IN 1910, around four million Africans lived in cities. By 2007 there were 373 million, and demographic projections suggest that there will be around 770 million in 2030, more than the total number of city dwellers in the entire western hemisphere today.Footnote 1 Even if these estimated numbers of urban dwellers are far from accurate, they reveal the amplitude of the recent urbanization process in Africa which is said to be unique in world history and, according to the Executive Director of UN Habitat, the second biggest challenge facing Africa after HIV.Footnote 2 This recent urban trend has been analyzed in some cases as ontologically distinct from other urbanization processes in the world. Since the early 1990s, the World Bank has recognized that the urban-based economy in the South has significantly contributed to Gross Domestic Product though recent reports have described the continent's urban growth as ‘pathological’ or ‘dysfunctional’, suggesting that, unlike the rest of the world, urbanization in Africa has not been accompanied by sustained economic growth or reduced poverty.Footnote 3 Global city research theorists suggest that the world has been witnessing an epochal transformation in the spatial organization of capitalism since the 1970s which has enabled cities to regain their primacy as the geo-economic engines of the world system.Footnote 4 They also contend that there is no global city in Africa, with Johannesburg arguably the only exception. More recently, a group of Western scholars and experts have popularized the idea that Africa's cities are either dominated by a series of self-regulatory systems working outside state regulations, or are quasi-slum cities in which three-quarters of the population are slum dwellers.Footnote 5 These expert analyses, although very different in their foci, all share a ‘failed state’ developmentalist perspective, an increasingly common stance in some circles of political science that views African states and cities as ungovernable, in crisis, and incapable of implementing public policies.Footnote 6
However, a less normative vision has recently suggested that cities should not be seen only in terms of what they lack but rather on the basis of what they are and how they arrived at their contemporary configurations. Analyzing Africa's cities within the larger framework of the global South serves the dual purpose of avoiding their construction as exceptional and reinserting them in broader academic debates that are not confined to the continent's history.Footnote 7 According to this argument, it is imperative to include Africa's cities in a larger context in order to look at them comparatively and move beyond a perception of urban Africa as being essentially different from the rest of the world. In this renewed and contentious academic interest in the continent's cities, what is the place, and role, of history and historians?
The production of knowledge on Africa's urban past is far from being the monopoly of a set of scholars who would label themselves urban historians. In dealing with Africa's urban past, scholars are instead confronted by at least three distinct bodies of knowledge. First, there are an impressive number of monographs and edited collections which constitute an increasingly important field in quantitative terms but which have not yet led to the sustained development of synthetic overviews.Footnote 8 This is overwhelmingly the work of urban historians who may be distinguished from those who might be considered as only ‘passing through their territory’ (that is, the city).Footnote 9 African urban history is a specific subfield of research that first developed in the USA and Europe in the late 1960s and then in Africa from the late 1970s onwards. This history examined the ways in which Africans shaped the patterns of urbanization and how urbanization influenced African social practices. Secondly, there are scholars who do not claim to be urban historians nor make the town their territory but whose interests in related topics constitute a central, and sometimes more valuable, body of research on cities than urban history itself. Finally, there is the production of other social scientists that includes an historical perspective within the framework of a more general interrogation of cities in Africa and the world. Considered in this broader perspective, the historiography of cities in Africa is considerably richer than the limited subfield of urban history.
For understandable reasons, attempts to merge the body of knowledge produced on Africa's urban past with the more general trend of urban studies have been very few but have nevertheless opened up cross-fertilizing fields between history and the social sciences.Footnote 10 Understanding Africa's urban past and presenting this in a comprehensive analytical manner is actually faced with two major – and probably not so new – challenges. The first is to critically analyze the categories which are commonly used by scholars and policymakers and which are supposedly characteristic of the cities of the continent. In the past, cities were seen as Islamic or colonial; today they may be analyzed as postcolonial, cosmopolitan, informal, or yet to be globalized. A second challenge is to move beyond one of the major biases of urban history and urban studies that often remains trapped in a localism that considers local strategies, regulations, and actors in isolation.Footnote 11 This article seeks to avoid the traps of localism and the temptation of categorization in an attempt to understand Africa's historical urban dynamics. I suggest that research on Africa's urban past may benefit significantly from a more thorough engagement with African history, urban studies, and social sciences in general. Historians may also contribute to this debate in ways that demonstrate that history matters but not in a teleological and linear way.
Analyses of urban Africa in relation to world history and the history of the state are suggested as two analytically productive ways for exploring the dynamics of Africa's urban past while contributing to ongoing debates in urban theory, state-formation analysis, and world history. Unlike current studies of globalization, which do not tell us much about the historical depth of interconnections between Africa and the rest of the world,Footnote 12 world history tries to understand the processes involved in the exchange of commodities and social and cultural practices between continents. Such an approach does not need to include the West nor, therefore, to involve a study of Western capitalism in order to pay increasing attention to the social, cultural, and environmental effects of the globalization process.Footnote 13 To insert the continent's cities in world history provides a method of analysis that does not depend upon the lens of European history nor does it oblige consideration of Africa's cities as necessarily African cities. The major metropolises of the continent are actually not only African cities. Cairo belongs to the Middle East, the Mediterranean Sea, and the African continent; the port cities of the Indian Ocean may be referred to as ‘Afrasiatic’ from the early modern period; while Johannesburg can be said to be an elusive metropolis because of the multiplicity of registers in which it is either African, European, or even American.Footnote 14 It is suggested here that the notion of African cities is inappropriate – albeit largely used in the academic literature – as it implies a common history among cities of the continent while underestimating the heterogeneous global influences that have shaped them. Privileging a broader reading of the continent's cities is also a necessary academic challenge in a context in which Africa has been excluded from research on and theorizations of the global city.
