognize how colonial and postcolonial contexts are entangled with our sources and understandings.

Power Associations and Colonial Surveillance

One of the earliest descriptions of a Komo performance to appear in print dates to 1885, when French military officer Joseph-Simon Gallieni published his account of a French government-sponsored mission through the Bélédougou and Ségou regions of present-day Mali between 1879 and 1881. The mission contributed to the French government’s larger effort in the late nineteenth century to assert its authority over the three-corner region and larger areas of western West Africa. Burkina Faso (formerly Haute Volta), Côte d’Ivoire, and Mali (formerly Soudan français) achieved independence from France in 1960. The presence of Komo in Gallieni’s narrative, written in the late nineteenth century, as well as in additional reports by French officials indicates that the organization existed by the mid-nineteenth century, if not earlier.²⁰ Exact details of the origins of Komo or any other power association remain undetermined, although different explanations circulate (see also chapter 4). French colonial administrator Maurice Delafosse (1912, vol. 3, 175) posited that Komo, also known as Koma, was universal and had existed for millennia. Without citing his sources, he suggested that Koma or some version of it appeared to operate across West Africa and in the basin of the Congo River (Delafosse 1912, vol. 3, 123, 174).

As French officials and missionaries imposed their authority in vast areas of West Africa in the late nineteenth and early twentieth centuries, they recognized the importance of power associations to local populations and focused attention on the institutions. At times, they viewed power associations as potential threats to their political and religious missions.²¹ Catholic missionary Joseph Henry published a foundational text about populations he recognized as Bamana and their religion in 1910, and he dedicated significant space in his book to the analysis of power associations and their activities. Henry describes a variety of boliw and the organizations, or “cultes,” linked to the objects, including “Tji Wara” (Ciwara), Komo, Kono, and Nama (1910, 127–156). Consistent with his view of boliw as “fetishes,” the missionary explains that each boli becomes “a person, an individual” through its construction (1910, 128). He regards the objects with disdain, writing that they “are nothing less than intrinsically evil beings in revolt against the Supreme Being, the Creator God and Master of all things” (1910, 133). With a similarly dismissive tone, he characterizes the population he identifies as Bambara as one attached to its condition and not eager to change (1910, 9). Henry further notes that he hopes that the information he published would aid missionaries as well as colonial officers, traders, and travelers in their activities (1910, 11).

Given the goals of government officials and religious missionaries across western West Africa in the early twentieth century, perhaps it is no coincidence that they evaluated power associations with concern. While their analyses vary, the foreign observers often validated their suspicions, at times only tacitly, and offered assessments of how to manage the organizations they considered concerning or threatening. Writing in the first years of the twentieth century, Louis Le Barbier, who served as a chargé of the colonial mission in French West Africa (Afrique occidentale française [AOF]) before 1918, asserted that power association chapters “have literally terrorized all of the Bambara cantons” (1906, 17). The evaluation insinuates that he did not consider the organizations beneficial. He adds, “They have lost, due to contact with [French] civilization, much of their strength, but without ceasing to be still powerful” (Le Barbier 1906, 17). He thus implies the possibility that the organizations’ activities might continue to wane as French presence in the region waxed.

Throughout the first half of the twentieth century, government officials monitored power associations’ activities in their attempts to gain control over the organizations and the region. Occasionally, they may have considered ways to incorporate the organizations into their structures. French administrator Louis Tauxier (1927, 281–282) notes that in the area of Ségou, a local tribunal instituted by the French government required non-Muslims to swear oaths over Komo, whereas Muslims vowed to tell the truth over the Koran. The French administration appears to have accepted the importance of Komo rather than dismissing it in an area where some individuals were Muslim and others were not, at least in some instances.²²

A challenging dispute brought a former French colonial administrator by the name of Roland Colin into a memorable encounter with Nya at the end of 1953 and beginning of 1954.²³ From 1952 to 1954, Colin worked in Sikasso; I spoke with him more than fifty years later for my research. After he received news of an intense land conflict between residents of Diélé and “Nienzérebougou,” two towns near Kignan, Colin, his guard, Maliki Sidibé, his interpreter, Bakary Diallo, and his wife, Renée Colin-Noguès, traveled to the two towns to call a meeting of people involved in the disagreement. Colin (2004, 296) reports that he wanted to defuse the situation before someone committed a criminal offense. He asked residents of each town to tell their side of the story. In the colonial administrator’s observation, the accounts on each side seemed consistent, and the position of each side seemed plausible. Colin states he did not know how to proceed.

Diallo, the interpreter, reportedly asked to talk with Bénogo Diarra, the complainant from Diélé, and Sirikoro Bangali, the complainant from Nienzérebougou. Diallo returned to Colin to explain that Diarra and Bangali agreed to turn to the Nya power association to resolve the situation. Reasoning that the oldest Nya chapter in the area was the chapter of Diélé, over which a man named Fadio Diarra presided, the men agreed that Fadio Diarra would oversee the proceedings. Someone brought the chapter’s power objects to the base of the tree where the Nya leader and other men had gathered. Bénogo Diarra swore an oath affirming that if he harvested millet in the contested land but had not planted it, he would die if he ate the millet. Bangali swore the same oath after Diarra. As soon as the two men finished their oaths, the tension dissipated.

Colin learned later that after the men gathered before the Nya power objects of Fadio Diarra, Bénogo Diarra tried secretly to sell the millet he gathered from the contested land to a merchant. However, people in the two towns were aware of the oaths, making it impossible for him to offload the grain. His actions made his culpability clear. In response, leaders of a well-known Komo chapter in Nienzérebougou prohibited blacksmiths, the makers of farming equipment, in the town and any nearby towns from touching any agricultural tool for any resident of Diélé, where Bénogo Diarra resided. No blacksmiths lived in Diélé, and residents of the town eventually presented offerings to the Komo chapter in Nienzérebougou. Whether the incident marked the first time a colonial administrator ever turned to a power association to resolve a problem, as Colin stated in conversation with me, Colin’s account of it demonstrates that at least one colonial administrator remained open to local conflict-resolution methods and the activities of power associations.

At other times and in other places, officials and missionaries attempted to undermine power associations’ authority, bolster French influence, and discourage or contain resistance to colonial rule or Catholicism. In 1918, Le Barbier clearly expressed his concern about the influence of power associations in western West Africa: “We must not trust [secret societies including Kono and Nama] and keep our guard with respect to them: between hostile and skilled hands, they are capable of being a formidable instrument of resistance to our empire” (Le Barbier 1918, 19).

As French awareness of the organizations and their arts grew, colonial administrators considered how these groups used assemblages, installations, masks, and performances in their operations. In his 1924 publication on populations of Ségou and “Kaarta” that he recognized as Bamana, French administrator Charles Monteil reasoned that destruction of a boli did not usually result in the elimination of the “culte” or association that created the object (Monteil 1924, 254–255). Monteil’s statement may obliquely refer to repeated French colonial efforts to destroy power association objects or French anxieties around the relationship between the organizations and their objects.

