Introduction: Africa's tragedy

In a controversial speech delivered to an audience of students and professors in Dakar on 26 July 2007, newly elected French president Nicolas Sarkozy offered a few meditations on Africa's relationship to the world, past, present, and future. After acknowledging the ills of the slave trade and colonialism, and the historical bonds they created between Africa and Europe, Sarkozy recuperated ‘history’ to make a very different point about the continental malaise.Footnote ¹ “Africa's tragedy,” he argues,

A disquieting blend of contradiction and historical fantasy, the French president's escapade into ethnophilosophy (Mbembe 2007) provides a fitting overture to the questions of history tackled in this paper, in at least two ways. First, negatively, the speech paints the very image of Africa that I seek to question. It is ironic that, in denouncing the myth of an African eternal return, Sarkozy recuperates exactly the kind of imaginaire he so vehemently rejects. By resurrecting the old Hegelian fetish of timeless Africa written out of modernity, he becomes the charmed audience of an Africa that has existed but in the minds of foreign observers (Mudimbe 1994; Miller 1985)! This leads him to embrace the curious notion that ‘Africans’ have somehow been insufficiently ‘historical’ or entrepreneurial in their engagement with the world – a thesis with which this paper takes severe exception.

Chiding the Dakar allocution for its shortcomings is not entirely fair game, in that the speech owes more to post-Enlightenment myth-making than to serious historical analysis. The fact that ‘Africans have histories too’ is not exactly startling news; in fact, exorcizing Hegelian demons from representations of Africa's past has been a central pivot of africanist scholarship since the time of independence. There is, however – and this is the second point – a more opaque level, on which Sarkozy's views show uneasy resemblance to academic conventional wisdom. What this convergence reveals, as I show below, is that the tropes that animate Sarkozy's narrative (the Atlantic world and colonialism, and Africa's tragic relationship to them) also organize competing forms of writing about the African past. The salient question, then, is no longer whether Africans have history, but the kinds of histories of Africa that have been written, the kinds of historical imaginations they have permitted, and the kinds of alternatives that are possible.

The following article examines these questions in the context of northern Senegambia, the area lying between the Senegal and Gambia river valleys and covering the present country of Senegal, where I have been conducting archaeological and archival research since 2001 (figure 1). Senegal has been a fertile terrain of historical, and to a lesser extent archaeological, engagement, producing research that speaks very keenly to the analysis of African experiences in the face of oceanic forces. The reflections that follow address this ‘historiography’ and the particular concerns that have driven it. In doing so, however, I also seek to make a broader theoretical point about understandings of Atlantic encounters in general.

Figure 1 Northern Senegambian states, ca mid-19th century.

My argument is that academic knowledge of the Atlantic past in Senegambia (and other parts of West Africa) has been governed by experiences of the slave trade and colonialism told and remembered in the present, which have recursively fashioned the terms of historical discourse in the region. In assuming, or reacting against, anticolonial readings of global encounters, analysis has narrowed the historical confines of Atlantic Senegambia to the slave trade and colonialism, and the nature of their impact on regional societies. These two moments have crystallized into ready-made abstractions that framed the African past around a given set of dynamics, while obscuring the historicities and experiences that may have exceeded these expectations. Following David Scott (2004), I propose that we extract ourselves from this historical project by redefining our relationship to the political and historical problems posed by the Atlantic era – in other words, by formulating new questions instead of supplying different answers to the quandaries of an earlier generation of historians.

One alternative, I suggest, is to reconceptualize the ‘problem-space’ of the African Atlantic around issues of materiality, temporality, entanglements, and the productivity of power.Footnote ² Specifically, I seek to expand recent archaeological advances in the study of Africa and its diasporas during the past five hundred years (e.g. DeCorse 2001b; Kelly 2004; Ogundiran and Falola 2007; Reid and Lane 2004b; Stahl 2001b), by exploring how artefacts and texts can be mobilized to produce narratives of the African Atlantic sensitive to different temporalities of action and rhythms of social experience. Drawing inspiration from important studies by Stahl (2002) and Ogundiran (2002), I propose that artefact trajectories, and the regimes of value that guided them, can shed promising light on these dimensions of Atlantic encounters, by offering a conduit into embodied experiences and the local negotiation of global forces. Turning to the Siin province in Senegal, I use recent archaeological and historical research to illustrate how object circulation and appropriation can complicate conventional readings of Atlantic dynamics in the region. Material from the Siin shows how Atlantic objects and forces unevenly rippled across the area, opening different strategies of action and forms of subjectivity as the region was gradually absorbed into a political economy increasingly configured by outside agencies. Finally, I reflect on the possibilities afforded by this analysis, and its potential for repositioning the Atlantic in relation to the past and future of Senegambia.

Atlantic perspectives on Senegambia

Pace Nicolas Sarkozy, Senegambia's immersion into worldwide circuits of political economy has been portrayed – often with reason – as nothing short of revolutionary. However, scholarship has generally been divided over how detrimental Atlantic forces were and how much agency Africans exerted over their own fate. On one side, an earlier generation of historians minimized the impact of enslavement, arguing that the balance of trade systematically favored Senegambia until the 19th century, and afforded new possibilities for economic development and political accumulation (Curtin 1969; 1975; Fage 1969). By contrast, a rival camp equated the Atlantic era with great socio-political turbulence marked by the rise of predatory states, rampant slave-raiding, crises of production, famines and generalized insecurity (Barry 1998; Becker 1988). Working through Atlantic commerce and colonial regimes, world capitalism, according to Barry (1979, 41), dealt a ‘fatal blow to all fields of artisanal production and interregional trade’, thus robbing local communities of their autonomy and binding them in relationships of dependency with exterior markets (Bathily 1989). Africans were forced into a broader process of peripheralization that relegated the continent to the margins of the world economy, and stalled its development relative to other parts of the globe (Rodney 1982; Wallerstein 1976).

While there is no question that the advent of Atlantic commerce transformed the course of African societies, one may wonder the extent to which broad-brushstroke portraits of development or devolution actually capture the complex nature of historical experiences in Senegambia. It seems paradoxical, for instance, that these momentous transformations should often be discussed as if they had taken place on a largely static canvas of endlessly reproducing social institutions and traditions (Barry 1998; Cissoko 1967; Curtin 1975; Pélissier 1966). The disjointed collage of political turbulence onto a fixed cultural backdrop seems to replay old anthropological dualisms splitting state from society, agency from structure, history from stasis, production from reproduction, and so on. In doing so, historical tableaux have consigned the precolonial cultural past to a qualitatively separate sphere of history, one bypassed by the operation of time, and fading into the penumbra of collective memory. Note here the (metahistorical) rapprochement with Sarkozy's views.

More questionable still is that Atlantic dynamics should have had uniform social consequences, that evenly spilled across all realms of social experience; that a vast region, of tremendous cultural and political diversity, should be adequately captured by single historical movements; or that, indeed, cultural structures may not have been overly affected by these changes. Forcing Senegambia into sweeping stories of (in the last instance) African autonomy or powerlessness that already presume the form, direction, and dynamics of history and culture simply obscures the distinct shape and character of once-living communities.