However, this world perspective is not to be looked at in isolation from other scales of analysis. Charles Tilly once admonished urban historians to stop oscillating between ‘the time-space particularism of local history and grand timeless, spaceless processes, causes and effects’. Tilly's cure for this disease was to exhort urban historians to admit that their turf was quintessential social history, and to turn back to interpreting ‘the ways that global social process articulate with small–scale social life’.Footnote 15 To follow up Tilly's suggestion is one possible way of inserting Africa's cities into world history and taking social history into consideration which, since the 1980s, has been one of the most innovative fields in the historiography of cities in Africa.Footnote 16
It is also important to consider the role towns and cities have played in the history of state-formation and explore the effects and limits of state action on the making and shaping of the continent's cities. The point is not to return to a nationalist reading of urbanization process or to write the history of Africa's cities caught within the teleological framework of the rise of the nation-state.Footnote 17 It is instead to participate in an important current debate on state formation in Africa and to understand how local, national, and transnational actors have forged and remade the state, an interpretation that challenges ‘failed state’ analyses which ‘tend to reify African states as ahistorical “things”’.Footnote 18 Analyzing the multiple, ambivalent, and non-linear city/state relationships on the continent is a way of moving beyond this normative vision of the state while facilitating the interrogation of this relatively unexplored issue in the analysis of state formation in Africa. Such a perspective also opens up novel analytical approaches for understanding the urbanization of the continent. African states and cities have a long history but this current analysis limits its scope by drawing upon examples predominantly from the last two centuries.Footnote 19
AFRICA'S CITIES IN WORLD HISTORY
Various labels have marked the last thirty years of urban studies and the literature on urban history in Africa. From the criticized Islamic and colonial city paradigms to more recent global city theory, Africa is alternatively perceived at the core or at the periphery of a process of academic categorization that concerns most of the cities of the south. In a way, these approaches have been developed to either include the continent's cities in, or exclude them from, a world history perspective, though neither has been entirely able to escape an oversimplifying vision of Africa's urban change. Concentration on large-scale, anonymous structures and processes has neglected the life experiences of ordinary residents that are at the core of social history. After reviewing the different forms in which Africa's cities have been categorized, I will look at ways of thinking about them both locally and globally, and will use the history of urban labour and the process of historicizing world imaginaries in Africa as two examples that illustrate the potential articulation of world history and social history.
FROM ISLAMIC CITY PARADIGM TO THE GLOBAL CITY THEORY
From the 1930s to the 1990s, the Islamic city paradigm enjoyed considerable popularity especially in North Africa. The Islamic city (that is, a set of political, economic, social, and cultural characteristics supposedly shared by towns and cities of the successive Arab and Ottoman Empires between the seventh and nineteenth centuries), was mainly defined negatively on the basis of a number of elements it lacked according to Max Weber's prototypical city: the regularity and institutions of the Classical city; the political autonomy of the medieval town; urban planning; links with the countryside. The Weberian idea of the medieval town as a parasitic body, a foreign entity encrusted on the land, a ‘consumer city’ rather than a producing city, was refurbished by Hellenist Moses Finley and successive Orientalist scholars.Footnote 20 In the last thirty years, several researchers have denounced the absurdity of using Islam as a conceptual framework to account for the urban phenomena of countries with such varied historical traditions.Footnote 21 In sub-Saharan Africa, indigenous forms of urbanism predated Islamic expansion and had no links with Arab trade networks (Aksum, Zimbabwe, Benin City, Old Oyo, Ile Ife, Jenne Jeno).Footnote 22 On the East African coast, Swahili towns were polyglot, multiethnic frontiers, and composed of various population groups (Arabic, Indian, African) while (Sahelian) cities ruled by Islamic elites (Timbuktu, Djenné, Gao, Katsina, Kano) were equally places of mutual influence between Northern and Western African societies.Footnote 23 Despite these factors, the Islamic city paradigm only underwent a gradual demise and is still sometimes considered central in the historiography of cities of the Arab and African world.Footnote 24 Resorting to the consumer and the parasitic city models to explain ancient towns in sub-Saharan Africa similarly remains in common use.Footnote 25
The ‘colonial city’ is another widespread label used both within and beyond the African continent to refer to the city under colonial – and by extension – apartheid rule. It may refer to a particular moment during which colonialism was portrayed as a power demarcating, racialising, and ordering urban space.Footnote 26 The overemphasis on colonial control, segregation schemes, and the ‘sanitation syndrome’ has indisputably played a role in the development of the colonial city paradigm. A large body of literature has helped to shape the understanding that all European powers used (hygiene/sanitation) policies to enforce a clearer division between Europeans and Africans. However, this was also a late nineteenth century world phenomenon related to heightened European racial consciousness and therefore more than just a strictly colonial practice.Footnote 27 The debate on segregation and the control of space has been most vivid in South Africa where cities have often been seen by historians as the origin and centrepiece of apartheid legislation as well as a privileged site of anti-apartheid activism.Footnote 28 This overemphasis on the perspective of the colonizers rather than the colonized has led to inconclusive typologies common to many syntheses of Africa's urban past which ‘tend to divide between the essentially African and the essentially colonial city’.Footnote 29 In some cases, analyses reduce the African practices of the city to a colonial perception of disorder while the reader is left to infer what Africans thought about these developments and the extent to which the laws were obeyed.Footnote 30 To view colonial cities as ‘dual cities’ is misleading since colonial societies cannot be understood merely in terms of a ‘European versus Indigenous’ dichotomy.Footnote 31 It omits the agency of African societies, their capacity to overcome such divisions, to ignore them or even to imagine them differently. The colonial ordering of urban space was incomplete as the colonial powers lacked the resources to enforce segregation ordinances while many Africans circumvented colonial regulations even in the stricter eastern and southern African colonial regimes.Footnote 32 In South Africa, this academic – and understandably political – obsession with the segregated nature of the city has, more than in any other part of the continent, been radically criticized since the end of apartheid for viewing state and urban control in teleological, monolithic, functionalist terms. Furthermore, this perspective consistently underestimates the world influence of modern architectural and planning movements in the country while forgetting to trace the relationships between the local and the national state, and does not pay enough attention to township dwellers' practices and imaginations of cityness.Footnote 33 This latter criticism highlights the necessity of giving up a dual city analysis though it arguably remains necessary for the continued exploration of the extent and limits of discrimination policies imposed upon African urban dwellers during colonial and apartheid times.Footnote 34