By the early twentieth century, French visitors to the three-corner region confiscated or oversaw the seizure of certain power association objects. In 1910, Fr. de Zeltner published an article describing the works that he donated to Paris’s Musée d’Ethnographie du Trocadéro and that I consider in the next chapter. According to Zeltner, Muslims and non-Muslims in an unnamed town in the cercle de Bamako complained that the objects’ owners, who reportedly led a Nama chapter, engaged in extortion and intimidation. Zeltner writes that local guards succeeded in taking the objects “without any difficulties” (1910, 364). The account raises the possibility that the guards worked under the direction of French authorities. Zeltner adds that the action restored calm. His statements further suggest that he considered acquisition of the objects justified and even beneficial for the community, at least from a French perspective, as the act undermined the authority of Nama leaders who purportedly disturbed local communities.

Confiscation of three power association objects in a town near Bamako may not have been an isolated instance of a colonial intervention designed to diminish the influence of a power association chapter. According to colonial reports, the French administration at times destroyed power association objects. A 1911–1913 report on the town of Gongasso, near Sikasso, in present-day Mali indicates that the community once had owned a branch of “Komon,” presumably a reference to Komo. A commandant ordered destruction of the town’s Komon by fire in 1910.²⁴ The report does not specify the catalyst that sparked the iconoclastic act, but perhaps colonial administrators sought to rein in activities of power association leaders in order to minimize potential threats to French authority, or maybe they were concerned about local conflicts.²⁵

Government officials and Catholic missionaries at times monitored power associations and attempted to dominate them, a perspective that appears even in Gallieni’s early account of the Koumou performance at Sibi. In 1927, more than forty years after Gallieni traveled through Sibi, Tauxier insisted on the French government’s command over Komo, then recognized as a formidable organization in the AOF. Tauxier described two examples of nineteenth-century Muslims overpowering Komo—one through the sale of Komo performers into slavery and the second through force of a gun. But Tauxier also stated that many Komo leaders resisted Muslim incursions, hinting at the organization’s tenacity. He indicated that by contrast, Komo leaders submitted to the crushing force of the French colonial administration. He explained that the organization’s leaders recognized French authorities’ capacities to defeat them and preferred to avoid conflict. Tauxier concluded, if erroneously, that in principle Komo leaders only performed religious functions and that they could retreat to that position under French rule (Tauxier 1927, 299–302). By relegating Komo to the domain of religion, Tauxier attempted to diminish the organization’s political import and potential threat to the French government.

Throughout the colonial period, French officials tried repeatedly to determine what exactly Komo and other power associations were and how they operated. They sought general information and occasionally focused on specific descriptions of certain chapters or leaders. They also continually evaluated the colonial administration’s relationship to the organizations as well as the groups’ impact on local activities. In his preface to an article on Komo by Moussa Travélé, identified as an interpreter for the colonial administration, French administrator Henri Labouret urges colonial officers to understand “the languages, institutions, customs, [and] mentality” of local populations (Labouret in Travélé 1929, 127). Labouret seems most concerned about the stories of allegedly murderous or otherwise disruptive activities of Komo leaders that Travélé details in his text (Travélé 1929, 146–150). One of the stories displays the naiveté of the colonial administration even as it provides a glimpse of local-level politics. As Travélé describes it, Komo practitioners in an unnamed town fooled white authorities into thinking that the Komo practitioners’ Muslim rivals sought to revolt against the colonial administration. Yet according to Travélé, a Komo member’s conversion to Islam precipitated the conflict that in turn led to the Komo practitioners’ accusations of insurgency (1929, 148–150).

Several years later, in 1931, French ethnographer Marcel Griaule led the Mission Dakar-Djibouti through the three-corner region, where Michel Leiris and Éric Lutten joined him in reportedly stealing boliw from power association houses (Leiris [1934] 1981, 102–105; [1934] 2017, 152–156). The collecting mission contributed to Griaule’s interest in articulating philosophical and cosmological underpinnings of arts from present-day Mali. He subsequently worked with French ethnographer Germaine Dieterlen, and she collaborated with Malian scholar Youssouf Tata Cissé. Dieterlen and Cissé attempted to explicate a complex cosmology for Bamana Komo associations and to demonstrate the organization’s philosophical sophistication (Dieterlen and Cissé 1972). Scholarly interest in the foundations, cosmologies, and traditions of power associations grounded the organizations in an unchanging ethnographic present (see also Colleyn 1988, 79–83). It cast the associations as politically impotent and diverted attention from power association leaders’ acute abilities to respond to and negotiate their everyday circumstances in the present. Framing power associations as Bamana or Mande institutions further denied the organizations’ political importance (Amselle [1990] 1998, 14–15; see also Amselle [1990] 2010, 54; cf. Förster 2018, 136–137).

Some scholars viewed power associations within a town as institutions that formed a harmonious whole. In his book The Bambara, Dominique Zahan, who studied with Griaule, identified N’domo, Komo, Nama, Kono, “Tyiwara” (Ciwara), and Korè as six jow or “initiatory societies” that together “constitute an organic whole” (Zahan 1974, 15–24). Yet the divergent lists of jow that Zahan and other observers in and beyond the three-corner region have recorded suggest there is no single coherent system (see also Diamitani 1999, 44–45; Austen 2007, 38). Their various configurations demonstrate that power associations do not—and may have never—fit together to form a whole. A single town can support more than one chapter of the same organization and no chapter of other organizations (see also McNaughton 1979a, 20–21). For instance, by June 1953, the town of “Nagna” near Sikasso in present-day Mali apparently supported three Nya chapters and two Wara chapters.²⁶ In the early 2000s, some towns in western Burkina Faso supported multiple chapters of the same association, whereas other towns did not support any.

The variability of power association chapters within different towns and from town to town reflects the particular concerns and ambitions of certain individuals operating within distinct contexts. Even the missionary Henry recognized variation in each chapter’s history, explaining that “one knows the origin of each [chapter], and each initiate, even if little instructed in its rites, will tell you if it comes from the north or the south, the east or the west, if it is male or female, and every person with a white beard can draw the family tree” (Henry 1910, 127). Now, as in the past, certain people seek to install a power association chapter, join one, or otherwise invest in one when misfortune or some other significant event befalls them, their families, or their communities. Depending on their resources and goals, they decide which chapters they would like to link themselves to and which experts they would like to consult.