While recent scholarship has attempted to sail a more balanced course, in part through greater attention to regional specificity (Klein 1968; Searing 1993), researchers have not always managed to escape cut-and-dried bifurcations: positing favorable terms of trade or negative balance sheets, agency or submission, autonomy or dependence and so on. Nor have they avoided the tendency to write regional history from the standpoint of the best-documented polities, which were also the most implicated in commercial slavery and most susceptible to its disruptions (Klein 1992; see critiques in Diouf 2001 and Thioub 2002). And when authors have examined – often astutely – political dynamics in smaller or more peripheral polities (Galvan 2004; Klein 2001; Searing 2002), ‘culture’ has by and large remained construed as a retro-projection of the colonial ethnographic present into precolonial contexts (but see Baum 1999; Shaw 2002).

The salient point, here, is that a certain structure of argument shaped in the anticolonial (nationalist, pan-Africanist, dependentist, Afro-Marxist) discourses of the 1950s, 1960s and 1970s has also defined the terms of historical inquiry; the scope, direction and rhetoric of analysis; and the kinds of pasts that were possible. The twin ghosts of the triangular trade and colonial violence acquired particular significance in the political moment of post-independence national struggles. As particularly brutal forms of exploitation, yet ones that propelled the ‘forward march of global modernity’, these epochal moments are interlaced with a complex array of emotional, moral and political positions. The resulting dialectics of guilt and anger, of exoneration and accusation, of accountability and significance, considerably impacted the politics of the present in newly independent nations, and thus contributed significantly to shaping postcolonial academic discourse in Africa (Cooper 2005; Mbembe 2002). They offered a set of compelling causes to explain Africa's position in global political economy, and simultaneously outlined the obstacles that needed to be overcome to foster an alternative to the present situation. Reflected in the mirror of the present, this idea of history provided a moral and political compass delimiting particular species of historical imaginations and anticipated futures.

Reimagining the ‘problem-space’ of Atlantic pasts: historicities, materialities, trajectories

The crucial issue, as Scott (2004, esp. 23–57) aptly remarks, is the extent to which the interrogations and vistas of post-independence scholarship continue to be valid in the present, and how ‘old’ questions may continue to shape how we think about the past. The key, he argues, is not to pin ‘new’ answers to the old spectres of the past, but to redefine ‘problem-spaces’, by posing different questions, which can then mobilize the past to fashion new projects and expectations about the future. Thus, in the case of Senegambia, alternative histories lie perhaps less in demonstrating the impact of the Atlantic slave trade, or how much it actually stifled or lifted historical trajectories. Perhaps more important is the need to shift our lenses to different realms of historical inquiry targeting the temporal qualities, embodied experiences, and physics of power of Atlantic encounters. In this light, we should perhaps try to comprehend the Atlantic era as a configuration of discursive, material and practical relations, enmeshed in fields of power that both limited capacities for action and yet also afforded the ability to act – with varying degrees of awareness of the consequences – in the face of change (e.g. Foucault 1994, 298–325, 326–48; also Scott 2004, 127–28).

Important here is the study of Atlantic processes in their situated expressions, in a light attentive to the contexts and contingencies of continental history. Only then can we begin to retrieve what Bayart (1993) describes as the ‘true historicity of African societies’, where the past is no longer assimilated to scripted histories, but seen as a complex matrix of entanglements (sensu Mbembe 2001, 14–15) – that is, bundles of shifting temporalities, durations, deviations and spatialities that overlap, interlace or clash with each other just as they unevenly shape different actors in different fields of practice. From this vantage point, Atlantic experiences cannot be inferred from a bottomless ethnographic past that separates state and society, where local fates move at the behest of outside forces. The historicity of experience in Senegambia lay somewhere else, on a more ambiguous terrain where the political, the economic and the cultural were profoundly entwined, where the meeting of large-scale economies and local worlds was mediated by cultural logics, moral economies, political configurations, and conflicts. And, as will be shown below, the archaeological evidence accumulated so far seriously questions whether the introduction of European merchandise was as revolutionary in Senegambia as historian Abdoulaye Ly (1990, 251) once rhetorically surmised.

To explore the influence of political economic forces on local societies, we must tackle how local regimes of value and practices of social distinction shaped the African reception, consumption and desire for imported products (Stahl 2002). Drawing inspiration from a thriving anthropological literature on exchange and consumption (e.g. Appadurai 1986; Munn 1986; Myers 2001; Sahlins 1994; Thomas 1991), recent scholarship has usefully underscored the need to unravel how African modes of valuation, some of which pre-date the Atlantic trade, helped to recontextualize trade materials in ways that were locally meaningful, and how, in turn, these global contacts opened the way for new horizons of social practice (Ogundiran 2002; Piot 1996; 1999; Stahl 2001b; Weiss 1996). In other words, we must document the cultural translation of economic encounters before and during the Atlantic era, and the different forms of experience forged in these interactions.

Because they tend to conflate diverse levels of social response into generalized trends, familiar abstractions (regions, world system, the state, society, etc.) often gloss the minutiae of cultural entwinements. Thus, instead of packaging different domains of social practice and their material expressions, we might be better off unhooking them from one another, and examining their respective entanglements with outside systems of objects (Dietler 1998; Stahl 2001b; 2002). Here, objects – and their circulation in and out of spheres of production, exchange and consumption; between different political actors; along or across social boundaries – seem to provide a salient point of analytical entry (Appadurai 1986; cf. Van Binsbergen and Geschiere 2005). Retracing, to the best of our ability, the historical trajectories of different categories of object (Appadurai 1986, 13) can lend insights into how cultural choices guided the consumption of Atlantic commodities, and how, in turn, the circulation of trade imports introduced new resources and situations that in time could be converted into new material experiences, cultural logics and social relations (e.g. Lesure 1999). Because the relationships between objects and the people who use them are always (re)calibrated in given domains of social action, materiality, subjectivities and the webs of socio-semantic and historical relations binding them are enmeshed in a process of mutual constitution (Keane 2005; Meskell 2004; D. Miller 2005). In this respect, objects work to embody or produce social values in the context of cultural action, thus creating, reinforcing or diffusing the social orders that coordinate human exchanges: hierarchies of objects are linked to hierarchies of people; tastes for things translate into judgements of people; and object worlds and the ‘places’ they inhabit can effectively re-enshrine, transgress or abolish social difference, depending on the political projects that mobilized them (Keating 2000; Mills 2004). It is the nature of these links – their histories and qualities – and the contexts that shape them that must be determined.

The pasts materialized in objects can thus critically revisit portrayals of the African Atlantic as part of a worldwide process driven by a coherent (economic) logic, offering instead glimpses of a more constellated history of human transactions. Object trajectories thus privilege neither global sense nor local sensibilities, nor do they gloss the relations of power in between; rather, they open varied, if jagged, angles on these historical entanglements, looking at the production of global trends in localities and the fashioning of local worlds by forces well beyond them (Appadurai 1996). Interestingly, here the fragmented geography of exchange embodied in artefacts turns into a methodological vantage, as it lends us access to the ‘awkward scales’ (Comaroff and Comaroff 2003) on which social experiences of the Atlantic were fashioned and played out. Cultivating this ‘awkwardness’ in our own scales of analysis permits us to illuminate the tensions and (dis)articulations between these different realms of social, spatial and temporal extensions – a move absolutely crucial to capturing the ‘incompleteness’, uneven intensities, and frictions constitutive of different moments of globality, colonialism, and capital expansion (Cooper and Stoler 1997; Ferguson 2006; Hall 2000; Tsing 2005; Trouillot 2003).