A similar ambiguity revolves around the postcolonial city label, commonly used to qualify cities which have been developing since the end of the colonial period. This approach implies the comprehension of an enduring common colonial legacy that unites Africa's cities and other cities of the South in a postcolonial framework.Footnote 35 In certain instances, this notion, still poorly discussed in academic arenas, nevertheless articulates something more essential than a particular historical moment, a condition either radically different from the colonial city or, conversely, one which cannot rid itself of this legacy – both readings that impoverish our understanding of the colonial and postcolonial urban past. The colonial situation, instead of being analysed as a total social project, as suggested by Georges Balandier,Footnote 36 is limited to urban planning, technologies of control, and the civilizing mission. Sources of conflicts are said to emanate mainly from the spatial separation between colonizers and the colonized, while the contemporary city is perceived as a fluid and more variegated space shaped by conflicts of a more complicated nature (rich against poor, long-standing residents versus newcomers, youth against elders, citizens versus immigrants …).Footnote 37 This latter approach does not always avoid the trap of introducing forms of binary discourse that surprisingly tend to classify colonial and postcolonial cities into ontologically distinct categories. In this process the concatenation of historical periods and the various influences that have shaped African city life are forgotten. Moreover, seeing Africa's cities only in terms of their colonial and postcolonial relationships may preclude a fuller understanding of the multifaceted ways in which they have engaged with the larger world.Footnote 38
While the various attempts to integrate Africa in large historical frames have sometimes led to an essentialist vision of its cities, one of the more recent and innovative urban theories has left the African continent aside. Global city research looks at the changing forms of capitalism and its effects on urbanization at the world level including the concentration of capital, transnational corporations, and financial services industry in some specific metropolises. On the one hand, it entails an explicit critique of mainstream conceptions of globalization which presume that territoriality, borders, and places are becoming irrelevant.Footnote 39 On the other hand, it challenges state-centric approaches in the Western world by stressing the role major cities have played in reshaping the geography of capitalism and the rescaling of statehood.Footnote 40 Inevitably perhaps, critics of global city research have emerged especially from within Africa as it presents Western cities as the paradigmatic ‘model’ in terms of which all other cities are to be interpreted, regardless of their particular locations or histories.Footnote 41 Within such a framework, cities located ‘off the map’, in the South and especially in Africa, are almost invariably said to be lacking the characteristics that would qualify them as genuinely ‘global’ cities.
Given the shortcomings of the approach mentioned above, alternative notions of the global city have thus been developing more recently. However, using the notion of world city instead of global city to qualify metropolises outside the core of the capitalist system is probably of little help as this, again, has led to some doubtful classifications.Footnote 42 The globalizing city notion might be more useful as it construes globalization as a process, not a state that reifies and classifies cities of the world.Footnote 43 Looking from within the globalizing city of Accra, Richard Grant considered that there is a widening, deepening intensification and growing impact of global connections on the local economy and on local engagements with the world beyond.Footnote 44 The adoption of globalizing strategies is not exclusive to socially mobile transnationals in building gated communities or new foreign companies investing significantly in Accra, but also concerns a process from below including returned migrants, evangelical movements, international NGOs, and slum dwellers looking for transnational alliances. The fact that Accra is more connected today to the world economy, to transnational NGOs and migrant networks is an important point that invalidates part of global city research which, de facto, excludes the continent from the global city map.
However, many scholars agree that Africa constitutes a site of extremely uneven globalizing processes. Such unevenness stems from the movement of capital across national borders, in the process linking particular and dispersed sites of global relevance thereby leaving huge regions simply bypassed.Footnote 45 Furthermore, post-1970s contemporary urban dynamics need to be read simultaneously as disconnected from – or poorly connected to – the world economy. James Ferguson has dismantled the idea of any linear trend towards more permanent urbanization and greater connection of a city's residents to the world, thereby questioning the meta-narrative of modernization that has frequently been associated with twentieth century urbanization in Africa.Footnote 46 Such trajectories of disconnection may be more common than we think as processes of counter-urbanization have developed in several regions of the continent.Footnote 47 There are also many small and middle-sized towns (with less than 500,000 inhabitants) where future urban growth will be (proportionately) larger than in the metropolises:Footnote 48 the marginal nature of international flows of people and goods in such locations suggests that those flows will not provide an effective measure for analyzing the particular social, political, and economic dynamics characteristic of these burgeoning towns.Footnote 49 The notion of a globalizing city might be useful to define post-1980s Accra, but the self-prophetic dimension of this construct is of little help in understanding the social and economic heterogeneity of the continent. In all likelihood, all these labels will be unable to fully reintegrate Africa's cities into world trends without oversimplifying and homogenizing their multiple histories. One single notion does not appear to be sufficient for grasping the uneven integration of Africa's cities into world history. In the next section, the urban labour market in Africa is used as an illustration for understanding the relations between this heterogeneity and larger economic trends.
WORLD HISTORY AND THE MAKING OF AN URBAN LABOUR MARKET
From the gradual abolition of the transatlantic slave trade in the 19th century to late colonial stabilization policies and post-1970s structural adjustment policies (SAPs), the urban labour market in Africa seems to have been shaped by world economic trends and by decisions taken in international arenas in which African actors have apparently minimal bargaining power. A focus on historical changes does not, however, illustrate the resilience of African societies to these policies and the manner in which these policies were reshaped in tandem with both national and local circumstances. Recent research in social history and anthropology facilitates a reconsideration of the extremely uneven consequences of world economic history in Africa.