Marketplaces of Power

Aspiring power association leaders decide to pursue knowledge of an organization and to work with a certain chapter within what historian Gregory Mann describes as an “extremely active spiritual marketplace” (2003, 271). Mann’s observation pertains specifically to the area around San in present-day Mali in the mid-twentieth century, but across western West Africa, specialists have competed and continue to compete for clients and for followers, and they have appealed and continue to appeal to potential clients and followers from diverse cultural or ethnic groups. Such marketplaces of power promote exchange and competition.²⁷

Power association leader Karfa Coulibaly’s statement that he possessed a mask “more powerful” than the one in Anita Glaze’s photograph signaled that distinctions familiar to me through the publications I had read may not be so clear-cut. I had traveled to western Burkina Faso to study Komo and Kono in communities recognized as Senufo. Coulibaly lived in such a community and identified as his first language Sìcìté, which is classified as a Senufo language. Before my exchange with Coulibaly, I imagined he might tell me he owned a helmet mask similar to ones that art connoisseurs and other scholars have attributed to Kono and that appeared in some of the other photographs I carried with me. I had not thought he would associate the mask in Glaze’s photograph with Komo or Kono, because Glaze calls the masquerade Kunugbaha and links it to Poro, which, on the basis of my reading, I then understood as a separate phenomenon (see fig. intro.9). But clearly Coulibaly saw his own work in relation to and even in competition with the mask that Glaze photographed and identified with Poro male initiation.

While observers report the presence of Poro in diverse communities across western West Africa, authors including Glaze have recognized Poro as an organization at the core of Senufo communities.²⁸ The capital P on the term when written in English suggests that Poro is a proper noun that designates a specific institution. Elsewhere I offer a different assessment, namely, that poro is a term that describes a type of institution, one that promotes the exchange of powerful knowledge and the safeguarding of that knowledge, rather than naming a specific institution. The term may refer to a compulsory male initiation association within a community or a voluntary association (Gagliardi 2014, 185–199). As a word that designates a type of organization rather than names a specific organization, poro requires a lowercase p.

A general understanding of poro allows for the possibility that the term poro encompasses Kono and other power associations and organizations. And if Kono is a poro organization, then Coulibaly’s suggestion that he owned a mask more powerful than the mask the performer in Glaze’s photograph wears no longer seems to hinge on a comparison of different things. A closer look at arts attributed to organizations in the region further signals that forms associated with Kono, Komo, and other poro organizations are not always distinct. Komo chapters invest in helmet masks characterized by gaping jaws, sharp teeth, pointy horns, and elongated ears. Komo, like certain other poro organizations, supports performances that fight crime, including malevolent sorcery, and that address other challenges individuals and communities face. As we will see, the distinction between benevolent and malevolent action and thus between the goodwill of a power association leader or other specialist and the hostility of a malicious or criminal sorcerer depends on one’s position relative to a particular action or series of actions (see also McNaughton 1979a, 9–10; 1988, 11–14, 42–50; 1995).

A photograph in a collection of images attributed to French Catholic missionaries Gabriel Clamens and Michel Convers, now in a private collection in Belgium, shows masqueraders wearing masks that art connoisseurs, scholars, and observers might associate with Komo (fig. 1.2). Seven men appear in a line, and several of them stand or sit with musical instruments. Two men wear helmet masks with wide-open maws atop their heads, leaving their faces visible. Feather-covered garments cover their bodies. The men’s masks resemble the performer’s mask in two photographs identified as “Komo mask dancing” and reproduced in Zahan’s book The Bambara.²⁹ Handwritten words on the back of the photograph now in the private collection in Belgium offer different information. They indicate that the photograph illustrates poro masqueraders from “Dyawala” (Diawala), a town in present-day Côte d’Ivoire located about sixty miles northeast of Korhogo and in the broad area often regarded as Senufo.

In 2004, when I was beginning my research on power associations and their arts, I had not yet arrived at these conclusions. When I returned to Los Angeles after my first visit to Sokouraba, I looked back at Glaze’s Art and Death in a Senufo Village to see if I could better understand how Kunugbaha might relate to Kono or Komo. According to Glaze, Kunugbaha masquerades belonged to blacksmiths, a group often linked to Komo. She also indicates that Kunugbaha performances likely differed from other “Poro helmet masquerades” (Glaze 1981, 142, 207–209).³⁰ Spotting an endnote marker at the conclusion of one of her paragraphs describing different masquerades, I turned to the back of the book to read the additional information she had to share (Glaze 1981, 237). Citing a personal communication from April 1978, Glaze explains, “Patrick McNaughton’s data on the blacksmith masks of the Bamana [and here she seems to mean Komo masks] suggest that there is a close historical relationship between Komo masks and the Senufo blacksmiths’ helmet masks.” She thus gestures toward a link between Komo and Kunugbaha. She adds, “Though this is not the place for comparative analysis, there is good reason to believe that the connectives can be reconstructed on the basis of form and content.”³¹ Glaze’s reference to her communication with McNaughton further encouraged me to pursue my questions.

FIGURE 1.2. The back of the photograph reads, “Dyawala [Diawala, Côte d’Ivoire], Poro.” Unrecorded photographer. Image courtesy of the Archives of A. R. Arthur, Belgium.

The communication with McNaughton that Glaze cites anticipates themes developed in “African Borderland Sculpture,” the College Art Association panel McNaughton organized in February 1987 and the eponymous essay he published in the August 1987 special issue of African Arts dedicated to the same theme. McNaughton writes, “Studies of African sculpture style are most fruitful when they include explorations of cultural change and artists’ motivations. Politics, commerce, the stature of sculptors, and the relationships between sculptors and clients have all been shown to shape style, and a growing body of research demonstrates that it is no longer possible to view the shape of African art as dictated by hermetically sealed geographic, ethnic, or traditional boundaries” (1987, 76). In the same essay, McNaughton questions the relationship between so-called centers and peripheries. He proposes looking for the borderland, namely, the “conceptual space where forms and ideas of diverse origins are contemplated and rearranged, where creativity is at a premium and its enactment results in lively history” (McNaughton 1987, 77; see also Frank 1987, 2007, 2021). McNaughton’s conception of the borderland seemed to apply to Sokouraba and environs, and it helped me justify my decision to locate my study of the arts of Komo, Kono, and other power associations long framed as Bamana or Mande in a Senufo-speaking neighborhood of Sokouraba. The town is located beyond the “Mande heartland,” or center, at the borders of present-day Mali and Guinea, and it is also located beyond Korhogo in present-day Côte d’Ivoire, a city framed as the center of Senufo-ness since the early twentieth century.

My research takes as its starting point the idea that focusing on the Bamana-ness, Mande-ness, or Senufo-ness of power associations and their arts overshadows the dynamic exchanges and specific contexts that shape and are shaped by the organizations and their arts. I find that Komo, Kono, and other power associations in western West Africa are organizations that foster the exchange of powerful knowledge across vast interpersonal networks. The organizations’ leaders use their knowledge to address individuals’ and communities’ concerns, fight crime, and otherwise ensure well-being. Aspiring power association leaders seek to study with the most knowledgeable experts, and a most knowledgeable expert may or may not be someone who shares a common background with an aspiring power association leader. Similarly, my desire to learn from any of the people who have taught me has had nothing to do with my interests in learning from people who share a cultural or ethnic background with me. Rather, I have wanted to study with the most knowledgeable experts I could access in various locations. Power association leaders likewise develop expertise as they navigate collegial, and at times adversarial, relationships that transcend cultural, linguistic, religious, and geopolitical boundaries. Through exchange, collaboration, and competition, power association leaders maintain and promote assemblages, installations, and performances central to each local chapter’s identity and activity.

