Revisiting colonial legacies in knowledge production on customary authority in Central and East Africa

ABSTRACT Renewed attention on customary authority in both scholarship and development interventions renders it pertinent to revisit how contemporary engagement with this form of authority is still informed by colonial legacies. These legacies include: first, the penchant to see customary authority as solely invested in ‘chiefs’, rather than being relational and multifaceted; second, compartmentalized approaches that emphasize chiefs’ role as political authorities, while overlooking ritual, medicinal and spiritual aspects; third, misanalysing the role of female agency in the customary domain; and fourth, drawing on dichotomies that are often heavily inscribed in Western understandings, in particular, the modern versus traditional and state versus non-state divides. A growing body of work, however, has overcome these biases and developed more nuanced understandings of customary authority. Building on this work we propose to approach both the constitution of customary authority as well as knowledge production on this social institution in terms of ‘contested coproduction’. This concept helps focus on the socially constructed boundaries between different categories, and to see customary authority as a contextually shaped product of both structure and agency. It, therefore, advances the project of developing general conceptual tools that can capture the bewildering variety of expressions of customary authority while still enabling comparison.

on how customary authority is imagined, understood, crafted and drawn upon within African socio-political orders. 3 Examining the socially constructed nature of the boundaries between the modern and the traditional, and the state and non-state spheres, has importantly advanced the study of customary authority. Taking these dichotomies as a given risks introducing a priori understandings of 'the state' and 'modernity' that heavily draw on Western models. 4 Additionally, using dichotomous notions as a point of departure may obscure the complex manners in which elements defined as 'modern' or 'traditional' are mutually imbricated, overlap and co-constitute one other. 5 Many scholars now recognize that customary authority is dynamically shaped by elements of modernity and 'stateness', while it also co-constitutes 'state', politico-administrative and 'modern' authority in various ways. 6 Overcoming dichotomous conceptualizations has also moved forward the ongoing debate on the 'invention' versus the 'authenticity' of tradition, allowing these notions to be seen in relative rather than absolute terms. This debate is intrinsically connected to wider discussions about the comparative influence of colonizers versus the colonized; structure versus agency; and the nature of colonial rule, 7 in particular, the extent to which it relied on direct, violent imposition or less coercive modes of governance. 8 No clear consensus has emerged in any of these debates, in part because the historical processes under study diverge widely per context, as do the range of factors shaping them. Following scholars highlighting historical diversity, we believe that this variety of processes and outcomes makes the quest for generalization futile. Therefore, focusing on explaining differences appears more useful than searching for putative definite answers. 9 One reason for the diversity of customary authority is that it is the constantly evolving product of inherently messy processes of 'contested coproduction' with contingentand contextually differentiatedoutcomes. 10 These processes shape and are shaped by structural factors, such as political-economic relations and political cultures, but they are also the product of the agency of both Africans and non-Africans. This agency is not limited to rulers and other elites. As Spear argues 'tradition was reinterpreted, reformed and reconstructed by rulers and subjects alike'. 11 In sum, taking a cue from Feierman's conceptualization of tradition as discourse, 12 a contested coproduction approach focuses on the interplay between structure and agency. This reflects a broader tendency in the humanities and social sciences to see structure as a process. Whether labelled 'structuration' in Giddens's vocabulary, 13 or framed in terms of Bourdieu's 'habitus',14 this line of thinking emphasizes 'the customary' as a social institution that informs, and is instantiated and shaped by, everyday practices. This social institution has important discursive and imaginative dimensions. For van Dijk and van Rouveroy van Nieuwaal, customary chiefship has to be imagined in order to be enacted. 15 These imaginaries and related discourses, in turn, are intricately bound up with knowledge production. As highlighted by Vansina, tradition or custom has cognitive aspects. 16 These aspects lead to a complex interplay between tradition as a grid of intelligibility, and the social realities shaped by and perceived through this grid, allowing tradition to change while sustaining the idea of timelessness. 17 The interplay between knowledge and social practice thus stands at the heart of the contested coproduction of customary authority. 18 How knowledge production shaped the production of customary authority has been well documented for the colonial era. To organize local governance, the colonial authorities embarked upon a vast process of documenting, mapping and classifying peoples, their traditions and their rulers. This knowledge was subsequently diffused among Africans by missionaries, schoolteachers, colonial administrators and the army, being appropriated, resisted and redefined. In this manner, it came to shape the ways Africans see themselves and their (imagined) past. 19 As Ranger observes, 'the invented traditions of African societies … distorted the past but became in themselves realities through which a good deal of the colonial encounter was expressed'. 20 A growing body of literature documents similar processes of contested coproduction for the post-colonial era. This literature generally puts a greater emphasis on contestations between various African actors, 21 although parts of it focus on the ways customary authority is shaped by development and peacebuilding agencies' resources and discourses. 22 The crucial role of knowledge production in the constitution and evolution of customary authority renders it important to identify biases within the ways this knowledge is generated. We observe that in the colonial era, knowledge production on customary authority was importantly shaped by on the one hand, ideational factors, in particular, political ideas and imaginaries, and on the other hand, requirements of rule and policy implementation, whether of colonial authorities or African rulers. 