The link between the decline of the international slave trade and the subsequent increase in domestic slavery in the nineteenth century urban economy remained obscure twenty years ago.Footnote 50 This is probably due to the fact that for a long time ‘many studies dealt with the general history of the states or communities in which ports were situated, rather than with the specific history of port towns themselves’.Footnote 51 We now know from several monographs that, locally-speaking, the transition from exporting slaves to ‘legitimate commerce’ created an acute demand for new labour, increased domestic slavery in and around coastal towns, while the population of slave ports (for example, Accra, Ouidah, Lagos, Luanda, Cape Town, Zanzibar) increased dramatically in a few decades, and sometimes even within a few years.Footnote 52 In the history of the continent this change should not be seen as exceptional, as the integration of slaves into major towns and their surrounding hinterlands was central in the making of major hinterland cities especially in West and Central Africa.Footnote 53 It remains to be explored how the abolition of slavery in coastal societies transformed emerging urban labour relationships and how the repercussions of widespread slavery affected social relations in various regions of the continent throughout the colonial period and beyond.Footnote 54 In nineteenth century Lagos, wage labour held little appeal as local slave owners struggled to slow the decline of slavery and hold on to their bondswomen and men by using marriage, overlordship, patronage, and polygyny, and sought dependents not only for their work but also for the political support and social prestige they embodied.Footnote 55 Former slaves in Zanzibar adopted new cultural urban forms and leisure (for example, dialogic poetry exchange, dance performance, competitive songs and football) to escape their servile conditions.Footnote 56 In French West Africa, many veterans, most of whom were slaves or ex-slaves, had no desire to return to villages where they would be subordinate to elders or former masters.Footnote 57 Their integration into different urban environments remains to be understood but reveals in some cases an early process of individualization and social distinction (for instance, through the acquisition of land, bars, and new forms of attire).Footnote 58
While the transformation of slavery into new labour relationships remains to be fully explored at the town level, we know that late colonial stabilization policies were obsessed with urban workers. The main aims of these policies were to cut off the solidarities that had developed in the urban riots of the 1930s and 1940s between the ‘floating population’ and the more permanent urban workers, and to transform a potentially ‘anarchic’ floating population into a veritable working class which, it was hoped, would become increasingly differentiated in terms of salaries, promotions, and qualifications.Footnote 59 Practically, however, the policies had differentiated and unwanted localized effects. In Tanganyika, maintaining the political legitimacy of chiefs and headmen proved more important to local administrators than did the implementation of a stabilization policy and raising the social welfare of working class urban Africans to international standards.Footnote 60 In Nigeria, railway workers were able to use the perks of their jobs – salaries, housing, medical and other benefits – to support large households and become important patrons, rather than form the ideal small, nuclear families advocated by official policy.Footnote 61 In several French and British colonies, instead of disciplining urban workers, stabilization policies generally led this section of the population to actively claim more social and political rights for their particular groups or categories.Footnote 62
To a certain extent, the post-1970s SAP sought to reverse the effects of these late colonial stabilization policies. It had the effect of shrinking the number of public servants and formal employment in many cities of the continent and reducing what International Monetary Fund (IMF) economists denounced as the urban bias of African governments (taxing heavily rural products and giving high wages to urban dwellers who were also the first to benefit from public investments in education, health, and infrastructure). The popularization of a new notion of the informal sector by the International Labour Organization and several generations of academics, policymakers, and bankers to describe all activities that escape taxation or state registration suggests that the cities of Africa and the global South have entered a ‘new era’ marked by a general ‘informalization’ of the urban economy.Footnote 63
This change was not as dramatic in all African countries and the division between the pre-1970s and post-1970s periods sometimes seems too clear-cut while case studies show more complicated situations as well as unexpected elements of continuity and change. Firstly, if we define informal activities as activities not registered, taxed, or regulated by the state, then they must be considered as a continuation of colonial history rather than a novel feature of the post-1970s era. There may also be a historical and gender bias in the informal sector scholarship. The idea that colonial cities were exclusively male preserves while women stayed in the country where they reproduced peasant livelihoods has been challenged successfully by recent research on women's urban experiences that has helped to remap our understanding of the urban colonial landscape.Footnote 64 Women involved in trading, dyeing, brewing, or prostitution were part of colonial and postcolonial informal urban economies and were probably more prone to join the non-regulated economy because of their loose connection to administration, their will to escape taxation, and fundamental commitment to following their own agendas.Footnote 65 Moreover, once in town, their strategies differed widely from one place to another according to different possibilities of access to land, housing, and work, and according to local and national regulations that were, in many cases, more determinative than the 1980s SAP rupture.Footnote 66
Secondly, instead of viewing the implementation of SAPs as responsible for a general informalization of the urban economy, several studies have indicated that large metropolises such as Nairobi, Johannesburg, Cape Town, Abidjan, Lagos, Maputo, and Dar es Salaam have started, in the last decade and sometimes earlier, to regulate their trade, reconquer their centres, and reassess their prerogatives in terms of spatial planning and taxes.Footnote 67 These regeneration projects are also aimed at implementing a World Class City agenda, representative of an increasing concern among city managers and politicians to attract investments, raise land values, and clean up the streets from traders. These policies indicate the difficulty of dissociating informal from formal forms of labour in Africa's cities as well as in the more general economy of the continent.Footnote 68 They also invalidate the idea, suggested by Rem Koolhaas and his colleagues, that regulated colonial cities have gradually transformed into paradigmatic informal African equivalents.Footnote 69 The exploration of social history and world economic trends remain imperative to avoid homogenizing African urban worlds.
HISTORICIZING WORLD IMAGINARIES
To move beyond the excessively narrow economic approach of globalization, more innovative research has looked at the everyday social practices of inhabitants shaped by ‘globalised imaginaries’ to highlight the worldliness of contemporary African life forms.Footnote 70 As mentioned by AbdouMaliq Simone, African urban residents have developed a ‘worlding’ from below as they are able to operate at larger scales in a broader world through the mobilization of religious practices, modes of dress, food, and musical taste.Footnote 71 These perspectives share a vision of Africa's cities in which world cultural repertoires are both locally produced and imagined, and propose a comprehensive reading of the city which avoids both localism and the global city paradigm. The risk, however, inherent in globalization studies is to omit mention of the historical dimension of these processes as well as the limits of world connections.Footnote 72 An additional risk is to read the city as a unified actor – considering a metropolis as cosmopolitan, for instance – assuming that its residents univocally share a (singular) world repertoire.