In an overview of power associations published in conjunction with Bamana: The Art of Existence in Mali, a 2001 exhibition that opened at the Museum Rietberg in Zurich and the now-defunct Museum for African Art in New York, McNaughton states, “We can never reconstruct the history of all the power associations” (2001, 173).³² He remarks that “[this inability to reconstruct history] is unfortunate, because it involved high levels of enterprise and expertise in the hands of a great many people.” However, we can return to published accounts and unpublished records to uncover details that allow us to place certain power association chapters within more specific historical contexts. Extant information is partial, fragmentary, and often impossible to verify. Still, it facilitates a shift in focus from general statements about power associations to more specific assessments. An important point to dispel is the characterization of power associations as distinctly Bamana or Mande. While authors frequently characterized the organizations as Bamana or Mande throughout the twentieth century, references to the organizations from the same period repeatedly indicate that other communities supported the institutions. As early as the late nineteenth century, French military officer Louis-Gustave Binger (1892, 81) described houses for Komo at the edges of towns he regarded as Senufo in a travelogue recounting his 1887–1889 journey through the three-corner region. Reinhard Hugershoff—who traveled to southwestern Mali as a member of the 1907–1909 Deutsche Inner-Afrikanische Forschungs-Expedition (German Inner-African Research Expedition) led by Leo Frobenius—also documented Komo and Kono in places he referred to as Senufo (fig. 1.3). Notes from the expedition record that Kono houses commonly appeared near town centers.³³

FIGURE 1.3. Helmet mask identified as “Mask of Komma [probably Komo] of the Senufo, Pig’s Head,” as observed and possibly acquired in 1907–1909. Ink drawing on paper by an unidentified artist [possibly Reinhard Hugershoff]; 12.7 × 16.7 cm. Frobenius-Institut an der Johann Wolfgang Goethe Universität, Frankfurt/Main, KBA 10183. © Frobenius-Institut.

Once we understand that the identification of discrete cultural or ethnic groups reinforces colonial frameworks and tends to disregard historical dynamism, then any colonial reference to a particular cultural or ethnic group requires extra scrutiny. Still, Binger’s, Hugershoff’s, and other colonial-era observers’ references to power associations in communities that they recognized as neither Bamana nor Mande indicate that there is a mismatch between the tenacious idea that power associations and their arts are distinctly Bamana or Mande phenomena and the historical record, which clearly links the organizations and their arts to other labels. The discrepancy may hinge on colonial logic, but it also points at weaknesses in such faulty logic that imagines organizations and arts as well as people and knowledge necessarily cling to a single cultural or ethnic group.³⁴ I return to this point again and again because the idea that power associations are specifically Bamana or Mande institutions and the related notion that the organizations’ arts are Bamana or Mande endure in present-day presentations of the organizations and their arts, despite abundant evidence to the contrary.

Throughout the first half of the twentieth century, colonial administrators identified power association chapters in locales they labeled as Senufo, just as Binger and Hugershoff had. In 1911, the administrator Edgard Maguet explained that “all the Senufo have gris gris, among which the most widespread are ‘Ouara’ [Wara], ‘Konon’ [Kono], and ‘Gomon’ [probably Komo].”³⁵ Four decades later, colonial administrator Colin signed a set of reports detailing information he gathered in towns near Sikasso.³⁶ The reports document the presence of “fétiches,” a term that seems to refer to power associations and perhaps particular power objects, in communities with populations that the official recognized as predominantly Senufo. The terms Tyiwara (Ciwara), Komo, Kono, Wara Mousso or Nia (Nya), and Waratié (Wara) appear on his lists, as do other names that I did not encounter in western Burkina Faso. For example, a May 31, 1953, report on the town of Wakoro identifies 826 of the town’s 875 inhabitants as Senufo, at least 116 of whom are described as Senufo “of the blacksmith caste.”³⁷ The document refers to several power association chapters in the town, including a Komo chapter that its leader, Ntio Sanogo, reportedly brought to Wakoro from “Worofora” two years earlier as well as a Tyiwara (Ciwara) chapter that the father of Manianzié Coulibaly, the chapter’s leader at the time, apparently acquired from a locale in the cercle de Ségou identified as “NTila.”

Population information in Colin’s reports points toward diversity in the towns the official visited. His repeated references to “assimilated” populations also signal the possibility for people’s identities to change. The May 31, 1953, report on Wakoro characterizes another 49 of the town’s inhabitants as “Bambaras assimilés sénoufos,” or “Bamanas assimilated as Senufos.” A report from the town of “Sourounto” (possibly Souronto) dated five days later identifies 173 of 204 residents of the town as Senufo and sixteen townspeople as “Peuls assimilés sénoufos,” or “Peuls assimilated as Senufos.”³⁸ The “assimilated” populations referred to by still other reports in the collection reinforce the idea that identities across western West Africa developed in relation to each other, often within tumultuous contexts of political expansion, consolidation, and threats predating European colonization (see also Amselle [1990] 1998). And with conversion as well as migration or changes in occupation, a person’s status could change within a lifetime (see also Launay 1982, 1992, 1999).

Inhabitants of the region have also long spoken a variety of languages. People whose first languages differed from the first languages of their neighbors may have found it useful to learn other languages or rely on a regional lingua franca to forge relationships. Jula, one of the region’s lingua francas, is classified as a Mande language. The prevalence of Jula and other Mande-language terms in the names for power associations and in the organizations’ performances may reflect specialists’ desires to communicate to the broadest audiences.³⁹

Colin’s reports offer rare and important insights into the specificities of single towns in the vicinity of Sikasso in the mid-twentieth century. They hint at more detailed histories. They evince the shifting nature of identities and further suggest that migration directions within individual towns—as well as across the greater three-corner region—varied in relation to complex, local-level concerns and circumstances. While we still need to uncover more details in order to understand local-level histories, their interconnections, and their broader relevance, data available in Colin’s reports make clear that each power association chapter in the Sikasso region in the mid-twentieth century was embedded within specific networks of relations. A sampling of details in the documents provides a glimpse of local-level histories often barely detectable in studies of power associations and their arts.