23 The biases resulting from the imprint of these factors have been reproduced in later periods, not least as both ideational and political factors have continued to shape knowledge production on customary authority. These biases include the following: first, the propensity to equate 'traditional authority' with the institution of chiefs or kings; 24 second, compartmentalized understandings of customary authority that only focus on certain facets of authority (in particular political dimensions), while obliterating others (such as spiritual and ritual aspects); 25 third, the penchant to downplay or misanalyse female agency; and fourth, the above-mentioned tendency to conceptualize customary authority in dichotomous terms. 26 A growing body of literature avoids these biases and instead, approaches customary authority in a holistic, relational and emic manner, studying how it is constituted and experienced within the particular socio-political orders of which it is part. 27 This special issue builds upon this more nuanced literature. It reunites papers from different disciplinary backgrounds that focus on customary authority in both the past and the present in a variety of geographical settings. The resulting longitudinal view foregrounds the centrality of the non-political dimensions of customary authority, in particular, ritual and spiritual aspects, and the crucial role these dimensions play in the ways customary rule is constituted and legitimized. 28 Furthermore, several papers highlight the analytical inadequacies accruing from equating customary authority with the institution of 'the chief', which leads to obliterating other customary forms of leadership and authority. 29 The collection also examines the intricate manners in which 'modern' or 'state' authority, including its performative dimensions, is shaped by notions and practices of custom, even while not always explicitly labelled as such. 30 Moreover, some papers show how 'modern' authority might be read, interpreted and evaluated against the backdrop of imaginaries of traditional authority. 31 Taken together, the papers in the collection further advance the project of working towards more refined conceptualizations of customary authority. They accomplish this in part by suggesting or further developing new theoretical approaches, such as seeing the customary in terms of 'capital' 32 or performances of authority, 33 or distinguishing 'customary authority' from other 'authority that is customary'. 34 The rest of this article proceeds as follows. We first reflect on the geographical scope of the special issue, and briefly trace the history of the contested coproduction of customary authority within the area covered by the different contributions. Subsequently, we situate the special issue contributions in relation to three inter-related debates surrounding the study of customary authority in Africa, which provide an insight into the biases in knowledge production identified above. These debates relate to: first, the invention of tradition; second, the composite nature of chiefship; and third, the impact of dichotomous thinking on understandings of customary authority. We end by outlining ways forward in the study of this dynamic institution, calling for comparative, holistic and interdisciplinary approaches that take a longue durée perspective.

Diverse histories and geographies of customary authority
The case studies explored in this issue are from Eastern and Central Africa, in particular, South Sudan, the Democratic Republic of the Congo (henceforth the Congo), Tanzania and Rwanda. These countries, all bordering the eastern Congo, are deeply interconnected through social, cultural and historical networks. One example is the Cwezi-Kubandwa complex, which revolved around spiritual forces and mediums and was entwined with forms of customary authority. By the nineteenth century, this complex had spread to present-day Burundi, Rwanda, Uganda, northwest Tanzania, and the eastern Congo, an area where diverse (Bantu) languages came together. 35 In addition, the contemporary border zones of these countries can be conceptualized along the lines of Kopytoff's frontier model, constituting clusters of intertwined and multi-layered institutional networks marked by high levels of interaction, exchange and mobility. 36 One such region is the area around Lake Kivu, connecting present-day Congo and Rwanda. 37 Another is the zone between what are now South Sudan and Northeast Congo. As of 1700 AD, imbalances in organizational scale and political competition between populations in this zone bolstered a spiral of institutional innovations based on cultural borrowing and adaptation. These developments resulted in a network of different yet interwoven cultural-political complexes. However, the establishment of colonial borders subdivided these frontier regions into national spheres, which often obscured their status as a cultural crossroads. 38 While interconnected, there was much variation in the pre-colonial socio-political and cultural institutions of the different countries and sub-regions featured in this collection. A clear example is the divergence between 'forest' and 'savannah' complexes. 39 The colonial era induced further variations in these regions' political pathways. All countries discussed in the special issue were subject to British (or Anglo-Egyptian in the case of South Sudan) and Belgian colonial rule, with Rwanda and Tanzania also having been under German control before the First World War. Both the British and the Belgians implemented a system of indirect rule, where state-appointed customary chiefs became part of the administration, and functioned as intermediaries between colonizers and colonized. Yet the ways in and moments at which this system was implemented differed per region. 40 Differences in indirect rule were to a large extent the result of divergent challenges 'on the ground', relating, amongst other factors, to demography, environment, political events and economic patterns. These challenges were often resolved through trial-and-error, rather than blue-print solutions, leading to idiosyncratic outcomes in the co-optation and deployment of customary authorities. 41 Despite these differences, what colonizers had in common was that their own political imaginaries heavily shaped their approach towards indirect rule. Many colonial administrators and scholars came to take monarchies as a yardstick for African socio-political organization. This tendency was informed by their own societies' orientation towards centralized, hierarchical, territorialized forms of political organization sanctioned by tradition. 42 By comparison, other socio-political orders, in particular, those with less hierarchical authority structures and high levels of territorial fluidity, were assigned to 'lower' stages of civilization. 