Historicizing various competing world imaginaries among city dwellers elucidates the messy process by which people get connected to worldwide influences and helps to identify groups (who might be defined along social, gendered, or generational lines) who claim to share a form of cosmopolitan life. More importantly for our purposes, it also sheds light on the very processes by which becoming cosmopolitan was equivalent with becoming urban: both marked a distinction from the countryside. Heterogeneity and historical contingency shaped how each group or individual, in each town or city, came to imagine themselves as belonging to several worlds at the same time. For instance, at the core of the social history of Bulawayo (Zimbabwe) in the 1930s and early 1940s, stands a figure named Sipambaniso. He was alternately a police officer, welfare officer, railway adviser, union leader, executive of the Bantu National Congress, and was involved in modern music, dance, football, and boxing. He was considered to perfectly represent the interests of the long-term residents of the Bulawayo at the time.Footnote 73 His capacity to embody these diverse cultural and political repertoires aptly captures what it meant to be urban in mid-20th century Rhodesia. In other urban locales, other world repertoires were appropriated and merged with local idioms. In Kinshasa and Brazzaville, appropriation entailed the swift adoption by young Congolese migrants of sports, movies, and new musical genres at the crossroads of American, African, and European influences.Footnote 74 Gangsters known as tsotsis in Johannesburg participated in the emergence of a ‘youth subculture’ made up of a mix of American and local influences dominated by specific language, dress code, violent practices, and exacerbated masculinities.Footnote 75 They claimed a strictly urban origin, and looked down upon and confronted other groups considered as backward such as the Marashea gangs whose activities were organized between Sotho rural settlements, gold mining areas, and urban enclaves.Footnote 76
Becoming cosmopolitan and becoming urban were thus contested historical processes everywhere, though especially in regions with longer established urban traditions (for example, the Swahili coast, hinterland West Africa, North Africa). In these places, Islam played a central role in the process of dissociating oneself from a rural, ‘backward’ environment and being associated with real or imaginary prestigious external links. It is significant that Swahili and Hausa city-state elites and those in northern Sudanese cities, especially after the fifteenth century, claimed Arab or Persian origins.Footnote 77 Timbuktu and Walata between the thirteenth and fifteenth centuries were multiethnic centres of commerce but as time passed, fewer and fewer scholarly families claimed a West African ethnic identity and ‘by 1800 there were almost no scholarly families left in Walata and Timbuktu that claimed anything other than a Arab identity’.Footnote 78 Similarly on the Swahili coast from the sixteenth century onwards, to be civilized, one had to be a coastal townsperson: the term mwungwana signalled urban culture and the elite's sense of belonging was demonstrated by social markers such as stone houses, prestigious external origins, Islamic practices, and specific modes of public behaviour.Footnote 79 This status of urban belonging became increasingly attractive in the late nineteenth century. Participating in cosmopolitan town life, public rituals, and dance societies was a source of prestige and became central to being recognized as a member of the urban coastal society. Nevertheless, demands for inclusion by people from the interior were highly contested by urban patricians who sought to heighten their exclusiveness by identifying themselves with overseas regions associated with Middle-Eastern ancestry, Islamic authority, and international commerce.Footnote 80 The word wanyamwezi was used by urban coastal people as a derogatory term connoting ‘uncivilized’ or ‘rural’, though the boundaries separating Swahili townspeople from the supposed barbarians of the hinterland were shifting, permeable, and extremely ambiguous.Footnote 81 These examples illustrate what it has meant to be urban and to be cosmopolitan in Africa at different moments in history. They also suggest how these strategies, when pursued by elites, could be intimately tied to processes of state building.
RECASTING CITY/STATE RELATIONSHIPS
Exploring the co-constitution of cities and states in the continent's history is a less than straightforward exercise. The long tradition of research in history, geography, and historical sociology has focused on the central albeit ambivalent role played by towns and cities in building and rescaling the state in Western Europe.Footnote 82 This is not the case in Africa where the history of city/state relationships in a longue durée analysis is often either ignored or more recently dealt with through a normative framework of analysis. Thirty years ago, John Peel recommended that ‘a satisfactory African urban history must be about politics … which directs attention to the larger unit of the state in which West African urban analysis must be set’.Footnote 83 These recommendations have not always been followed by historians and political scientists. From the perspective of economic history, trans-Saharan and transatlantic trade and Western capitalism are effectively seen as the determining forces of the urbanization process. Such a perspective tends to downplay the role of political leaders at the local, national, or regional levels in shaping multiple urban forms on the continent.Footnote 84 From the perspective of political science, the town-state nexus from the precolonial period to the present sometimes appears but has yet to be fully explored. In general, towns and cities have not been highlighted as significant factors in the history of the state in Africa. They are at best briefly mentioned in, and sometimes totally excluded from, the most influential books on the historical formation of states in sub-Saharan Africa.Footnote 85 The central argument of Jeffrey Herbst's analysis appears to be an exception but does not totally escape a normative approach to the state in Africa. He suggests that the state in Africa does not extend beyond a core area, defined as either the capital or the critical urban areas and those rural areas with important economic assets.Footnote 86 The geographically limited authority of such states reveals their fundamental and historical weaknesses. Herbst uses the same ideal type definition of the state (a monopoly on legitimate violence, an autonomous bureaucratic apparatus, the embodiment of popular sovereignty, a spatially and territoriality coherent entity) that scholars have used to interpret state politics around the world.Footnote 87 Measured by this ideal type definition, ‘African states are often identified as failed not by what they are, but by what they are not, namely, successful in comparison to Western states’.Footnote 88 This analytical framework is thus at the core of the ‘failed state’ perspective frequently criticized by anthropologists and political scientists but which is nevertheless influential in the ways cities in Africa are often understood.Footnote 89 In other words, a long historical urban past in Africa might be seen as either marginal to state-building processes or as the manifestation of the inherent weaknesses of the African state in general.