The Siin: historical and archaeological contexts

Having delineated these few perspectives, let us now see how cultural transactions may have played out in Siin, a small coastal kingdom located in west-central Senegal (figures 1 and 2). Mirroring other regional polities, the Siin became entangled in the Atlantic world in the 15th century, as a supplier of slaves, salt, ivory, hides, textiles, grain and other natural products (Boulègue 1987; Brooks 1993; Curtin 1975). The Siin, however, provides an atypical entrée into regional dynamics, in that it remained a modest participant in the Atlantic system, secondary to the larger Wolof, Halpulaar or Mandinka polities surrounding it on all sides. The Siin is also interesting in light of its homogeneity and political organization. Many authors have inferred from oral traditions that the Siin was predominantly Serer in ethnic composition and highly centralized politically (Klein 1968; Mbodj 1978). In turn, they have proposed that these two characteristics may have combined with Siin's geographic compactness to shelter the region from the bulk of the disruptions generated by the Atlantic commerce. The documentary and archaeological records, however, provide a more nuanced assessment of Siin's political landscapes and their historical sensitivity to economic realities.Footnote ³

Figure 2 West-central Senegal: kingdoms and ethnic groups in the mid-19th century.

Less contested, however, is that while Siin remained active in coastal activities during the 17th century, it gradually distanced itself from Atlantic exchanges after that, probably as a result of the ascendancy of the Franco-British commercial hegemony which triggered a shift from hide- and textile-centred trade to one focused on slaves. As practices of enslavement intensified during the 18th century, fuelling a lucrative commerce in captives and the rise of internal slavery, the Siin may have been demoted to the rank of second players, in so far as the kingdom was never a major supplier of captives. Commercial activity, however, was not uniform, and the trade in captives and other supplies at Joal, Siin's coastal outlet, fluctuated wildly depending on political contingencies, regional conflicts and diplomacy, famines and so forth, with occasional surges in slave sales in times of war (Mbodj and Becker 1999). In times of peace, however, the Siin more readily supplied grain, cattle and other basic necessities to the French entrepôts of Gorée and Saint-Louis (Golberry 1802, 110–12; Le Brasseur 1977, 101, 121–22).

Against this background of relative disengagement, the Siin was gradually drawn into a codependent relationship with French coastal stations. And indeed, by the end of the 18th century, it had become locked with other kingdoms into a political economy of violence, where the cycle of debt and credit chained the fate of statecraft to foreign imports: horses, firearms, textiles and luxury items obtained from European merchants in return for slaves, food and other materials (Klein 1992, 32).

The passage into the 19th century ushered in profound reconfigurations in the economies of Senegambia, inspired by the British abolition of the slave trade in 1807 and gradual transition to ‘legitimate commerce’ (Law 1995). Existing merchant networks quickly compensated for the abandonment of human commodities, offering new economic options (gum arabic, incipient cash crops) that consolidated local kingdoms’ dependence on trade goods (Curtin 1981). Due to the lack of documents, little is known about the Siin during the first part of the 1800s, and we can conjecture that the region probably had limited interactions with French agents and authorities during that period. Following a change in metropolitan policies, it was absorbed into the sphere of French political influence in the 1860s, as part of a nascent colonial empire designed to satisfy the capital's need for new geographic frontiers and safe overseas markets for investments and economic surpluses (Klein 1968). Under these shifts in geopolitical gravity, African regions were given a new part in the rapidly expanding world economy: that of suppliers of raw materials and agricultural goods in return for goods processed or manufactured in metropolitan areas (Austen 1987, 42–46; Cooper 1993). Home to one of the most sophisticated agrarian systems in West Africa, the Siin became a centrepiece in the commercial apparatus of the Senegal colony, as one of the largest producers of peanuts in the French African empire (Klein 1979; Mbodj 1980).

Because of its peculiar engagement in Senegambian political economy, the Siin offers a welcome counterpart to those histories focused on major slave-trading states disproportionately represented in archival correspondence. More specifically, the region can help to redirect historical attention towards the manifold experiences and trajectories that comprised the Senegambian Atlantic as a historical space (Diouf 2001). While shedding initial light into these dynamics, the documentary record for the region is very uneven, and best exploited in juxtaposition with archaeological material.

Of the more than 180 sites (including loci subsequently marked out as sites) identified during the first regional survey in Siin (covering slightly more than 6 per cent of the province), over 110 could be definitively ascribed to the Atlantic period. Of these, over 40 habitation sites fell into Phase Va of the regional ceramic chronology (ca 1400s–1700s), while more than 90 belonged to phases Vb and Vc (18th–early 20th century), denoting the multicomponent nature of many of these occupations, and considerable overlap over time.Footnote ⁴ Atlantic-period settlements unfold as a mosaic of small and larger villages, ‘hamlets’ and more impermanent sites (figures 3 and 4). They present a number of long-term similarities, notably limited material accumulation indicating a pattern of residential mobility over time, and a certain structural homogeneity within phases, as various classes of site remain ‘relatively’ undifferentiated in their material expressions. Thus most sites will consist of a suite of trash and habitation mounds, varying only in number, size, surface features and artefact density. Interestingly, the size or material abundance of a site seldom seems to correspond to its remembered political importance (Richard, n.d.). Social space is notoriously difficult to ‘read’ due to the impermanence of sites and invisibility of architecture.

Figure 3 Settlement distribution in coastal Siin, 18th and 19th centuries. The grey rectangles indicate survey areas.

Figure 4 Settlement distribution in central Siin, 18th and 19th centuries.

The settlement landscape also exhibits considerable geographic and historical variation. Squarely centred along the coast and riverine channels before Atlantic contacts, the village habitat gradually relocated towards the hinterland after the 1500s. During the same period, habitation sites experienced a dramatic increase in number (and reduction in size and organization) to give way to a dispersed human landscape, made up of numerous small (often 1.5 hectares or less) and seemingly transient settlements.

Royal memory, village traditions and written documents were collected and consulted to anchor these sites more firmly in cultural context. While these sources must be treated critically, they often yielded information on the historical sociology of the settlements, the social identity of their occupants (royalty, aristocrats, slave-warriors, retainers, servants and domestic slaves, commoners) and their inscription in the remembered history of the kingdom, between recent migrations dating back to the 1960s and the more distant past of dynastic legends. Seven sites were selected for limited subsurface testing, offering a diverse sample of former capitals (Ndiongolor, Mbissel) and aristocratic settlements (Cupaan), villages associated with the foundation of the Siin polity (Dioral, Sagn Folo, Simal) and simple agricultural villages (Sorokh). Excavations consisted of a small number of trenches (generally two per site) and judgemental test pits sunk in trash middens and residential mounds, accompanied by extensive mapping and recording of surface features. Though not contemporaneous, these sites (or parts of them) nevertheless were occupied at various times in the past six hundred years. Their artefactual inventories, combined with those retrieved on surveyed sites, provide the raw material for the analysis below.