43 These more horizontal forms of authority had limited legibility to the colonial authorities, appearing cumbersome from the point of view of establishing effective control. 44 Portraying certain socio-political orders as less civilized or in decay, in turn, helped legitimize the colonizers' vast project of the social engineering of customary authority. This project encompassed the territorial and social regrouping of people into chiefdoms and 'tribes', the cornerstones of colonial socio-political organization. It also entailed the appointment of paramount chiefs and sub-chiefs to administer these 'tribes', even where no such rulers had existed before. This restructuring transformed many socio-political orders in profound ways. The effects were particularly dramatic for horizontally organized or segmentary societies, which often did not conceive of themselves as a single ethnic group. 45 This changed, however, through systematic exposure to the knowledgeproduced by missionaries, administrators, soldiers and scholarsthat was drawn upon to guide colonial social engineering. Being diffused through various colonial institutions, these narratives of belonging and history came to shape Africans' understandings of their origins, traditions and social identities. 46 Post-colonial knowledge production on customary authority has similarly been shaped by the complex interplay between political imaginaries and requirements of rule, of Africans and Europeans alike. In the 1950s, inspired by modernist notions of progress and development that dominated the intellectual climate internationally, European and African elites started to consider traditional authorities as 'archaic', 'feudal' and barriers to progress. Moreover, given that these authorities had played crucial roles in upholding the colonial order, they were often framed as reactionary and symbols of the ancien régime. 47 In independent Tanzania, where the Nyerere government was heavily attracted to socialist visions of modernization, these negative views led to the abolishment of customary authority. 48 Similarly, in Rwanda, the monarchy was abolished after a referendum held in 1961. In Uganda, the new constitution of 1967 suppressed the political authority of the kingdoms. 49 By contrast, in many areas of the independent Congo, despite an ill-fated attempt to suppress them in the early 1970s, customary chiefs continued to be an official part of the state administration. 50 In South Sudan too, chiefs remained a formal part of the administration, as reflected in customary courts being recognized as the lower rungs of the justice system. 51 In addition to differences in post-independence regimes' policies, numerous other factors induced further variation in customary authority's post-colonial evolution. These factors include: the evolution of central state power and centre-periphery links, regimes of land and natural resources governance, the relative salience of discourses of ethnicity and indigeneity, and the emergence of violent conflict. 52 How these elements impacted customary authority was in turn shaped by broader political-economic processes. The rollback of what was framed as the atrophied, bloated, indebted, African neo-patrimonial state in the 1980sin part the outcome of structural adjustment policieswent hand in hand with a reorientation of aid efforts towards (supposed) 'non-state', 'community' and 'local' actors. 53 Chiefs, generally perceived as representatives of 'local communities', were now expected to ensure 'local participation' in governance and improve the legitimacy and efficiency of public service provision. 54 Renewed prominence for chiefs in donor interventions chimed with post-Cold War African intellectual and political developments linked to democratization, decentralization and the intensification of identity politics under the influence of 'globalization'. 55 In some contexts, such as Uganda, this led to what has been called a 'resurgence' of traditional political structures, 56 comprising their restoration, formal recognition or expanded authority. 57 Moreover, political elites increasingly tried to tap into customary chiefs' authority to attract votes, reinforce local control, and legitimize their rule. At the same time, chiefs were 'rediscovered', by scholars and political actors alike, as embodying supposedly 'authentically African' values, in contrast to imported and imposed Western institutions. 58 These developments went hand in hand with growing entrepreneurial tendencies among chiefs, in part enabled by deregulation and other neoliberal reforms, that were closely connected to their roles in mediating access to land and other natural resources. One expression of this entrepreneurialism was chiefs' aptitude for collecting rents from extractive industries or agribusiness, 59 leading in some places to 'rentier chieftaincy' 60 or a 'traditional-industrial complex'. 61 Another development that led to changes in chiefs' post-Cold War era roles was the outbreak of violent conflict, which occurred in all of the countries discussed in this collection except for Tanzania. Generally serving as intermediaries between 'the state' and the population, chiefs now also became brokers between their communities and insurgencies. 62 In some contexts, such as northern Uganda, their role was yet further transformed by influxes of humanitarian aid and peacebuilding efforts. 63 As this overview indicates, the factors shaping manifestations of customary authority are manifold, spanning numerous societal domains, historical eras and geographical scales. Consequently, the contested coproduction of customary authority has followed distinct patterns in different contexts, with varied outcomes. 64 This variability, however, has been partly obscured by the homogenizing grids of intelligibility through which colonial observers perceived customary authority, and the subsequent imprint of colonial legacies on post-colonial knowledge production. Yet a growing number of scholars acknowledge the fundamentally diverse nature of pre-colonial, colonial, and post-colonial situations, and how they impacted customary authority. 65 In line with this body of work, we believe that the analytical task of explaining variations in customary authority can only be achieved through structured, focused comparisons of specific sub-regions. 66 This exercise falls beyond the scope of this special issue. However, bringing adjacent histories together might provide preliminary indications. In addition, this collection develops conceptual approaches that do not predetermine or essentialize the content of customary authority, but instead focus on its situated, socially constructed nature. Such approaches facilitate describing and analysing similarities and differences across regions.