In an April 12, 1953, report on the town of Ngana in the cercle de Sikasso, Colin credits a man named Bangari Ballo, identified as Senufo, with having established the town at an unspecified moment in the past. Ballo reportedly left another town known as Ngana but located in the cercle de Bobo-Dioulasso in order to find new land to cultivate. Colin refers to the Malinké family of a marabout, or specialist with knowledge of Islam, named Fadima Kéita who apparently came from the town of Nanguila in the “pays mandé” or “Mande country” to settle in Ngana in the cercle de Sikasso around 1930. Thus, according to Colin’s documentation, Bangari Ballo migrated from east to west, whereas Fadima Kéita migrated from west to east. The same report points at additional complexities with respect to the identities and migrations of families in Ngana in the cercle de Sikasso. Colin describes the family of Lamine Sangaré as formerly Peul and locates the family’s origins to the town of Nouna, perhaps a reference to the town of Nouna currently located northeast of San and not far from Djenné. Colin adds that Sangaré’s family “lost its language and had completely assimilated to the Senufos.” And he writes that yet another family, the one of Bamana griot Flatié Sarro, traced its origins to Ségou. Migrations into Ngana were not unidirectional.

The same April 12, 1953, document about Ngana indicates that origins attributed to different power association chapters in the town do not correspond with recorded origins for families overseeing the chapters. Factors other than family origins may have contributed to decisions about the acquisition of each new power association chapter. The town’s leader at the time, identified in the report as a descendant of Ballo, the town’s founder from the Ngana near Bobo-Dioulasso, apparently presided over a “Waratie” (Wara) chapter in the town that the father of the town’s then-current leader brought to Ngana from a place called “Missala.” Relatives of the town’s leader’s family named Niana Ballo and Zanga Ballo reportedly oversaw Ngana’s “Wara Mousso” or “Nia” (probably Nya) chapter and Konon (Kono) chapter, respectively. A Zanga Ballo, whose relationship to the town’s leader remains unspecified, led the town’s Komo chapter. The document also indicates that the Nya, Kono, and Komo chapters in Ngana came from chapters in the towns of “Tankabougou (Ngolasso),” “Kouloupembougou (Ngolasso),” and “Kabala (Ganadougou),” respectively. Zié Diarra, presumably from a different family, led the town’s Tyiwara (Ciwara) chapter, which Colin mentions came from “Banko (Dioïla)” (fig. 1.4).⁴⁰ As families and power association chapters migrated in various directions and at different moments, individuals forged relations with people in disparate locales. Unfortunately, Colin recorded little information about the exact decisions informing each move or each acquisition of a power association chapter. However, it is hard to imagine that people made decisions to resettle or acquire a power association chapter haphazardly. Installation and maintenance of a chapter, like resettlement, require significant time, labor, and expense.

FIGURE 1.4. “Tyiwara de Ngana, et le Maître du culte,” perhaps Zié Diarra, n.d. (Colin-Noguès 2006, 52–53). Photo: Renée Colin-Noguès.

With chapters scattered across the three-corner region and beyond, aspiring power association leaders past and present choose what knowledge to pursue within an active marketplace of power. Some aspiring power association leaders may decide to connect to a particular chapter because it has earned renown. The names of some places appear repeatedly in the reports from the area around Sikasso that Colin completed as a colonial administrator in the early 1950s; it may be that chapters known for their successes were concentrated in those locales. One place-name that appears repeatedly in the reports as the source for a number of chapters is “Dougouni.” For example, the town of N’goloniebougou apparently supported “Dityan,” “Waratié” (Wara), and “Wara Mousso. – (Nia)” (probably Nya) chapters that traced their origins to Dougouni. One of two Tyiwara (Ciwara) chapters in Kabarasso, Waratié (Wara) chapters in Nagna and in Sourounto (possibly Souronto), and a Dotian chapter in Wakoro also reportedly located their origins in Dougouni.⁴¹ A search for Dougouni in the GEOnet Names Server database administered by the United States National Geospatial-Intelligence Agency yielded geographic coordinates for four separate populated places in present-day Mali, suggesting the place-name mentioned in Colin’s reports may or may not correspond with a single geographic location.⁴² If the Dougouni cited in Colin’s reports all correspond with the same place, then the town appears to have once fueled the spread of power association chapters across the landscape.

A report for Dougouni appears in Colin’s set and highlights other factors to consider in assessing local-level histories. Colin identifies 738 of the town’s 958 inhabitants as Bamana, another 106 of them as Bamana of the blacksmith “caste,” and 86 more of them as “Sénoufos assimilés bambara” or “Senufos assimilated as Bamana.” His classificatory terms again signal the mutability of identity. What is not clear is whether Colin recognized as Bamana people who spoke a language identified as Bamana or people who had not converted to Islam, the latter a possible meaning for the term (see Colleyn 2009a, 7–8). Interestingly, his report attests that 949 of the town’s 958 inhabitants were Muslim. The statistic indicates that conversion to Islam may not have contributed to Colin’s determination of whether a person was Bamana. In addition, Colin reports that Islam had replaced all of the power associations in the town.⁴³ Yet he provides a list of chapters, perhaps because he heard that they had operated in the town in the past. When read in the context of reports from other towns citing Dougouni as an origin for power association chapters, the information raises the possibility that people in the town may have abandoned power association chapters that once earned fame. It is also possible that townspeople who supplied Colin with information preferred to claim that the chapters had declined in prominence even if the chapters were still active.

Massa and Islam

Rural towns across the three-corner region have been connected to broader spheres for decades. Another set of reports Colin authored yields insights into political and religious activities in rural towns in the vicinity of Sikasso in the late 1940s and early 1950s. Several months after Colin began his government post in Sikasso in May 1952, he traveled through an area south of Sikasso called the Folona (see also Colin 2004; Colin-Noguès 2006). A January 2, 1953, document he signed offers an overview of observations he gathered in towns in the Folona between November 13 and December 4, 1952.⁴⁴

In one section of the document, Colin assesses the influence of political parties on people in the region.⁴⁵ He refers to “quite violent political unrest” fueled by soldiers from the area who had fought for France during the Second World War, rallied against the French colonial administration, and supported the Rassemblement Démocratique Africain, an anticolonial political party. Colin notes that tensions climaxed in 1949 and that by the time he traveled through the Folona, the political situation there had become calmer. Colin also specifies that as he traveled, he advised people to focus on cultivating millet rather than engaging in dead-end arguments. He then concludes the section with a brief reference to what he calls the Dieu de San and indicates it has a “whole different influence” on the area. He dedicates subsequent pages in the report to the Dieu de San, a movement that in various documents he references by other names, including Massa, Allah Koura, and Fétiche de San. According to Colin, people in the Folona commonly called the movement Massa, meaning “the powerful” in Jula.

The influence of Massa bears consideration in a study of power associations because people who invested in Massa, like power association leaders, sought to harness potent materials and energies in order to secure well-being for themselves and their communities. Colin’s comments indicate that Massa superseded power association chapters in importance in some towns, and residents of some towns actually abandoned or said Massa required them to abandon their power association chapters or the power objects at the core of the chapters’ activities.⁴⁶ Yet despite Massa’s influence across the three-corner region in the mid-twentieth century, power associations, rather than Massa, seem to have endured into the twenty-first century.