Regional assemblages in Siin are heavily dominated by local pottery, which accounts for over 90 per cent of their content and volume. Ceramic inventories generally feature low-fired globular jars and bowls of various sizes, decorated with roulette impressions, incisions and slips (Richard 2007, chapter 8). Most Atlantic-period sites will also yield a small collection of other artefacts, comprising local manufactures and trade imports, and largely composed of (in descending order of proportion) glass bottles (mostly wine or liquor, made in Britain, France or Holland), Venetian and Czech glass beads, tobacco pipes, oxidized metal objects and European ceramics (ibid., chapter 9). Although preliminary at this point, artefactual evidence manifests rather counterintuitive trends for a period so closely associated with displacement and flagrant inequality. At first glance, one of the most striking features of Siin's Atlantic assemblages is their homogeneity. The same kinds of local ceramics and factory-made goods (bottles, beads, pipes) are found across the region, and show limited spatial segregation. Instead, there seems to have been little, if any, accumulation of imported objects before the 19th century. The other key factor is the lateness of European-produced assemblages, which gravitate mostly around the post-1850s era. At the broadest of scales, regional archaeological patterns raise serious questions about the material impact of the oceanic economy, and its effects on political and economic power. How do these patterns break down when examined through finer sets of lenses?

Assorted entanglements: coastal trade and the logic of composition

Atlantic exchanges, early on and continuing into the 19th century, were locally mediated by what Jane Guyer (1993) has called a logic of composition (also Guyer and Eno Belinga 1995; and Guyer 2004, for a critical expansion). That is, social power was created not only by accumulating rare or precious commodities, but by selectively combining or composing different classes of goods, and the substances, knowledge, relations and use-values that they embodied. Value thus inhered not solely in particular goods but in the assortment of a variety of commodities.

Throughout the Atlantic era, commercial mémoires list long compilations of objects, which in shifting arrangements permitted certain categories of local or other goods to be secured. Thus in the early 17th century, Ruiters (1969) documented equivalences between imported and local goods – what could be obtained in return for what – making it all too clear that local tastes in consumption regulated the nature and direction of exchanges. In 1610, Lancaster, while briefly stopping on the Petite Côte, complained that although ‘the natives catch much fish, they [the crew] did not buy any, the expedition being devoid of crystal beads, white and blue counter [glass beads], bloodstones [carnelian beads], knives with yellow handles, iron bars, liquor bottles, without which one can procure neither cattle, nor goats, nor chickens, nor wood, nor water, etc.’ (quoted in de Moraes 1993, 143). In 1720, André Brüe indicated that beads had to be strung by colour and in particular arrangements to be saleable on the coast (Brüe 1720). In the mid-1760s, Demanet (1767, 246–47, 250) remarked that baubles were absolutely necessary to obtain captives and foodstuffs, in association with iron, liquor, beads, weapons, ammunitions and gunpowder. Ten years later, Le Brasseur (1776) lamented the indispensability of assortments in return for certain local products.

Collectively, these various accounts underscore the importance of compositional practices in coastal trading, and the distinct modes of valuation on which they rested. Just as wealth flowed from the strategic mobilization of people and relations, ‘things’ could be selectively combined in various arrangements to meet local demand and tastes. While ‘real values’ slid up and down against absolute price scales depending of the combination of goods being exchanged (Demanet 1767, 242–50; Doumet 1974, 54–56; also Guyer 2004), imported commodities became imbued with cultural value and meaning as part of trade assemblages, whose composition varied on the basis of what was being transacted. The widespread nature of composition in African commerce at the time of contact suggests the probable antiquity of these modes of social consumption.

Liquid trajectories: notes on the social life of bottle glass . . .

General trends in the structure of trade, commercial relations and imported products fluctuated both historically and geographically, following shifts in the dialectics of supply and demand. In turn, the kinds and quantities of goods that flowed inwards impacted different spheres of local production and consumption in different ways. As elsewhere along the coast, historic documents stress the early popularity of trade alcohol in Siin (Lemos Coelho 1985, 1.6, 1.13). Many authors mention that liquor was more heavily prized on the Petite Côte than further to the south where populations were more heavily Islamized (J. Lemaire (1615), in de Moraes 1993, 168; Paris 1976, 23). Official correspondence between local administrators and French charter companies documents increasing local demand for alcohol and iron throughout the 1700s, with alcool de traite (trade liquor) moving to the strategic centre of commercial exchanges in northern Senegambia (e.g. Demanet 1767; Lemaire 1887). How the inflow of liquor affected cultural uses of alcohol in local settings during the 18th century is difficult to evaluate. One of the persistent difficulties we face is the virtual absence of evidence on alcohol consumption practices in coastal Senegal before the onset of Atlantic exchanges. Archaeological research has been too limited so far to suggest much of anything before the 16th century. On the documentary side, while it is known that populations on the Petite Côte made wine out of grain (millet) and the fruit of the palm tree at the time of early contacts with Europeans (e.g. Almada 1984, 37), coastal observers shed little light on the sociology of alcohol use, thus offering thin comparative ground for evaluating later archaeological trajectories.Footnote ⁵ More intriguing still, considering the recorded antiquity of the liquor trade, is the fact that the latter has left practically no material trace on regional sites before the late 18th century. Preservation factors may in part account for this phenomenon. It is also possible that alcool de traite was traded in perishable, non-glass containers that would have no archaeological visibility.

This changes with the turn of the 19th century. In effect, the ubiquitous presence of alcohol bottles on archaeological sites in the Siin offers a material reminder of the central importance of liquor in international exchanges and processes of colonization on the coast of Senegal.Footnote ⁶ While liquor was but one element in a complex assortment of barter items during the 17th and 18th centuries, it rapidly eclipsed other goods during the 19th century and came to dominate the sphere of commodities traded on the coast. Case bottle assemblages clearly illustrate the reign of gin as the trade import par excellence after the 1850s (figure 5). Writing in the last decade of the 19th century, Noirot (1892, 452), who administered the Siin province, admitted that ‘genever liquor’ was the main item traded on the coast and indispensable to obtain local products (see also Mbodj 1978, 336; Rocaché 1904). The sheer quantity of gin bottles also speaks to the dynamics of colonial capitalism and supply circuits in Senegambia at the time. The liquor trade reflects the economic success and productivity of the Dutch distilling industry, whose products inundated African overseas markets through the marketing channels of other European nations. Before that, as archival references appear to suggest, alcohol imports were probably more diverse in provenance, and included French liquor and rum from the Caribbean, and possibly wine (Lemaire 1887, 53).

Figure 5 19th-century gin case bottle.