Situating scholarship on customary authority: the invention of tradition debate
In the 1980s, inspired by Ranger's dictum of the 'invention of tradition', 67 it became fashionable among scholars to emphasize the colonially engineered nature of customary authority and law. At the time, this was a useful corrective to understanding 'the traditional' as rooted in a static, pre-colonial, and often idealized past. 68 In the 1990s and 2000s, however, following Ranger's misgivings about the term 'invention', 69 this focus received pushback from those pointing to the limits of colonial social engineering, emphasizing historical continuities and African agency instead. 70 Spear, one of the leading voices in this rebuttal, highlights the 'constraints imposed by tradition itself'. 71 Rather than 'inventing' rules, norms and identities, the colonial administration harnessed pre-existing, albeit ever-evolving, social institutions. Moreover, for Spear, colonial rulers relied on African agents for the appropriation of rules and norms. He concludes that 'agency must be seen as a function of discourse because people debate issues of the present in terms of ideas and beliefs drawn from the past'. 72 Yet he also recognizes the socio-economic constraints that shaped both colonizers' and Africans' agency. Indeed, a range of scholars, including Berry and Boone, have highlighted the importance of structures of wealth production and accumulation in shaping arrangements of rule, including those surrounding customary authorities and their relations to 'the state'. 73 The debate on how structure shaped agency in the colonial era is closely related to the question whether colonization induced fundamental historical discontinuities. In the revision of his 'invention of tradition' thesis in the early 1990s, Ranger admitted that his earlier insistence on invention had introduced 'an ahistorical dualism' between the pre-colonial and the colonial eras that risked overstating colonialism as a rupture. 74 Yet shortly after, Mamdani's Citizen and Subject foregrounded the transformative nature of colonialism, 75 sparking debates that continued into the 2000s. 76 While these debates were never conclusively resolved, a relatively broad consensus has emerged thatalthough there were no rigid boundaries between the two 77colonial social engineering was generally stronger in horizontally organized societies than in relatively centralized socio-political orders. 78 What has remained a bone of contention, however, is the extent to which colonial rule created ruptures through its violent character. Some authors highlighting continuities, such as Spear, describe colonial rule in rather benign terms. This is reflected, for instance, in his observation that 'colonial policy derived from ongoing negotiations and compromises with Africans and among themselves'. 79 However, studies foregrounding the violent repression of 'traditions' deemed 'subversive'such as the articles by Van Bockhaven and Eggers in this issue 80suggest that this portrayal is incomplete. In the Belgian Congo, the colonial authorities intervened in certain areas in a profound and often heavy-handed manner to shape customary authority, including by dislocating and eliminating recalcitrant authorities, appointing compliant chiefs, and reworking rank orders and rules of succession. 81 Certainly, as other colonial regimes, Belgian colonial rule worked via complex interlocking processes that included persuasion, socialization, and subject formation. 82 Yet less directly coercive techniques were enacted against the backdrop of, and partly derived their efficacy from, what Mbembe has called 'the founding violence' of colonialism. 83 The debates on continuities versus ruptures and the relative influence of agency versus structure have continued in relation to the post-colonial context, 84 including in respect of the post-Cold War 'resurgence' of customary authority. Some of those emphasizing the profound transformations of customary authority over the past two decades argue that these changes warrant the use of the term 'neotraditional' (in contrast to re-traditionalization). 85 However, this term would seem to problematically imply that previous iterations of traditional authority were not dynamic and constantly evolving.