Some authors have suggested less essentialist ways of exploring state-building in Africa. The historicity of the state, the embeddedness of its bureaucratic organizations in society, the interactions between state officials and non state actors, the material dimension of statehood, and the importance of accumulating basic legitimacies are all key dimensions that apply to African states as well as to many states in the world.Footnote 90 Scaling down the focus of analysis from the level of the territorial state to local arenas or to socioeconomic areas (such as mining or border areas) has been suggested as another possible way of evaluating the manner in which the state is built from the margins or from local situations.Footnote 91 Two other analytical possibilities are advanced here. First, understandings of the ‘state’ and the ‘town’ have, for too long, engaged with these as separate political entities even though several historical studies have recently emphasized the necessity of analyzing them within a common analytical framework. Second, cities have been privileged places in which multiple authorities compete over state functions and, as such, should contribute more to our understanding of the history of the state in Africa than is the case at present.
MAKING THE TOWN, MAKING THE STATE
There is a clear correlation between state formation and processes of urbanization in various regions of the world and Africa is no exception. This is in spite of the examples of stateless though urbanized societies, for instance Jenne-Jeno, and states without towns but with mobile court capitals, such as Burundi, Rwanda, and Buganda until the nineteenth century.Footnote 92 As in Europe, most African states needed the revenues generated by trade activities concentrated in towns and ports in order to strengthen their power over a territory and to finance war. This was the case for the long, urbanized past of both North Africa and areas of sub-Saharan Africa, where governments were even more dependent on taxing trade given the low population densities dispersed across vast hinterlands. In several parts of the continent, town and state formation were simultaneous and cumulative processes. Control or protection of long distance trade, warfare, and a slave-based economy were central elements for town prosperity that, in turn, reinforced state power. The nineteenth century Sokoto Caliphate provides a significant example. The long history of Hausa city-states, the capture of slaves by the cavalry destined for work in the textile industry or for settlement in colonizing villages around various towns, the expansion of Hausa merchants throughout West Africa and their connection to regional and international trade networks, all contributed to the economic, demographic, and urban expansion of the Caliphate.Footnote 93 The most powerful and populous state in sub-Saharan Africa at the time was also the most urbanized.
Warfare was an integral part of state formation and was a crucial stimulus not just for social and political change but also town growth. There are many examples of towns and states formed by displaced groups or developed by the arrival of ‘refugees’. For instance, the collapse of the Songhay ‘Empire’ in the late sixteenth century led to the re-emergence of Walata (in the western Sahara) as a notable centre of Islamic literary production thanks to the migration of cleric lineages from Timbuktu.Footnote 94 The arrival of warriors and merchants from the inland Niger Valley led to the creation of the town and state of Kong (in today's Ivory Coast). The migration of merchants and religious lineages from Gao and Timbuktu to Katsina and Kano contributed much to the rising importance of Hausa city-states after the seventeenth century.Footnote 95 In many parts of the continent, warfare led to the multiplication of city walls which, in turn, led to a greater concentration of people seeking protection. This emergent settlement pattern allied to the practice of warfare also facilitated taxation and the control of trade while at the same time imparting prestige to rulers. Significantly, in the 19th century, the two most urbanized regions in Africa south of the Sahara, Hausaland and Yorubaland, were also the densest groupings of African walled cities.Footnote 96 Despite the evident significance of the role played by warfare in the development of urban centres, a systematic, continent-wide analysis of this dynamic has not been undertaken.
Nineteenth century warfare was also decisive in shaping town institutions, integrating ‘strangers’ into local societies and making the state. In Accra, military organization and warfare determined the form of Ga institutions and the dynamics of political competition.Footnote 97 In nineteenth century Kumase, warfare and military organizations such as asafo became a force to be reckoned with and contributed much to emerging distinctions between the ‘urban’ and the ‘rural’.Footnote 98 In the town of Ibadan, founded in the late 1820s, chieftaincy titles were allocated according to practical criteria of military merit and maintenance of a following by warrior-founders of compounds (ile).Footnote 99 The city did not emerge out of a coordinated project managed by an overarching authority as had been portrayed in earlier academic research. Instead, the conflicting oral histories collected by Ruth Watson clearly demonstrate that the ‘residential group was constantly changing as new refugees arrived, built up their military retinues, and then departed to set up their own ile, taking their own group of followers and their families with them’.Footnote 100 Ibadan is symptomatic of both the affirmation of the nineteenth century military state and indigenous forms of African urbanism: in less than half a century, it became simultaneously the most powerful state in the region and one of the largest cities on the continent.
As state formation and town growth are aspects of the same historical process, it is near impossible to disentangle the terms for ‘state’ and ‘town’ in several African languages. The foundational concept of Yoruba political sociology is the term ilu, commonly translated as ‘town’ or community.Footnote 101 An ilu is both a town and a polity and Yoruba people do not make a conceptual distinction between the two.Footnote 102 In Dahomey, urbanity was not only defined by size and the concentration of people, but by political autonomy and the role of the town as an administrative seat: to referred to a town or settlement of any size while togan (the chief of the to) was a provincial governor within the kingdom.Footnote 103 In the Ga language (southern Ghana), the term man most commonly denotes ‘town’ and has wider social and political connotations such as people, nation, or state. There are also a large number of names that distinguish townspeople enjoying full civic rights (mambii) from people of the bush.Footnote 104 ‘The very things that made Accra a state – organized political, military, legal and religious institutions – are also the things that made Accra distinctively “urban”.’Footnote 105 These examples suggest that networks of towns directly shaped precolonial states and militate against the idea that the latter developed independent of the former.Footnote 106
In several parts of Africa, a wide range of negotiations, compromises, and conflicts came to define complex relationships between European officials and local elites. Precolonial urban-based institutions were often important in making the early colonial state as many colonial territories came into existence through them. The old and new networks of towns became the architecture of emerging bureaucratic colonial states (with expanded though limited administration and public services). As in rural areas, colonial administration was unable to afford the cost of services for the urban population and was forced to tolerate institutions that were beyond its control and could be used to subvert or evade its dominance. This was the case in cities created by the colonial powers – for instance, the chefs de quartier in FWA or autorités coutumières in Congo, as well as old settlements like the Kasbah of Algiers, the walled city of Tripoli, the Senegalese Four Communes, the townships of Accra, Ibadan, and Kano, the ‘old town’ of Zanzibar and Mombasa, and the royal neighbourhoods of Ouagadougou and Kampala.Footnote 107 In the latter cases, the influence of chiefs and townsmen of precolonial origin continued long into the twentieth century.Footnote 108 In a context characterized by the quick turnover of colonial staff, policy changes, and slow and uneven bureaucratization processes, many of these authorities dating from precolonial times maintained or acquired significant authority (see below).