The liquor trade had a considerable impact on local consumption in 19th-century Siin. This is clearly illustrated in what European sources portrayed as the generalized practice of drunkenness on the coast. Colonial observers perceived the abusive local alcohol consumption as a sign of degrading mores and cultural anarchy (Bérenger-Féraud 1879; Pinet Laprade 1865). On one level, this bespeaks a conscious attempt to highlight the vitiated morals of the local aristocracy, and to strategically exploit the rhetoric of oppressiveness and violence to justify colonial intervention (Richard 2007, 166, 212–13, 215). On another level, the consistency with which writers of all backgrounds discuss local liquor consumption and inebriation over time suggests that alcoholism in Siin was no simple political fiction or colonial exaggeration, but a social pathology that intensified to the beat of global flows. This is expressed all too concretely in the hundreds of liquor bottles collected on regional sites, from aristocratic residences to peasant villages. Survey evidence thus confirms that alcohol consumption was not only limited to social elites, but encompassed the full swath of the Serer social spectrum in the 19th century. In this respect, the liquor trade indubitably eroded the local social fabric, as the landscape of violence and instability that crystallized in the second half of the 19th century would seem to suggest. More broadly, the situation in Siin joins a long list of examples stressing the part of alcohol in colonial world-building, as a vehicle of social and economic relations, an instrument of exchange and a component of imperial technologies of disruption and domination (Dietler 2006).

As with all material histories, however, there is a more nuanced side to the dynamics of 19th-century cultural consumption. For beyond the veil of rampant alcoholism, variations in artefact assemblages seem to lend suggestive light to the social dimensions of drinking and the role that alcohol may have played in the fashioning of social distance (Richard 2007, 607). Regional archaeological transcripts indicate that alcohol containers remain abundant on all 19th-century sites, yet a disparity emerges in the content of bottle assemblages between royal and aristocratic residencies and settlements inhabited by non-elites. Specifically, the former not only revealed denser and more diverse bottle assemblages, but also featured higher proportions of wine bottles in relation to gin case bottles. While this trend may simply index differential access to commodity circuits, excavated material suggests a more subtle set of cultural dynamics at play, denoting the enmeshment of wine in local grids of taste-making and practices of distinction, and its mobilization for the pursuit of social power.

Evidence of the latter emerged during our examination of a site cluster near the village of Ndiongolor, remembered in oral traditions as a secondary royal residence. The complex is also known to have episodically served as primary capital in times of political instability or succession crisis; most recently, it was the residence of renegade king Sanumoon Faay, who reigned in the 1870s. Our excavations at the largest locus – a small trench near the centre of a plaza littered with late 19th-century artefacts – uncovered a feasting pit, which contained a dense bottle inventory consisting almost exclusively of wine and demijohn bottle fragments, in blatant contrast with the expected high frequencies of gin bottles common to most regional sites.

While this provides meagre ground for generalization, in light of the repeated association of wine with aristocratic contexts and its presence in prestational ceremonies in Siin, it is tempting to wonder whether wine may have evolved into an icon of social difference, publicly singling out elite consumption practices from those of commoners (see Dietler 2006; Hamilakis 1999). As a highly public and collective ritual experience, feasting frequently forms an arena for the elaboration, reproduction or consolidation of strategies of power (Dietler and Hayden 2001). The feasting ground excavated in Siin could represent an instance of what Dietler (2001) has called ‘diacritical feasts’, where the display and consumption of precious or prestige items serves to mark and remake social boundaries. Contemporary documents provide discrete allusions to such an elite aesthetic in the Siin (Boukar Djilas n.d.), with resonances in other regions, where Bordeaux wines enjoyed considerable popularity within royal circles (Lamartiny (1884), in Gomez 1988, 151; Mollien 1818, 44). Adding independent voice to our case, oral memory in Siin recalls that wine (rather than any alcohol) was an essential element in the initiation ceremonies of high functionaries of the state by the end of the 19th century (Faye 2003).

The fact that an imported commodity held central ground in local ritual events and in the making of political personas speaks volume about the appropriation and recontextualization of foreign objects into local social practices and logics of power. While different variants of liquor may have shared a common path across the various domains of consumption in Siin, wine may have become imbued with social significance, translated as ritual potency, and replaced local and imported liquor as an indispensable embodied substance associated with political power. Substituting, differentiating and demarcating between imported products enabled aristocratic elites to alter regimes of taste, by creating hierarchies of objects and practices that spelled out particular readings of social bodies and judgements of people (cf. Lesure 1999; Weiner 1985). Because it was readily available to the average villager, and thus less pliable as a material and symbolic medium, gin thus was kept out of the calculus of invented traditions. At the same time, based on colonial and ethnographic sources, it is very likely that, among commoners, trade liquor continued to emulate the ceremonial or ritual role played by wine in aristocratic circles, as it was commonly used for libations, propitiation of ancestral spirits, annual festivals, public ceremonies, funerals, work parties and so forth. Still today, while these practices have subsided with the growing influence of Islam, the fragments of commercial liquor bottles continue to adorn the base of spirit or ancestral shrines in Siin, especially at the time of Serer traditional festivities.

Proposing that wine and trade liquor may have shaped parallel spheres of social distinction is clearly tentative at this point. Yet it offers an incentive to examine archaeological assemblages with an eye for material paths and diversions, and in turn for the windows they open on the sociologies inspiring these changes (Appadurai 1986, 16–17, 21). Diversions from socially prescribed uses and flows of objects may well refract in fragmentary light episodes of the social history of value in Siin, the symbolic contests and manipulations to which it gave rise, and thus the cultural anxieties of power in times of uncertainty.Footnote ⁷

Of other (small) things: trade, cultural strategies and regimes of value

Against the growing popularity of iron and liquor in Atlantic exchanges (Curtin 1975), glass beads continued to be imported at a steady rate. In European correspondence, beads are often depicted as essential items of trade, modes of payment, gifts or components of customs assortments along the coast, and were incorporated into local aesthetic systems and adornment practices (see references in DeCorse, Richard and Thiaw 2003). Coastal writers make frequent references to African men and women clad in bead paraphernalia, and the fervent demand that existed for them. This trend was vividly captured by Demanet (1767, 245–46) who marvelled at

Documentary sources seem to echo anthropological and archaeological analyses in other parts of the continent, which have underscored the centrality of beads to cultural and power transactions: as barters; tokens of currency; money; luxury goods; and ritual, religious or decorative objects (Graeber 2001, chapter 4; Kinahan 2000; Stahl 2001a). Unfortunately, the partiality of archaeological information makes it difficult to understand the role which beads played in processes of social and ritual reproduction in Siin, and how their introduction transformed local practical embodiment. Clearly, beads were not exactly a novelty at the time of early oceanic encounters but had long been involved in local and regional exchange networks (Richard 2007, 363–64) – another reminder that the Atlantic trade often did not bring items that Africans did not produce or had no access to, but simply redirected economic gravities away from hinterland areas and grafted itself onto existing commercial circuits (Thornton 1998). Alas, while bead assemblages have been recovered from funerary contexts in the Senegal River Valley, megalithic and tumulus belts, and the Saalum Island Delta, beads and other imported materials are glaringly absent from Siin sites before the 17th century. While this probably in part reflects problems of archaeological sampling, it also raises pointed questions about the extent of Siin's connection to Saharan economic spheres.