A man called M’Peni Démbélé but also known by other names reportedly began Massa in Wolo, a town near San, in 1946, after he received instructions for the movement from an otherworldly entity he encountered in a field. Before he established Massa, Démbélé apparently studied Nya.⁴⁷ Massa was spreading quickly across and beyond the broader three-corner region by the late 1940s and early 1950s, as inhabitants increasingly turned to Massa as they heard about the success it brought other people. Colin dates the movement’s arrival in the Folona to 1951, when delegations from almost every town in the area traveled to San to acquire the power object necessary to install Massa. Colin relays that by the time he traveled through the Folona at the end of 1952, buildings for Massa stood in every town except Torokoro, and he characterizes the structures as “mosque[s] in miniature” (fig. 1.5).⁴⁸ He also states that after successful harvests across the Folona in 1952, delegations from many towns in the area returned to San to thank Démbélé for the harvest and receive updated guidance.

Within the “extremely active spiritual marketplace” that Mann describes in his study of Massa, inhabitants of the three-corner region in the mid-twentieth century must have considered options available to them to help them address daily concerns and achieve long-term goals. As Colin’s reports demonstrate with respect to the area around Sikasso in the early 1950s, such a marketplace consisted of various power associations as well as other phenomena, including Massa. Decisions about which institutions to support as well as which details to relay to the visiting colonial administrator differed from place to place. Following his visit to Torokoro on November 18, 1952, Colin noted that inhabitants of the town reported that they preferred to see what results Massa brought neighboring communities before investing in a chapter.⁴⁹ Elsewhere in the Folona, people who supported Massa in their towns apparently arrived at different conclusions regarding the activities of Massa and whether they could rely solely on Massa for their well-being then or in the future.⁵⁰ Institutions flourished and declined depending on perceptions of each one’s successes and failures and the particular dynamics of local-level politics.

FIGURE 1.5. “Case du culte syncrétiste du Massa (dieu de San) à Fourou,” n.d. (Colin-Noguès 2006, 32). Photo: Renée Colin-Noguès.

Observers’ reports suggest that across the three-corner region, marketplaces of power associations and other phenomena, including movements linked or seemingly linked to Christianity and Islam, have waxed and waned for decades. I have heard about and witnessed various commitments of resources to and the growth and diminishment of individual power association chapters as well as other endeavors designed to help people realize changes in their circumstances. A few people I interviewed in western Burkina Faso and northern Côte d’Ivoire remembered Massa and talked with me about it.⁵¹ Many more people recalled another mid-twentieth-century iconoclastic movement, l’eau de Moussa or l’eau de Sinématiali, meaning “water of Moussa” or “water of Sinématiali.” They described traveling to the town of Sinématiali in northern Côte d’Ivoire in the early 1960s to drink potent water understood to secure well-being, or they talked about other people who trekked to Sinématiali to consume the water.⁵² People also said that a requirement of drinking the water included abandonment of power objects. They mentioned big piles of discarded objects as a result of l’eau de Moussa. Several people noted that the movement coincided with their own or other people’s conversions to Islam. A few of them asserted that the conversions did not necessarily translate into lasting disavowal of their own or their families’ activities with power associations.⁵³

In the early 2000s, Do Ouattara, the son of the late Touba Ouattara, a distinguished Komo leader in Sokouraba before the elder’s death in 1986, worked to establish an evangelical Christian community in the town.⁵⁴ Evangelical Christianity and power associations may seem incompatible with each other, and Do Ouattara chose to pursue the former in an attempt to distance himself from the latter.⁵⁵ Yet to some extent, his activities as leader of the evangelical Christian community based in Sokouraba resemble the work of power association leaders: Ouattara marshals tangible materials and intangible energies in order to help people overcome illness, manage relationships, and otherwise effect change in their lives. He also seeks to combat sorcery.

For some observers, Islam may also seem at odds with power associations. When Colin wrote the report he signed in early January 1953, a central concern for the colonial administrator was the relationship of Massa to Islam. Colin reflected on whether Massa was linked to Muslim practices or was at least compatible with them. He doubted a “reasoned and concerted” connection between the two but concluded that Massa had the potential to serve as an “intermediary step” between work with power objects and Islam. Colin marshaled as evidence characteristics of Massa he observed in the Folona, including the disappearof power objects as a result of the movement, Massa’s prohibition against owning pigs, its fasting requirement, its insistence on followers’ daily ab and the gatherings outside its mosque-like buildings.⁵⁶ Colin reasoned that inhabitants of the area might become disillusioned with Massa after a few bad harvests, and he wondered whether they would then return to older practices or prefer to turn to Islam.⁵⁷ Concerned about the latter possibility, the administrator recommended careful surveillance of Massa. More recently, Colin (2004, 223) has explained that the French colonial administration focused significant attention on Islam and anti-French sentiment to which Islam might have been linked.⁵⁸

Colin’s reasoning that Massa could serve as an intermediate step between power associations and Islam posits a clear distinction between the two. Yet he also recognized the possibility for their coexistence in the same document. Power associations and Islam have long intersected in the three-corner region, as people have sought a variety of means to access resources in this world and the otherworld in order to effect change in their lives.⁵⁹ Power associations, Catholicism, and other forms of Christianity have also converged and diverged. Colin (2004, 300) himself recounts that to his surprise, Bénogo Diarra—one of the complainants in the contested land dispute Colin oversaw in November 1953 and a person who reportedly identified as Muslim—agreed to swear an oath at a Nya altar in order to resolve the conflict. Documentary evidence demonstrates that people living in the three-corner region have, for decades, if not longer, negotiated disparate phenomena, including power associations and Massa, that for some observers may seem mutually exclusive but in lived practice have not been.

An historian of religion could devote an entire study to an investigation of how Islam and power associations have coincided in various places and at different moments while keeping in mind Africanist scholar Louis Brenner’s conclusion “that neither ethnically-based notions of religion nor analytical distinctions based on a dichotomy between Muslim and non-Muslim provide a useful starting point for the historical study of religion” (Brenner 2000, 163). Arabic or Arabic-inspired script and Muslim concepts inform the construction of some specialists’ assemblages.⁶⁰ Oral histories of Komo and other power associations locate the original source of power association objects in Mecca, the city in present-day Saudi Arabia revered as the birthplace of Islam’s founder Muhammad and the preeminent Muslim pilgrimage site.⁶¹ In 1924, the French administrator Monteil described Komo as a particularly honored confraternity in Ségou and relayed oral accounts that credited “El Hadj Moussa” or “Fa Koli Sissokho” with bringing the first Komo power objects to Mali from Mecca.⁶² More recently, historian David Conrad (1992, 2001) has examined the relationship between Mansa Musa’s documented fourteenth-century hajj and Fajigi’s hajj. Fajigi is said to have carried with him to the Mande region power objects that became the foundation of Komo or other power associations. Conrad proposes that Mansa Musa and Fajigi may in fact be different names for the same historical person; if so, then the power objects integral to contemporary power associations may have their material origins in fourteenth-century Mecca, the sacred center of Islam. Yet an attribution of origins may emerge after the fact, thus reflecting concerns or ambitions of people making the assertions instead of indisputable histories.⁶³

Reading twentieth-century oral histories as concrete evidence of fourteenth-century events elides six centuries of history, disregarding the fact of change and reinforcing a nineteenth-century European idea that there is no history in Africa. Islam had spread across northern Africa and south into West Africa long before the fourteenth century. Twentieth-century accounts that identify Mecca and thus the Muslim world as the origin of power objects at the core of most power associations may reflect actual origins. They may also reflect late nineteenth- and twentieth-century circumstances and sensibilities rather than fourteenth-century histories. Henry and Monteil collected the information they published at a moment when memories of the destruction wrought by late nineteenth-century Muslim expansionist campaigns or stories of them were fresh. Perhaps people who lived in and beyond the three-corner region at the time sought to identify power associations with Islam in order to lend the organizations greater credibility in the eyes of Muslim neighbors.