Urban questions, especially those concerning labour, planning, poverty, and delinquency, became important issues within official circles both in Europe and Africa during and after the Second World War. These concerns favoured the creation of new departments in the late colonial bureaucracy.Footnote 109 The early colonial experience in towns might also have been decisive in setting up embryonic services and regulations though this has been less systematically explored. It seems from existing case studies that piecemeal responses to moral panics (over disease, unemployed youth, ‘unattached women’ or prostitutes) were subsequently transformed into public policies at the territorial level.Footnote 110 It is also possible – though this remains to be confirmed – that embryonic town services (municipal police, town planning, hygiene and welfare services) shaped the emerging central services of the colonies. This might represent a similar experience to that of Europe, where the emergence of the welfare state was often extrapolated from municipal experiments and the personnel involved in their operation.Footnote 111
While ad hoc town initiatives might have shaped the central bureaucratic state, it is also clear that colonial officials, like African political leaders, dramatically transformed urban networks. In the long term, public investments gave rise to different patterns of networks. In former French Western and Central African countries, the concentration of public investments in capital cities after the Second World War favoured ‘primacy’ that involved 30 per cent or more of a country's urban population living in its largest city. This arrangement meant that labour markets and infrastructure were concentrated in capitals which were then surrounded by networks of towns with relatively poor urban services, a model popularized by a famous formula, ‘the state stops at PK 12’, that is, 12 kilometers from the capital.Footnote 112 In most African countries, postcolonial leaders continued to invest disproportionally in capital cities neglecting both secondary towns and rural areas.Footnote 113 ‘Primacy’ as a colonial heritage of centralization still dominates Africa more than any other continent but has begun to decline over the last three decades.Footnote 114
CONTESTING THE STATE AND COMPETING AUTHORITIES OVER STATE FUNCTIONS
Post-1970s structural adjustment policies (SAPs) together with the reduction of state services and decentralization of government administration have facilitated the return of local power centres, increased the number of non-state actors involved in delivering services, and multiplied the patron-client ties and personal networks upon which people must rely to survive.Footnote 115 In itself, this process does not necessarily imply the weakening or privatization of the state.Footnote 116 Scaling down the focus of analysis to local arenas has recently helped to renew our understanding of the state in Africa. As argued by Christian Lund, the multiplication of parallel structures and alternative sites of authority (for instance, chiefs, political factions, hometown associations, neighbourhood groups, and vigilante organizations) has managed to ‘bring the state back’ into local arenas.Footnote 117 A growing though still limited literature on ‘cities at war’, in post-conflict situations, or towns beyond the control of government bureaucracies helps elucidate how the state can be built from the margins through unofficial agents imposing taxes on populations, building embryonic administrations, capturing the resources of humanitarian aid (for example, Goma, Lumumbashi, Juba),Footnote 118 or developing financial centres for money laundering, receiving migrant remittances, and recycling the goods of the regional economy (for example, Nador [Morocco], Benguerdane [Tunisia], Touba/Mbacké [Senegal]).Footnote 119
An historical approach that examines the contingent relationships constituted through conflicts and negotiations between state officials and multiple competing authorities in local and urban areas reveals processes of state construction and deconstruction in Africa. To oppose the rural and the urban here is probably of little help. Juxtaposing urban Africa as a privileged site of emergent civil society and of challenges to colonial and authoritarian regimes with tribal and despotic power in rural areasFootnote 120 overstates the role of urban political forms and undervalues the strategic place of rural constituencies and the importance of urban-rural links in the making of African politics. Political struggles during the colonial period have more often than not bridged the gap between the rural and the urban,Footnote 121 though this may be changing.Footnote 122 Nevertheless, urban areas should remain a significant focus for exploring day-to-day interactions between state and non-state actors.
Of radical importance are the effects of urban unrest and large-scale mobilization on the shape of the state. Heightened political tensions in cities have led to public actions and reinforced the role of bureaucracy as a means of reasserting the authority of the state over turbulent populations. Examples of such actions include the establishment of the late colonial state in British and French Africa after the urban riots of the 1930s and early 1940s, the creation of security services in Togo alter the revolt of Lomé in 1933, the notable reinforcement of the South African repressive apparatus after the 1960 Sharpeville massacre and the mid-1980s township revolt, and the recentralization of the Federal state stimulated by the 1966 anti-Igbo riots in northern and middle belt Nigerian cities that precipitated the civil war.Footnote 123 Moreover, ‘urban bias’ seems to coming back as a key concern for some African leaders after the so-called 2008 food riots.Footnote 124
These large scale mobilizations are probably not the dominant form of political and social confrontation in most of Africa's cities. Less visible, though no less important, are a set of fragmented confrontations and negotiations between the urban poor and the state. In a number of Middle Eastern and North African cities, mass protests, labour unionism and social movements failed – at least until late 2010 – to improve the living conditions of any significant number of people. Instead, poor people fostered what Asef Bayat calls quiet encroachment, characterized by largely atomized and prolonged mobilization with episodic actions by individuals and families to acquire the basic necessities of their lives, and to get access to collective resources (land, shelter, piped water, electricity) and to public space (street pavements, intersections, street parking places).Footnote 125 The contemporary urban landscape of the poor in several large metropolises seems dominated, on the one hand, by the apparent lack of public policies and, on the other, by the everyday bargaining of citizens with state officials and local political leaders. In Lagos, Nairobi, Casablanca and Cairo, the poorest neighbourhoods are, however, not necessarily abandoned by the central or the local state, as the ‘failed state’ literature contends, but instead remain strongly tied to it through networks of intermediaries, caïds, and political or union leaders, relationships that result in multiple forms of political belongings.Footnote 126 Such a situation might not look unfamiliar to historians of Africa, as large areas of cities under colonial rule were also poorly serviced and often dominated by forms of patronage and corruption, and under the nitpicking control of the state officials.Footnote 127 A comprehensive and continental assessment of what has fundamentally changed in urban based patronage and in the poor's access to services, land, and housing from the late nineteenth century to the present is not to be found in the existing literature.