Although the absence of robust archaeological baselines prevents us from comparing pre-15th-century contexts with ones from the Atlantic period, it is possible to pick up the trail and movement of beads across later historical settings by juxtaposing material archives against the textual records. One notes, for instance, a discrepancy between 16th- and 17th-century European sources that show the economic significance and cultural appropriation of imported beads, and their near absence from archaeological sites dating to that period. Although this incongruity surely finds some of its origins in post-depositional distortions, conditions of preservation and the difficulty of identifying early bead specimens with accuracy, the absence of beads may simply indicate that an area like the Siin may not have been as active or involved in local exchanges as portrayed in European sources. Alternatively, it is possible that, in the early decades of the Atlantic trade, the circulation of overseas commodities like beads, and access to them, may not have been as widespread as documentary evidence suggests, but subject to tight control on the part of political elites and dignitaries. Archaeological evidence is too limited to date to suggest differences in the geography of bead distribution that would support preferential consumption or their use as insignias of prestige and social valuation, but it is quite possible that relevant contexts have yet to be investigated. In effect, some 17th-century sources comment upon the local use of beads to make and mark social difference (Almada 1984, 24; Dubois (1669), in de Moraes 1998, 145), a form of ‘practical logic’ that has been documented in other parts of Africa (Ogundiran 2002; Stahl 2002).

By contrast, the omnipresence of Venetian and Czech beads on historic settlements in Siin offers unambiguous evidence of the widespread demand for imported beads in the Siin during the 18th and 19th centuries. This is all the more interesting because this period appears to have experienced a precipitous decline in bead imports (Curtin 1975, 312, 318), as liquor and weapons and armaments supplanted earlier imports. The ubiquitous presence of beads on 18th- and 19th-century sites in various parts of Senegal (DeCorse, Richard and Thiaw 2003; Thiaw 2003) leaves little doubt that beads remained important elements of the regional economic and cultural landscapes, as many colonial authors have observed. A more nuanced expression of the continued cultural significance of beads in Siin (as commodities, currencies, body adornments and part of local sartorial grammars) can be seen in the growth of a local bead tradition during the 1700s, which probably has earlier roots in the 17th century. While perhaps fortuitous, we may wonder if the archaeological emergence of a local bead-making industry may signal a connection to the expansion of Atlantic commerce and shifts in commercial fluxes, although the meaning of these entanglements is uncertain. Were local clay beads used to emulate their imported counterparts, by whom and in what social situations? Did they acquire different values, functions and social messages? Perhaps clay bead production manifests a cultural response designed to substitute for episodic shortages or the possible decrease in bead circulation documented by historians. Perhaps local artisans may have tried to harness the local taste for beads by producing beads emulating their imported counterparts. Where these artefacts would have fitted in local systems of taste, and the kinds of cultural meanings, values and social messages they would have commanded, are uncertain. The small nature of the assemblage, and limited information on the contexts of clay bead use and production, only afford hypotheses.

Although less well studied because of its limited archaeological visibility, the fate of cloth seems to be telling yet another story. The Siin is described in Portuguese and Dutch sources as a textile-producing area, whose products were valuable items of trade in various parts of the African coast (figure 6) (Ruiters 1969, 108). Because local cotton cloth also served a domestic market, it is possible that textile production kept the demand for imported fabric much lower in Siin than in other parts of Africa until the 18th century, and maybe later (Anonymous (1670), in Thilmans and de Moraes 1977). Other documents, however, show that by the 1630s, Petite Côte traders purchased cotton cloth from the Gambia River from which they fashioned their valued garments (de Moraes 1973). Contemporary observers also allude to the possibility that local cloth may have embodied forms of social differentiation during the 17th century (de Marees (1602), in de Moraes 1993, 54; Fr Gaspar de Sevilla (1647), in de Moraes 1995, 363; Lemaire 1887, 54–55). Certain garments and textiles seem to have distinguished nobles and elites from commoners, in conformity with many other societies in Africa and elsewhere, where cloth and clothing, by virtue of their pliability, tactility and gradability, and because of their intimate linkages to bodies and selves, constitute veritable ‘social skins’, and were deployed and redeployed as political artefacts in daily negotiations of social geometries (cf. Comaroff 1996; DiPaolo Loren 2001; Hendrickson 1996). Because of its clothmaking industry, Siin was centrally involved in coastal exchanges during the Portuguese and Dutch periods (Dapper 1971; Lemos Coelho 1985). At the end of the 17th century, however, the rise of French and British commerce led to a rapid diminution of the trade in local textiles (de Moraes 1972a; 1972b). These changes probably heavily impacted Siin's economy and its position in exchange circuits, since the region lost an important source of revenue and medium of exchange. The gradual peripheralization of Siin in the growing Atlantic economy may date back to these reconfigurations, after what appeared to be an enterprising start.

Figure 6 ‘Negro weaver’ (late 18th century) (by R.C. Geoffroy de Villeneuve, 1814); image reference VILE-180, as shown on www.slaveryimages.org, sponsored by the Virginia Foundation for the Humanities and the University of Virginia Library.

What is less clear, however, is how the collapse of a coastal market for indigenous textiles impacted Siin's clothmaking industries. Artisanal production probably experienced a decrease, and indeed European accounts report that imported cloth was increasingly popular throughout the 18th century. We can also legitimately wonder how the surge in foreign cloth influenced the meaning and uses of textiles in Siin, and mechanisms of distinction reported a century earlier. How were imported textiles incorporated and reframed into local social aesthetics and body politics? What kind of value did they acquire, and did the meaning of local cloth get redefined in the process? Transformations surely occurred, but their cultural significance unfortunately lies beyond archaeological interpretation at this point. A very small collection of highly variable spindle whorls scattered across the region and spanning at least three centuries bespeaks considerable heterogeneity in manufacture. This certainly demonstrates some historical continuity in the practice of cotton spinning into the 19th century (Geoffroy de Villeneuve 1814, 179–83), although the concurrent scarcity of spindle whorls in recent archaeological contexts may reflect the fierce competition which imported textiles posed to local clothmaking industries.

Because of small sample sizes, no real geographic or temporal patterning emerges from the artefact assemblage. Following oral and historical accounts, one might have expected to see some spatial clustering of spindle whorls in the vicinity of royal settlements and attached satellites, where servants and craft specialists (including weavers) appear to have resided. At the same time, ethnographic and historical accounts also remind us that spinning probably was a domestic activity practised by all women, while loom-weaving may have been confined to slave specialists, artisans or casted weavers (see discussion in Richard 2007, 210). While we should be cautious about assuming continuity in craft traditions and technologies, particularly in textile production (Roberts 1996; Stahl and Cruz 1998), the heterogeneity of the spindle whorl assemblage does not contradict the possibility of part-time or unspecialized cotton spinning, and suggests the decentralized production of cotton threads.