In relation to the structure versus agency debate, discussions on post-Cold War transformations have partly focused on chiefs' growing role in development interventions. This role can be captured by Bayart's notion of 'extraversion', which emphasizes Africans' agency, even in a situation of structural macro-economic inequalities and aid dependency. 86 At the micro level, development anthropologists foreground how customary authorities act as 'development brokers' and 'gatekeepers', being intermediaries between villagers and development organizations. These studies demonstrate how rather than being 'puppets' of wealthy aid organizations, these brokers co-create development by shaping its meanings and harnessing it for their own projects. 87 As such, external aid interventions engage in the contested coproduction of customary authority. Komujuni and Büscher's contribution to this special issue shows how, in their quest for 'legitimate' and 'locally rooted' leaders, development and peacebuilding agencies in the Acholi area of Northern Uganda have concentrated their efforts on customary leaders, much like colonial governments under indirect rule. 88 Large influxes of aid and its subsequent withdrawal changed the functioning and position of Acholi chiefs, forcing them to reconstitute new bases of legitimacy.
While there is general agreement in the literature that chiefs act as development brokers, disagreements have arisen about the extent to which chiefs-cum-brokers are shaped by and can actually shape development interventions. 89 Similar to discussions focusing on the colonial era, these disagreements seem the outcome of locally variable processes with differentiated outcomes, rendering generalizations difficult.

Compartmentalized versus composite understandings of customary authority
The invention of tradition debate has been clouded by particular conceptualizations of customary authority that originate in the colonial era, but that have been reproduced, in various forms and degrees, in parts of post-colonial scholarship. These conceptualizations relate to seeing customary authority as solely invested in 'chiefs', and approaching the latter primarily as political authorities. The result of these biases has often been downplaying or misconceptualizing the role of female agency within the contested coproduction of customary authority.
Colonizers' tendency to centralize power in the person of the customary chief promoted understandings of chiefship as a singular, fixed and rigid position. 90 This perspective obliterates that history has often produced multiple, competing candidates for chiefship; that chiefs' authority tends to be coproduced by other authorities that can be considered 'customary'; and that people can move in and out of the capacity to exercize customary authority. 91 A telling example of this bias is Mamdani's characterization of the chief as a 'decentralized despot', 92 which corresponds perhaps more to the colonial idea(l) of chiefship than to the flexible and negotiable position of chiefs in practice. 93 Scholarship on customary authority in the post-colonial era has often reproduced the singular focus on the figure of 'the chief', 94 correlating with a similar emphasis in international aid interventions. 95 As shown by Komujuni and Büscher in this issue, development and peacebuilding agencies in Northern Uganda heavily concentrated their efforts on individual chiefs, seen as representatives of local communities. This focus, however, was often misplaced because chiefs' authority is, in fact, relational and negotiated. 96 Seeing chiefs as the sole embodiment of customary authority goes hand in hand with colonial conceptualizations of chiefs as having primarily political authority, anchored in patrilineal succession and inheritance of patrimony (broadly corresponding to Weber's notion of 'traditional authority'). 97 However, within African socio-political orders, chiefly authority has historically combined authority instilled in (male) hereditary power positions with big man-like authority, which is based on personal talents and the ability to inspire awe and persuade. 98 The latter dimension, which is more akin to Weber's notion of charismatic authority, 99 was closely connected to leaders' spiritual qualities or at least their potential to attract and steward ritual institutions procuring metaphysical support and public healing. 100 Colonial social engineering and related scholarship, however, did not always take spiritual dimensions and brokerage qualities into account, being focused rather one-sidedly on institutionalized political authority. 101 Feierman argues that in their relationships with chiefs and lineage heads, Europeans relied primarily on patriarchal kinship authority, having no vocabulary for 'public healing' connected to the metaphysical. 102 Moreover, they tended to frame ritual practices in terms of 'religion', therefore approaching them in line with European Enlightenment ideals regarding the separation between church and state. 103 Owing to these ideals, healers, ritual specialists or mediums were generally not acknowledged as 'customary authorities', even though they could exercize considerable influence and often co-constituted the authority of chiefs. As suggested by Eggers (this issue), these figures held 'authority that is customary', but they were not considered 'customary authorities'. 104 She demonstrates how, in spite of colonizers' limited recognition of these aspects, chiefs in the Belgian Congo were forced to deal with the intertwined spiritual and physical insecurities of their communities, which at times prompted them to join the healing movement of Kitawala.