An impulse toward legitimizing may have persisted in present-day Mali, where scholars have continued to collect local oral references to Fajigi.⁶⁴ Power association leaders I have interviewed in Burkina Faso either never mentioned Fajigi during our conversations or told me that they did not recognize the name after I asked. According to ethnoarchaeologist Nafogo Coulibaly, whom I interviewed in Bamako in 2004, the postcolonial government in Mali aggressively promoted Islam throughout the country. He thought the government in the former Upper Volta (now Burkina Faso) may have been less focused on Islam.⁶⁵ The absence of detailed discussion of Fajigi during my interviews with power association leaders in Burkina Faso does not prove that Fajigi is not known there or is not an historical figure. However, absence of mention does raise the possibility that Malian citizens may have in certain moments felt more compelled than Burkinabe citizens to tie power associations to Mecca through stories of Fajigi.⁶⁶ More specifically, Malian citizens may have felt more inclined to perpetuate the Fajigi history in order to garner support for the organizations and the objects at their core within the Malian nation-state. They also may have referenced Mecca for other reasons. The fact that neither all Komo chapters nor all power association chapters necessarily claim an origin in Mecca further cements the impossibility of linking all power association chapters to Mecca or to Islam. As we continue to probe local-level stories and consider how they connect with other histories, we may refine our assessments and gain productive insights for recognizing the complexities of power associations and their activities rather than wedging the organizations into predetermined frameworks.

Local Chapters, Specialist Networks, and Ethics of Sorcery

Henry’s comment about elders knowing the genealogies of specific power association chapters combined with the detailed information about individual chapters Colin collected more than forty years later demonstrate the attention that inhabitants of the three-corner region have long paid to local-level histories. But when foreign observers instead focus on identifying and analyzing more general aspects of the organizations and their arts, they disregard how exactly a single chapter operates within a town, how it relates to other chapters, who constitutes its leadership, or how its leadership interacts with clients or other specialists. The framing also suggests similarities among objects, installations, architecture, and performances that three-corner region audiences likely considered as distinct. Indeed, each chapter’s existence, activities, and eventual demise depend on specific people and circumstances.

When Dahaba Ouattara and I encountered a woman and her child from Ouagadougou, Burkina Faso’s capital city, waiting to meet with Kono leader Drissa Traoré in his courtyard in Bougoula, we learned the pair had traversed nearly three hundred miles from the city to Traoré’s hometown because they sought to benefit from his expertise.⁶⁷ The woman and child could have elected to consult with any of a number of specialists between Ouagadougou and Bougoula, but they wanted Traoré’s input. We did not learn what problem prompted the woman and child to journey so far or how they decided to consult with Traoré instead of another expert. Perhaps comparable to patients in a medical doctor’s waiting room who expect other patients in the room to refrain from inquiring about their reasons for being there, clients waiting to consult with a power association leader also respect each other’s privacy. Still, some set of specific circumstances directed the patient or client to see the particular specialist. Traoré and other power association leaders earn distinction as they develop knowledge and skills that allow them to combat visible and invisible crimes as well as assist people with everyday concerns. A student who wants to perform well on an exam, a trader who wants to succeed in the market, a person who suspects a spouse of adultery, or an individual who experiences physical discomfort may consult with a power association leader for support. Clients travel from nearby towns and distant locales to meet with experts, extending the reach of a power association leader’s activities beyond his hometown.

An individual, family, or neighborhood may look to invest in a new power association chapter after experiencing a series of misfortunes.⁶⁸ Once a person or group decides to pursue a new chapter, the individual or a representative from the group as well as perhaps a few collaborators embark on a long process that people I interviewed characterized as the buying of a chapter marriage to an already-established chapter.⁶⁹ The idea that a person or group buys a chapter or marries one hints at the negotiations and exchange required to install a new chapter, as acquiring something or marrying another person in western West Africa has often entailed negotiations and exchanges with another person or group. The aspiring leader forges a relationship with the chapter from which he will buy his own or the one he will marry in order to establish his own chapter. He may also consult with other specialists to obtain additional materials, knowledge, and skills. Because the goal of a chapter is to help the leader, its members, and other people accomplish certain goals, an aspiring power association leader, like other specialists, may seek to consult with experts recognized for their particular abilities, rather than people who necessarily share the same cultural or ethnic identity. A regional lingua franca or other shared language among people who speak different first languages makes such interactions possible. Aims of power association leaders and their command of multiple languages result in a web of chapters that spans a culturally, ethnically, linguistically, and religiously diverse region.

Aspiring chapter leaders use materials, knowledge, and skills they garner from disparate sources to assemble power objects for the chapters they endeavor to establish.⁷⁰ Houses dedicated to power association chapters have appeared throughout the three-corner region for more than a century (e.g., figs. 1.6–1.7; see also fig. intro.3). According to people I met in western Burkina Faso, the overnight construction of a building marks the final step in a new chapter’s installation (see also Diamitani 1999, 114–117).⁷¹ Because the process of establishing a chapter exacts significant time, labor, and expense, some aspiring chapter leaders or groups begin the undertaking but never complete it. The late Moukanitien Traoré—who led a Kono chapter in the Burkinabe town of Kangala in the early 2000s, when I interviewed him, and who also was a maternal uncle of Dahaba Ouattara—referred to a Kono chapter in the town of Sokouraba that had not yet completed its marriage to Traoré’s Kono chapter in Kangala. Traoré asserted that the Kono chapter in Sokouraba had not fully established itself.⁷²

Power association chapters often maintain relationships with their source chapters and also forge relationships with other chapters, including ones identified with other associations. No overarching governing body appears to manage all the chapters of a single association. Rather, each power association chapter operates as a node within an intricate network of specialists and has its own history and particular relationships with other chapters. In report after report that Colin compiled, details about the leaders and origins of individual power association chapters abound. Interestingly, information Colin included in reports he completed in 1953 and 1954 differs in scope from data he recorded in similar reports compiled late in 1952, the year when he began to work in Sikasso. Colin records the types of chapters he found in each town in the earlier reports as well as the later ones. But in the earlier reports, he only occasionally specifies the name of a chapter’s leader or its connections to other chapters. Inclusion of more particulars in the later reports raises the possibility that over time, the administrator recognized the importance of connecting individual chapters to named people and other chapters.