While the Atlantic era was a narrative of disruption and turbulence, there is another side to the story that must not be overlooked. Atlantic entanglements entailed a certain ‘democratization’ of the trade economy, by bringing commoners into contact with new goods and possibilities. While kings and court retainers may have held special control over precious commodities (slaves, horses, iron), particularly early on (Almada 1985; Lemos Coelho 1985), textual references on rural women's bead assortments from the 16th century onwards hint that over time Atlantic commodities came to elude royal monopolies to fall within the reach of Serer commoners – something also attested by the unrestricted presence of European imports at all levels of Siin's settlement hierarchy. By the late 18th century, most villagers appeared to have been able to trade cereal surpluses in return for imported merchandise, as a way to fulfil the need for grain and provisioning in French trading posts. Le Brasseur (1776), for instance, deplored that ‘almost all the women on the continent have obtained [the right] to sell millet so they could request that the [trading] assortment contain all the baubles that they need. Have we ever traded such large quantities of millet without being obliged to give in exchange all the necessary merchandise?’ This congruence in documentary and material patterns lends credence to Searing's (1993, 51) suggestion that Senegambia's engagement with the Atlantic economy was structured by a ‘dual seller's market, one dominated by consumption goods valued by the peasants, the other by aristocratic prestige goods’. While the realities of political violence in the 18th and 19th centuries are inescapable, the convertibility of peasant goods and foodstuffs, and commodities acquired through the slave sales, ensured the widespread circulation of trade imports in the local economies and the participation of peasants in external exchange on an unprecedented scale (e.g. Piot 1996). These new exchanges and trade relations were elevated above the restrictions which traditional spheres of circulation and sumptuary laws imposed on peasant consumption and accumulation (Mbodj 1978, 81). By opening commodity circuits to peasants and non-elites, the Atlantic trade challenged local power relations, systems of inequality and the practices of distinction that reproduced them. This reached a particularly concrete expression in the 19th century, when peasants were able to trade agricultural products for weapons and ammunitions, which they could use to protect themselves against state violence (Klein 1968, 67; 1979). In return, the entry of peasants into a wider world of trade and consumption demanded adjustment on the part of elites, as the example of wine discussed earlier perhaps illustrates.

That a broader swath of Sinig society was drawn into global political economy thus does not entail its cultural subjection to interests and decisions fashioned in Western capitalist centres (Thomas 1991). The recent dates on trade import assemblages indicate that European goods do not make a visible material impact on the regional landscape until the 19th century. Colonial documents do not disagree with a picture of relative autonomy in Siin, even as villagers were encouraged to accept French authority through the medium of commerce. Thus, in the late 1800s, colonial administrator Ernest Noirot (1892, 457) lamented that the Serer were largely indifferent to European goods. Four years later, he reported that colonial authorities’ attempt to replace alcohol with imported cloth and make the latter the chief means of barter was largely unsuccessful, since the Serer were unwilling to let go of liquor and preferred African cloth (Noirot 1896, 54–55).

With the advent of the 20th century, however, a new page turned in the history of material practices in Siin. While, as a whole, Senegal's dependence on the world market had remained low until the late 1800s (Moseley 1992, 548), the explosion of colonial agriculture, the spread of monetization and taxation and growing reliance on French products effectively tethered local lives to the fate of peanuts, generally at the expense of local modes of subsistence, production and sociality (Noirot (1896), in Klein 1968, 174–75). Framed by the peanut trade and consolidation of colonial authority, this period was one of dramatic social and economic transformations, yet, true to nearly five hundred years of global enmeshments, one permeated by local negotiations and cultural syncretism (Galvan 2004; Mbodj 1978).

Conclusions (and openings): lights and shadows in portraits of the Atlantic past

The Atlantic era was a period of momentous social change, and reconfigurations rippled across a wide range of local social fields. Some of these changes have left tangible traces in the landscape, while other transformations in cultural practices and moral economies are bound to remain poorly known. One particularly under-studied phenomenon, for instance, is the impact of violence and depredations on social regimes of value. To illustrate, in a 1753 letter to the Company of the West Indies, French administrators informed the directors that the Siin, which was caught in a war with neighbouring Kajoor, ‘for more than 3 months cannot even supply itself with cattle; the famine obliged this country's people to eat them and they have become so rare than we are forced to furnish them from here’ (Conseil Supérieur de la Compagnie 1753). The same year, Estoupan (1754) attributed Siin's afflictions to

Considering the historical qualities of cattle as symbolic capital and collective wealth – the stuff of social contracts, relations and identities – in the region (Doumet 1974, 47, 91; Ruiters 1969, 119), the repercussions of this particular episode on local social logics must have been consequential. One wonders how the accretion of instability during the 18th and 19th centuries assaulted cultural conceptions of wealth, and the reproduction of social relations in peasant communities. In turn, the strategies and mechanisms developed by villagers to cope with insecurity are bound to remain obscure. The reality of fluctuations and reversals in constructions of social wealth becomes all the more apparent when we consider that a short hundred years after the episode described above, the decline of slavery and rise of the peanut trade may well have conspired to revive the importance of cattle as capital investment (Klein 1977, 352). Determining the fate of cattle, that prime emblem of social wealth, would be particularly crucial in unwrapping local practical logics of exchange and social construction. Livestock, however, remain frustratingly elusive in Siin's archaeological contexts, though work in other parts of the world, on faunal size and culling patterns, strikes a promising note for historical understanding (see Creighton 2006; Reid 2004).

This example reflects a perennial frustration for the study of African–European engagements – what Jane Guyer has called the dialectics of ‘turbulence and loss’ unleashed by Atlantic processes (in Stahl 2004a, 258). The advent of global connections engineered considerable uncertainty that dramatically reconfigured African societies, but also opened the way to new possibilities and trajectories. At times, however, turbulence was synonymous with loss, as certain strategies that had previously oriented political action fell by the wayside of history and memory. Our challenge, then, is to devise ways of retrieving these cardinal practices that may not have made their way into the ethnographic or documentary records.

In this article, I have tried to conjure up some of the historical fragments scattered by the turbulence of Atlantic history and its representations. I have drawn multiple sources to track the historical trajectories of different classes of objects across a variety of contexts. Even within a small area such as the Siin, and despite the disparity of evidence, analysis suggests that different fields of practice, production and consumption changed differently with the introduction of foreign commodities. This speaks to the braided diversity of local responses to global political economy, challenging conventional writings on Senegambia.

Rather than shifting wholesale or not at all – an impression sometimes left by historical treatments, particularly as regards matters of culture – material practices, via their agents, engaged unevenly with global market forces, and thus acquired their own social histories mediated by different conceptions of what is good, acceptable or possible, by different imaginations and representations of a changing world. It is in this context that we should understand why such avid smokers as Sinig villagers never quite adopted imported pipes (Richard 2007, 609–19), but quested after liquor and beads; why local cloth and imported textiles may have been locked in a complex dance of gains and losses; or why, indeed, the social trajectories of wine and gin may have bifurcated, as liquid commodities became embroiled in the politics of consumption. In this sense, I have also tried to show that any moment of consumption is also one of cultural production, or that production processes always already presume consumptive needs and desires. The same can be said of change and continuity, which are not historical absolutes, but whose expressions and ‘effects’ are tied to social architectures, moral economies and cultural sensibilities.