Historically, public healing has been fundamental to many Central-African political orders. Social problems like poverty, war, and famines were perceived as an illness of society stemming from disturbed relations with the spiritual world. Consequently, healing institutionsdiversely characterized as collective therapies, cults of affliction or religious movementsplay an integral part in the region's political history. 105 In the pre-colonial era, leaders had continuously tried to control these healing phenomena by being healers themselves or by tying ritual specialists to their courts. 106 Yet such healing associations could also constitute a counterweight to established power, manifesting themselves as 'therapeutic insurgencies'. Therefore, colonial governments saw them as potentially subversive. 107 Van Bockhaven (this issue) discusses how Belgian colonial policy legally cut off customary chiefs from these potentially insurgent power bases, creating a vacuum in the domain of social healing and other spiritual and judicial practices. 108 These practices encompassed violent and coercive ones, such as leopard-men killings (anioto), aimed at making people comply, purging society from threats, and eliminating opponents. While the colonial government often tried to suppress these collective therapieswith varying levels of success 109chiefs needed them to counter insubordination among their subjects, maintain moral authority and legitimacy, and keep rivals at bay. 110 Imprinted by these colonial legacies, part of the scholarship on customary authority in the post-colonial era has similarly approached chiefs primarily as 'political authorities', focusing on their relationship to 'the state' and their role in 'public service provision' seen along the lines of Western states. It has therefore paid scant attention to their activities in relation to public healing and spiritual matters. 111 This applies particularly to the more policy-oriented literature that focuses narrowly on chiefs' roles in distinct domains of governance, such as local justice and security provision, generally overlooking how these roles are shaped by ritual and spiritual dimensions. 112 The central position of the institution of the chief and the obscuring of its medicinal, spiritual and ritual dimensions have also led to overlooking female agency. 113 During the colonial era, feminine forms of public authority were marginalized as they largely operated outside the normative 'monarchical, authoritarian and patriarchal' order of the colonial apparatus. 114 However, women had considerable agency, notably in the domain of public healing. 115 Female mediums could become very powerful, and inspire revolts and insurgencies, usually to respond to social upheaval and distress. 116 Discussions of Cwezi-Kubandwa, in which women played prominent roles, show that the tendency to uniquely conceptualize these forms of power as counter-hegemonic is erroneous. Rather, the female leaders involved had variegated, variable and often ambivalent relations to dominant authority structures. 117 This complexity continues to mark female mediumistic powers in the present. Pendle's contribution to this special issue analyses how the Nuer prophetess Nyachol in South Sudan, who mobilized large amounts of armed men in 2012 and 2013, draws upon a long-standing tradition of divination to remedy disjunctions between the secular state and the spiritual so as to achieve social healing. 118 Specifically, Nyachol uses therapeutic resources to restore 'spiritual pollution' caused by killing and warfare, which the secular state cannot address. Similar to the cases discussed by Eggers and Van Bockhaven, this study reveals public healing figures' central role in debates on legitimate leadership, foregrounding their capability to create moral communities and imagine and advance alternative political orders. As argued by Feierman, much scholarship has neglected the influence of healers and mediums, especially female ones, because of their hidden or esoteric nature. Moreover, their impact has often been considered digressive, evanescent and unstable, in contrast to (masculine) institutionalized political power positions. 119 Beyond dichotomies: studying the social construction of binaries and boundaries The compartmentalized approaches described in the previous section partly result from the continuing imprint of dichotomous thinking on scholarship on customary authority. This is another legacy of colonial knowledge production, but which has been strongly reproduced in the post-colonial era. While having a cognitive appeal as classificatory devices, dichotomous approaches have hampered the study of customary authority by causing blindness for fluidity and interstitial activities. 120 Current scholarly debates identify and try to transcend three dichotomies that have structured previous scholarship on customary authority: the external/internal distinction; the state/society dichotomy, and the traditional/modern binary.
The imprint of dichotomies on the study of customary authorities is epitomized by Mamdani's idea of the 'bifurcated state', which took inspiration from Ekeh's notion of 'the two publics' in colonial Africa. 121 For Mamdani, the colonial state was divided into two domains: the 'civic sphere', which largely excluded 'natives', and the 'customary sphere' of 'tribal subjects'. This bifurcation was produced by a series of overlapping legal, political and spatial dualisms, anchored in distinctions between on the one hand, the urban-based civic sphere of rights and legal-administrative structures and on the other hand, the rural domain of customary law and despotic chiefs. During the colonial era, these distinctions corresponded to a divide between European colonizers and Africans, hence an external/internal dichotomy. 122 Later scholarship, however, has come to reject this essentializing approach to internal/ external boundaries. Investigating the complex interplay between colonizers' and Africans' discourses, institutions and practices, it concludes that drawing such boundaries is a relatively futile analytical exercise, even when they may be highly significant from an emic point of view. 123 Based on an analysis of narratives of the first Zulu king, Hamilton argues that colonialism created a 'historically conditioned dialectic of intertextuality between "western models" of historical discourse and indigenous traditions of narrative'. 124 Therefore, it is 'impossible to draw clear distinctions between the versions of the colonized and the colonizers'. 125 Other scholars, such as Feierman, similarly emphasize that colonialism created 'unique sets of local forms which were neither African nor European, but something altogether new'. 126 This hybridization also applies to the institution of the 'customary chief'.