FIGURE 1.6. “Case du Komo de Sanzana,” n.d. (Colin-Noguès 2006, 28). Photo: Renée Colin-Noguès.

FIGURE 1.7. Kono power association house, “Gomakoro,” Mali, 1985. Photo: Kate Ezra.

Information in Colin’s reports from the early 1950s also corresponds with McNaughton’s recognition that for more than a century, observers’ lists of power associations in western West Africa have varied. The consistent inconsistency in the lists points at the different constellations of power association chapters in any one place as well as the growth and decline of certain organizations. From the mid-twentieth century to the early twenty-first century, art-historical studies of power associations have tended to focus on Ciwara, Komo, Kono, and hunters’ associations, the organizations that have invested in masks with wooden headpieces and sponsored masquerade performances. Yet as we have seen, there are many power associations across the area. Nia (Nya), which Colin also refers to as Wara Mousso (meaning “female Wara”), and Wara, at times identified in Colin’s report as Waratié (meaning “male Wara”), also appear frequently in observers’ lists of power association chapters. Colin names other power objects, including Biendyougou, Dio, Dityan, Dotian, Kongozié, Manya, Namakoro, and Zantié, but the connection of the objects to larger associations is not always clear.

Aspiring chapter leaders may base the establishment of their power association chapters on knowledge that comes from an existing chapter. But specialists also seek to distinguish their knowledge and activities from the endeavors of their mentors or other collaborators. Specialists constantly search for new materials to use in their work and attempt to refine and expand their methods for managing powerful materials and energies. Their efforts reflect eagerness to sharpen their skills and to ensure that potential rivals will never know the full extent of their knowledge.

Fiery rivalry among local power association chapters defies European and North American conceptions of African villages as self-contained, harmonious entities where each association within a settlement serves a discrete function. Stories people told me about the August 6, 1986, deaths of one of Sokouraba’s most distinguished Komo leaders, Touba Ouattara, and his close collaborator, Bembèlê Traoré, who lived in the same town but in a different neighborhood, point to risks of competition within a single town. Traoré had worked closely with Ouattara, and according to Ouattara’s son, Do, the elder Ouattara shared his knowledge of Komo with Traoré. While Do Ouattara avoided talking about the deaths of his father and Traoré in great detail, he suggested that the two events were not mere coincidence.⁷³ After my interview with Do, my research associate, Dahaba Ouattara, offered an explanation. He speculated that Traoré may have been jealous of Touba Ouattara’s fame, prompting Traoré to harness his knowledge and launch an invisible attack on the elder Ouattara. According to Dahaba Ouattara, the elder Ouattara died from the attack. But before he did, he launched a retaliatory attack that resulted in Traoré’s death. The two elder specialists had long collaborated with each other, but competition between them purportedly prevailed and led to their deaths (see also Gagliardi 2014, 238–240; Gagliardi and Petridis 2021, 30).⁷⁴

The work that power association leaders and other specialists engage in cannot be separated from sorcery, or rather, the calculated manipulation of the natural environment and otherworldly energies to achieve certain ends. Kono leader Ibrahim Traoré distinguished between two types of sorcery during an interview. Using metaphor, he explained there is sorcery enacted to save people that does not necessitate the “eating” or “taking” of human flesh and sorcery executed only to destroy and “eat” people.⁷⁵ The line between benevolent sorcerers who work to save people from illness, failure, and death and malevolent sorcerers who kill and cause harm is not always clear. Traoré added that one cannot be a knowledgeable expert without inciting destruction at some point during one’s career. He offered an example: if a benevolent sorcerer saves the lives of children whom malicious sorcerers seek to capture for their own gain, the malevolent sorcerers may one day become angry at the foiled attempts to realize their goals. They may therefore seek revenge. As a result, the benevolent sorcerer who spared the children illness or even death may be forced to defend himself against the malevolent sorcerers’ attacks. His defense may precipitate the deaths of the malevolent sorcerers, thus causing harm.⁷⁶ Someone who uses his knowledge to protect innocent children from such wrongdoings—the latter actions that McNaughton refers to as “malicious sorcery” (1988, 48) and that also recall what Achille Mbembe calls “criminal sorcery” (1991, 120)—may suffer the wrath of vengeful malevolent sorcerers.⁷⁷

Power association leaders use their skills and tools to combat malevolent or criminal sorcerers. They harness their expertise in the battles they wage against other specialists and wrongdoers. But recognition of benevolent and malevolent actions is relative, at least to some degree, so the protection and success power association leaders and other specialists claim to afford their clients often imply injury elsewhere. Most combat between rivals takes place surreptitiously, even though some battles do take place in public. Recognition of attacks and counterattacks commonly emerges after the fact, when a third party attributes illness, death, or other misfortune to criminal sorcery or the work of a renowned specialist.

People familiar with the activities of specialists in and beyond the three-corner region recognize that certain calamities often constitute the only perceptible traces of otherwise invisible battles. In the early 2000s, people I met in western Burkina Faso often suspected that illness, death, or other misfortune resulted from either their adversaries’ consultations with specialists or criminal sorcery. Yet certain explanations for suffering did not necessarily preclude others. For example, internal hemorrhaging and a competitor’s intervention mutually explained the death of a young mother of small children who fell out of a tree while collecting shea nuts; similarly, HIV and local interventions to harm a man who slept with another man’s wife offered insight into the death of a diviner suspected of having an affair with a married woman.⁷⁸

As power association leader Ibrahim Traoré explained to me, actions that some people perceive as helpful appear harmful—and accordingly worthy of defense—to the people who conclude that they have suffered as a result of another person’s gain. For example, two suitors of the same woman cannot both successfully woo her and her family in order to secure a marriage. Both interested men may discreetly seek counsel or intervention from a power association leader or other specialist, although at most, only one man will be satisfied with the outcome. The man who is denied marriage to the woman might suspect the other man or his family of sorcery. The actions of a specialist may be viewed by one individual as benevolent and another as malevolent. Moreover, the side effects of such work can bring delayed injury, misfortune, or even eventual death to a person who requests a specialist’s interventions, people implicated in such activities, or any of their family members or relatives.

People in and beyond the three-corner region consequently remain alert to the possibility that power association leaders and other specialists have the capacity to engender either good fortune or damaging outcomes. Ambiguity is a characteristic of power associations that pervades their operations as well as their arts, including the objects they construct and manage. As power association leaders and other specialists work to accomplish their goals and advertise their skills to potential clients in near and far places, they create or maintain assemblages, installations, architecture, and performances that depend on certain individuals and specific contexts for meaning. In the next chapter, we will consider deliberate strategies involved in the making and maintenance of power association assemblages and installations within different domains.