Tracking these dynamics requires particular attention to time and space, and to the levels of social experience channelled at different scales and by different sources. Thus, in navigating between feasting pits and settlement networks, between aristocratic residences and peasant villages, between the coast and the hinterland, between Siin and other parts of Senegal, with an eye on commercial flows radiating from Europe, I attempted to craft a narrative sensitive at once to worldwide political economy and the eventual stranglehold of geo-capitalist relations, and to their imbrication and actualization in local worlds – the way historic communities in Senegal made and unmade their localities in the face of changing tides and forces they tried to domesticate but could not fully control. When archaeological evidence proved too limited, I attempted to rub whatever was there against historical archives to open spaces of interrogations – raising questions about ‘History’ as we understand it, attempting to spot contradictions in sources, and suggesting directions for historical inquiry.

Clearly, these are very preliminary steps that leave much unaccounted. In part, this is because the effects of loss and turbulence are doubly compounded by what we could call the ‘pathos’ of objects, to stretch Simmel's (1971) phrase, in that they embody not only the limits of social relations, but also the limits of their own ability fully to re-present what these relations entailed. Objects both delimit and work in fields of social relations, and when the latter are no longer present or have transformed into something else, objects only afford partial reconstruction of the broader sociology in which they were inscribed. Another archaeological difficulty is that we are often left to infer the processes of production (of things, values, peoples) and exchange through the recovery of artefacts in their contexts of use or consumption, which introduces shortcuts in our portrayals of material histories.

Within these constraints, little is said, for instance, about power, or gender, or the presence of the state, which were undoubtedly pivotal constituents of life (Richard, n.d.). Also, much of the interpretation stays within object classes, and fails to examine transactions between categories of goods and the regimes of value sustaining them (e.g. Comaroff and Comaroff 2006; Graeber 2001). For lack of historical and archaeological resolution, the analysis reads the process of value production and the kinds of people it created through a rather coarse sociology, separating elites from non-elites but struggling to refine, historicize or indeed problematize these categories. A more nuanced exploration of value must involve more systematic analysis of the properties of objects and their spatial expressions: qualities (colour, tactility, sensuousness, material, shape, size, interior–exterior, etc.) and quantities (numbers, ratios, densities, etc.), scales of expression and relationality (asymmetries, ranking, grading, equivalences, compositionality, etc.) (Guyer 2004). The relational element is particularly crucial considering the historical importance of compositional politics in Senegambia, which should encourage us to look at artefacts in the context of the broader assemblages of which they are a part.

At the same time, parts of the analysis show occasional flashes of what a ‘thicker’ cultural description might look like, and thus offer a few promising guidelines. For example, the different trajectories of objects brought together in the analysis inspire a move away from directional histories and towards a more complex understanding of historicity, glimpsed at the intersection of different material pathways. Relations of force are also immanent to the analysis, although they are treated mainly indirectly, through their effects (Foucault 1994). Rather than being sidelined, power is highlighted through its generative properties, which constituted African history in relation to a global architecture of inequalities. Instead of fitting into simple equations of imposition/coercion, political and economic forces defined more or less expansive fields of action in historic Senegambia. While the fields of African actors may have grown narrower and increasingly uneven over time, which constrained their ability to effectively impose or alter the direction of exchanges, exchanges were constantly negotiated within a frame of available options. For example, the existence of different spheres of circulation between social classes, linked to each other and to Atlantic supply routes through different social calculi and conversion logics, speaks to the kinds of insight that the combination of history and archaeology can produce. In turn, this evidence hints at the potential of archaeological research in helping us to understand how objects create value and patterns of action in historical situations, how they define different kinds of people, subjects and persons, and how they enter the making of social power within a larger topography of exchange (Graeber 2001; also Piot 1991).

Remapping the problem-space of the Senegambian Atlantic to take greater stock of African historicity and subjectivity is not solely an exercise in fact checking and fact making. Beyond factuality perhaps, it also rearticulates our positionality vis-à-vis the past and, simultaneously, the relationship between our historical imaginations and the present and future with which they converse. The historical imaginary of Senegambia has been haunted by the spectres of the slave trade and colonialism, historical tragedies that are faulted for plunging Africa into an abyss of alienation – from itself, from its resources, from the world, from its historical path and its legitimate future. Part chronicle of loss and suffering, part moral commentary, Senegambian history has unfolded as a narrative of melancholia and betrayal, underwritten by a deep nostalgia for a once authentic Africa, whose descent into ‘nothingness’ under the blows of external intervention has been methodically chronicled, leaving its historical potential unrealized. In these narratives, moreover, the Atlantic era and colonialism also served as mobilizing tropes, around which Africans would rally in their quest for meaning, self-identity and sovereignty (Mbembe 2002, 242). If such a project failed to crystallize, as Mbembe (ibid., 272) indicates, it is in part because it mobilizes history to construct political futures as the ‘recovery of an essential but lost nature – the liberation of an essence – or as a sacrificial process’. Absent from this discourse is a sense of identity as history. Likewise, in framing the African Atlantic as a space of loss swallowing the continent as a whole, or one polarized between dependence and autonomy, post-independence writers and their interlocutors distanced themselves from the very historicity of these global moments: the contingencies, hybrid engagements with the world and paradoxes that are part and parcel of Africa's historical experience continue to animate the African present and to inform its future. In turn, it is perhaps the lack of openness of post-independence imaginaries to these dynamics that drove their political visions into an impasse, disjointed from the realities of African experiences (Scott 2004, 220).

Does that mean, as Sarkozy intoned, that African political futures lie in coming to terms with the Atlantic slave trade and colonialism . . . by turning this page of history over and turning themselves over to global obsessions? If the outraged responses to the Dakar speech are any indication, substituting oblivion for morality plays will not do either. More pertinent, perhaps, as I have suggested, is the recuperation of the Atlantic moment (and its colonial extensions) as a historical space – indeed a space of historicity – that was not simply a process of negation but was absolutely fundamental to the shaping of African memories and identities into the present (Austen 2001; Ferme 2001; Shaw 2002). This past is there to linger, not necessarily as a burden, but as a reminder of the violence (symbolic and material) of any cultural encounter, and as a reminder that African experiences of the self and of the world are born out of their intersection with global historical flows (Mbembe 2001). In this light, in addition to providing historical depth on the politics of difference, geometries of power, and tensions arising from the creation of states and subjects in the postcolony, the Atlantic past also offers a historical key into the cultural aesthetics that have mediated and emerged in relation to Africa's engagement with the world at large (Mbembe 2002).

While it is necessarily partial, and suspended in webs of silences of its own creation (Trouillot 1995), this reformulation of the problem-space of the African Atlantic forces us to confront Africans as producers of history and culture. In this light, it may offer a historical inspiration for new cultural productions and imaginations. Because of their openness to the reality of contingency, and thus their outlook on the future, these alternative pasts can provide starting points for the production of new memories and subjectivities, both within African nations and also by linking the chronotope of the Atlantic to other historical imaginations of Africanity (Clarke 2004; Gilroy 1993; Hanchard 1999). What form will these links and imaginaries take, how can they can be forged, and how will they enhance and constraint African cultural and political expressions? Only time may tell.