Similar to the external/internal divide, the state/society dichotomy has increasingly been unpacked. In the 1990s, a body of political-sociological and anthropological scholarship started to explore the socially constructed nature of state/society boundaries. This work conceptualized 'the state' as a structure that is (re)produced through conjunctions of everyday practices, performances, and discourses. 127 Hansen and Stepputat coined the term 'languages of stateness'which are not limited to discourseto describe the institutions, practices, objects and symbols that help (re)produce 'the state'. 128 This line of thinking has become increasingly influential in Africanist scholarship. 129 One example is Lund's research on 'twilight institutions', or those 'in-between' the state and non-state, public and private domains. He demonstrates how 'languages of stateness' become a resource to claim and legitimize authority for state and non-state actors alike. 130 This more nuanced thinking on state authority has led to the recognition that rather than presuming a clear-cut dichotomy between state and non-state actors, authorities, including traditional ones, can be located on a spectrum of 'stateness'. Such a flexible understanding of 'stateness' has significantly advanced the study of customary authority. For instance, it provides a lens for looking at how chiefs and other 'authorities that are customary' draw on different discursive registers of authority. 131 The state/non-state divide has often been assumed to overlap with the dichotomy between tradition and its 'evil twin ', modernity. 132 In his original formulation of the 'invention of tradition' thesis, Ranger already problematized the modern versus traditional distinction by defining invented traditions explicitly as 'agencies of modernization'. 133 He saw the traditional sphere not as separate from the modern sphere but rather as co-constituting it. However, not all scholars see tradition as something quintessentially modern. For instance, Mbembe sees merit in distinguishing between lineages of the modern and the traditional. He considers customary authority the product of several cultures, heritages and traditions of which the features have become entangled over time, to the point where something has emerged that has the look of 'custom' without being reducible to it, and partakes in modernity without being wholly included in it. 134 Instead of taking the distinction between 'modern' and 'traditional' as starting point of analysis, which requires reifying these categories, many scholars now study the socially constructed boundaries between the two instead. This research examines how these boundaries figure in emic understandings of authority, rather than as Western analytical categories. 135 Thus Kyed and Buur analyse how the cognitive grid of the modern/traditional binary is 'employed both tactically and strategically in everyday practice and at the level of public discourses, national politics, and legislation'. 136 In sum, contemporary approaches reveal that, similar to 'languages of stateness', one can identify 'languages of customariness' or discourses, symbols, modes of (bodily) action, objects and institutions that evoke and enact the notion of 'customary authority'. 137 Becker (this issue) demonstrates how these 'languages of customariness' are drawn upon to 'perform' customary authority. 138 This is not the preserve of customary authorities; public servants and development actors may also draw on discourses, symbols and regalia of customary authority. A telling example is the long-reigning autocrat Mobutu Sese Seko in the Congo. Mobutu adorned himself with attributes referring to chiefly authority, such as leopard skin, and heavily employed re-invented traditions in his retour à l'authenticité (return to authenticity) campaign. 139 Becker's contribution shows that state agents may also draw on elements of custom almost subliminally, without those elements being recognized and labelled as 'customary'. Searching for manifestations of 'the customary' in post-colonial Tanzania, she finds traces in, inter alia, the staging of authority during mikutano, or public meetings involving NGOs and officials that centre on development interventions. The choreographed, disciplined and decorous nature of these public performances, the sense of protocol they breathe, and the discursive registers used for the expression of respect and obligation: all these elements, she argues, qualify as 'customary'. Such public displays of authority are reminiscent of the ways in which precolonial big men would demonstrate power by convening large public gatherings. Hence what appears 'modern' can be deeply shaped by custom. Moreover, custom is not only situated in discourses, but also in the ways things are done. In this way, Becker further underscores the reductionism entailed by solely locating customary authority in the institution of 'the chief'.