Ultimately, the government's inability to ensure security for its citizens is usually considered the most important indicator of contemporary state failure in Africa.Footnote 128 Historically, however, coercive practices including collecting taxes, recruiting labour, implementing sanitary regulations, and policing towns and cities have often been delegated to ‘Native authorities’, foreign companies and local communities.Footnote 129 The colonial establishment of trained and professional police forces was slow and uneven, and never met the needs of the population. Local leaders and elders frequently demanded the provision of security by indigenous authorities, self-defence groups, and vigilante and community organizations in areas perceived as threatened by youth gangs (in some Nigerian and South African townships this has been the case since at least the 1930s). Such provisioning became a cheap way for the state to unload the expense of policing onto local communities.Footnote 130 Nevertheless, the results of this delegation – or in certain cases, this capture – of security functions were unpredictable. In the mid-1980s, no one would have guessed that street committees, neighbourhood watches or vigilante organizations would, twenty years later, become part of the South African police service through community police schemes and reserve police forces.Footnote 131 It was also far from obvious that former night guards in southern Nigerian cities, denounced as vigilante groups by the federal state in the mid-1980s, would become officially recognized and paid agents of states in the early 2000s.Footnote 132
To cast the multiplication of actors performing or subverting state functions as the ‘privatization of the state’ or as ‘the end of the postcolonial state’Footnote 133 is to analyze such practices in terms of a one-way affair and ignore the capacity of local authorities to operate in the twilight between state and society, public and private.Footnote 134 Looking at sites of authority that are linked to state officials and perform state functions might allow us to see new forms of authority being generated in apparently failing states.Footnote 135 Achille Mbembe suggests that private actors who perform functions such as imposing taxes and create moral or political order through violence and other constraints should also be understood as doing so for their own personal material gain.Footnote 136 Both arguments are probably valid, depending on circumstances. By placing actors who claim state authority in a larger historical frame, we might move beyond this dilemma and grasp how African public authorities ‘wax’ and wane' as state institutions are never definitively formed.Footnote 137
CONCLUSION
Considering Africa's cities as dysfunctional, chaotic, failed, informal, or not globalized works to retain the Western city as the paradigmatic model against which all others are to be assessed. Attempts to categorize Africa's cities as Islamic, colonial, postcolonial, and cosmopolitan similarly shed little light on the urban change that has taken place on the continent. These categories tend to underestimate, or leave unattended, more localized and regional historical contexts, and ignore the fact that urbanization in Africa has been, as elsewhere, a layered economic, social, and political process. The heterogeneity of urban situations and the extremely uneven influence of other continents in Africa ultimately suggest that it is more appropriate to speak about ‘cities of Africa’ rather than ‘African cities’. Simultaneously, social history examinations have revealed how the specificities of localized situations have greatly shaped the everyday practices of ordinary residents. Given the sheer diversity of these situations, a comprehensive portrait of urban Africa thus remains a difficult challenge to achieve.
In this article, exploration of cities as places situated within country, state, and world networks has been suggested as a way to move beyond localism and unhelpful categorization, and to avoid fetishizing scales of analysis. Linking world economic trends more carefully to urban labour history may enable reconsideration of post-war modernization policies as well as the domination of the informal sector paradigm since the 1970s. Labour history has always been an important subfield within African urban history and new topics of analysis such as land, leisure, visual, literary and the performing arts and public policies relating to education, health, housing, water, security remain to be explored. Much recent historical and social scientific scholarship has also insisted on the necessity of thinking about the city and the state, and the city and the world, simultaneously. Being urban, modern, civilized, or cosmopolitan appear, in a number of cases, to be overlapping processes. Making towns and making states were also simultaneous historical processes in many African societies prior to colonial rule, while negotiations and contestations between a range of local and imperial actors shaped the early colonial state in the 19th century. Reasserting the importance of cities in history will likely help us to understand better processes of state formation in Africa today. Exploring, in more empirical detail, the making of the city and the state and their uneven connections to the wider world will make the banality of Africa's urban past part of the history of the urban world.
Today, scholarship on Africa's urban past shares with its European counterpart some common handicaps: a dearth of studies of small and medium size towns, which renders difficult a clear understanding of urban networks and their connections to the world economy; the decline of economic studies focusing especially on cities; and more generally, often absent connections between stories of individual cities and broader trends.Footnote 138 Not surprisingly, these gaps and omissions have proven a significant obstacle to the few attempts to write a comprehensive and comparative history of urban Africa. Our analyses of Africa's urban past must move beyond the narrow subfield of urban history, which remains ill-defined, too narrowly empirical, and disconnected from larger debates. Instead, we should return to one of the subfield's initial motivations: the desire to forge interdisciplinary connections between history and the social sciences.Footnote 139
e des méties de police au Togo (1885–1963)’, Phd., University of Paris 7, 2011.