The contribution of Watkins and Jessee (this issue) reveals that, even when not directly drawing on custom themselves, supposedly 'modern' officials can still be haunted by custom when the public evaluates their authority through the grid of 'languages of customariness'. 140 Exploring the legacies of the Rwandan umugabekazi (Queen Mother) Kanjogera (1895Kanjogera ( -1931, whose power position equalled, if not exceeded, that of the king, they demonstrate how the same gendered norms that shaped the evaluation of this queen's traditional authority continue to imprint critiques of post-colonial, post-monarchical female political elites. Imaginaries of 'the traditional' and related gendered expectations thus continue to structure evaluations of modern state authorities and political elites, long after the traditional institutions that gave rise to these imaginaries were abolished. To conclude, recent scholarship increasingly moves away from dichotomies that impede nuanced conceptualizations of customary authority, looking at their socially constructed nature instead. Indeed, studying these binaries is only useful when they correspond to lived realities, and vocabularies and ontologies rooted in the social contexts studied. But even when reflecting emic perspectives, dichotomies will always to some extent constitute abstractions propagating certain qualities and understating complexities and nuances. Moreover, for a Western readership, they may still retain essentializing connotations, prohibiting a full understanding of their emic uses. These observations raise the question to what extent the study of customary authority can benefit from conceptual languages and theoretical approaches that move away from dichotomies, even if those approaches initially served to study Western contexts. 141 For instance, similar to Hoffmann and Verweijen, one could study the customary in terms of the Foucauldian notion of governmentality and analyse the spread, forms and effects of the related technologies and techniques of power. 142 Hoffmann, Vlassenroot and Mudinga (this issue) suggest conceptualizing the customary as an additional form of Bourdieuan capital. 143 They demonstrate that even if 'customary capital' appears the emanation of 'authentic culture' and divine will, instilled in a chiefly position, diverse actors can capture and trade it for other kinds of capital. This makes customary elites attractive allies to external actors such as rebel groups and national politicians. These various examples show that adopting different conceptual approaches to power and authority than those anchored in the Weberian tradition can open up productive avenues for theorizing customary authority.

Conclusion
This article has reviewed the contested coproduction of knowledge on customary authority from the colonial past to the present, examining a range of scholarly debates and biases in scholarship, which are largely rooted in the colonial era. Exposing these biases is important in light of the renewed attention to customary authority in both scholarship and external interventions over the last two decades. As we have demonstrated, much knowledge production on customary authority has, to varying degrees, been shaped by the political imaginaries and projects of those producing the knowledge. This has often led to a heavy imprint of overgeneralizations, analytically unhelpful dichotomies, and compartmentalized approaches that foreground certain aspects of authority and agency while obliterating others. As the papers in this collection demonstrate, looking at micro-historical dynamics in conjunction with longue durée perspectives remains an adequate way to overcome these deficiencies and capture the situated, relational development of customary institutions in a holistic way. In this respect, scholarship can continue to take a cue from Vansina's macro-historical overview of large-scale institutional networks, wherein the Equatorial tradition is regenerated through innovations. 144 Similarly, it can take inspiration from Kopytoff's notion of cultural-historical continuities resulting from the dynamics of the internal African frontier. 145 In addition, we see scope for more interdisciplinary scholarship, marrying insights from anthropological and historical studies with those from political and development studies. Such scholarship is well placed to adopt structured, comparative approaches that look at a similar range of factors and processes in different contexts. Interdisciplinary and comparative research, of both synchronic and diachronic cases, can also help identify the relative prominence of the different factors that shape the contested coproduction of customary authority, thereby providing insights into the causes of the stunning variety of expression of this authority on the continent today. To achieve this, we need broad theoretical concepts such as 'contested coproduction' or 'customary capital' that can be applied across disciplines and are flexible enough to be used in comparative research, combining microand macro-levels of analysis.
A combination of longue durée, interdisciplinary, and comparative approaches will also allow for adequately analysingagainst the archival grainthose aspects of customary authority that have not always been fully or accurately studied, such as vigilantes, cults of affliction, initiation societies, and spiritually inspired insurrections. Rather than epiphenomena, these aspects are intrinsically connected to political authority and governance. For instance, control over land and people is inextricably interwoven with spiritual capacities such as (collective) healing and rain-making. While recent historical scholarship has increasingly studied these spiritual capacities, including the violence they entail, 146 only a few studies on contemporary customary authorities focus extensively on how they draw on spiritual, therapeutic and cultic aspects of violence. 147 This neglect is all the more ironic in the light of the abundant literature on the role of chiefs in peacebuilding and transitional justice. 148 At the same time, it underscores the urgent need for approaching customary authority in a holistic manner. Such an approach is not merely of theoretical interest, but is crucial for improving understandings of and interactions with the very violent orders that mark many areas of Central and Eastern Africa today.

Notes
148. See the various contributions in Huyse and Salter, Tradition-based Justice; and Zartman, Traditional Cures.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
The development of this special issue and the different contributions it comprises, including this article, were supported by the Centre for Public Authority and International Development (CPAID) at the London School of Economics and Political Science, through ESRC GCRF [grant number ES/P008038/1].