Let me begin by expressing my heartfelt thanks to Sarah Croucher, Audrey Horning, Tim Insoll and Peter van Dommelen for their thoughtful comments. It is not uncommon for academic critique to be seduced by the facility of criticism, to draw its impulse from what, with Freud, we might call the ‘narcissism of small differences’, and proceed with the obstinacy of the hatchet headed for the jugular . . . More artful, and far more difficult, is a form of engagement that gives more than it takes, that disputes but in the spirit of generous exchange. I feel that the respondents have accomplished precisely that. They have crafted a space of ‘hospitable’ discussion (cf. Derrida Reference Derrida2000), which rests on a double movement of ‘interruption’ (Westmoreland Reference Westmoreland2008): of their plane of thoughts, by bringing my article within the ambit of their conversations; and of my ideas, by making them a little less self-evident, a little less ‘at home’. This act of ‘making strange’ – when one's viewpoints acquire a different face as they are presented anew through the minds of others – is the cloth of which constructive dialogue is made.

As I read them, the responses also drive home a more general point about the hybrid constitution of academic writing, knowledge and authorship in a postcolonial world: that our arguments take shape and acquire significance in the act of addressing one or several audiences, that self-clarification is only achieved by exposure to others and their opinions (Butler Reference Butler2005), and that our ideas are constantly shot through with the outside world, and may follow unsuspected trajectories as they travel across different political fields and spheres of consumption (di Leonardo Reference Di Leonardo1998). Far from a liability, this relinquishing of control – accepting the overdetermined nature and openness of our narratives – also poses exacting terms for academic responsibility and integrity. Our indebtedness to different publics and interlocutors obligates us to our addressees. Our accountability to the past, archaeological evidence, and the stories we write is inseparable from our accountability towards the publics we write for and the descent communities whose pasts we write about.

These important questions of ethical responsibility have rightfully moved to the centre of archaeological inquiry in the past two decades, in Africa and beyond, and very much inform this reply – or indeed the obligation of replying (Bakhtin Reference Bakhtin1993), in a way that reciprocates the respondents’ generosity and does justice to their comments. While the commentators have raised more questions than space will allow me to examine here, one notes interesting convergences between them. And so, rather than a point–counterpoint progression, which would ultimately not be very useful (at least not in this context), I would like instead to use the responses as points of departure for broader reflections about postcolonial perspectives, colonial encounters and the making of Atlantic histories. I will do this addressing five clusters of themes: (1) epistemologies, (2) narratives and publics, (3) voicing and authorship, (4) materialities and (5) ethics and politics.

Postcolonial epistemologies, through thick and thin

In her response to the ‘Recharting’ piece, Audrey Horning expresses disquiet towards what she sees as a certain ‘thinness’ in the presentation of archaeological data, which makes her wonder whether I find much value in material evidence. These comments are entirely consistent with her important contributions to global historical archaeology. In effect, as Martin Hall (Reference Hall2010, 4) has recently noted, Horning (Reference Horning2007a; Reference Horning2007b) has emerged as one of historical archaeology's consciences, working tirelessly to keep its practitioners honest, and to hold them accountable to the promises/pitfalls of the craft. Particularly forceful has been her concern with the potential insufficiencies of postcolonial thinking in archaeology, and whether the field is truly relevant to our understanding of the past – whether it might not be an expensive (but fun) way of telling us what we already know. These points are well taken and command attention.

As I began to think of a way of engaging these comments, my mind somehow wandered to George Marcus's Ethnography through thick and thin (Reference Marcus1998). A play on words on Clifford Geertz's famous phrase, Marcus meditates on the analytical appropriateness of ‘thick description’ in a world no longer solely bound by the densities of ‘place’ but increasingly distended and displaced by global circulations. Sotto voce, the book seems to suggest that, on occasions, ‘thinner’ kinds of analysis might be better suited to capture the logics and labile expressions of global emergences, circuits and experiences (also Tsing Reference Tsing2005). Leaving aside the merits of Marcus's take on globalization, or the ‘degree of fit’ between ethnographic and archaeological situations, the book does raise an interesting point about the relationship between the mode of presentation, the calibration and robustness of evidence, and the nature of the problems we seek to tackle or points we seek to make. When are data enough to ‘say something’, and what constitutes acceptable support for interpretation? Are there contexts where ‘thinner’ expositions might be the way to go, or one way to get by in the absence of fuller investments?

Let me hasten to note here, that, in raising these questions and invoking Marcus, I do not intend to rake through the ashes of postmodernism in search for a cheap form of analytical evasion or justification for shoddy research, or to advocate a return to the endless regress of relativism. Nor am I unappreciative of deep empirical engagement and robust datasets. Surely, archaeological evidence is one of our safeguards against theoretical excesses, and its materiality both authorizes and lends palpable support to historical discourses (cf. Trouillot Reference Trouillot1995). But not all datasets are equal, and their quality, richness or absence also constrain what we can say, and how we can say it. This is especially true of many African contexts, where historical archaeology, despite tremendous advances in the past 20 years, often continues to trail behind with respect to datasets, empirical infrastructure, the kinds of question we can address, and what we can say about the past (DeCorse Reference DeCorse and Singleton1999). Such is the case for the Siin, for instance, where nearly all the post-1500 archaeological evidence for the region derives from my dissertation. In these contexts, where historical evidence is sparse (and confined to a thin coastal band until the mid-1800s), I would argue that archaeology, albeit thin and initial, is absolutely relevant to a richer understanding of Atlantic encounters, not so much in terms of ‘filling holes’ in the historical fabric, but in combining with other forms of evidence to generate different ways of looking – or to productively reconfigure problem-spaces, as I phrased it earlier.

Horning is correct that I could have provided a more precise and quantified account of archaeological materials. But I would also add that the article's thinness, its form, is also a product of its message. Surely, a big part of the article is devoted to discussing materiality, value and object trajectories, but those are mobilized to make a broader point about epistemology – one which, incidentally, expresses some reserve towards the promises of quantification. Debates over the Atlantic trade in Africa have often been conducted under the aegis of numbers and statistics, and operated by the principle that the discovery or addition of new evidence would help redress the mistakes of yesteryear and craft truer portraits of the past. Without denying that new empirical work has improved our understanding of the Atlantic trade (e.g. Eltis and Richardson Reference Eltis and Richardson2008), part of the intervention proposed in the ‘Recharting’ piece is to point to some of the epistemological limitations inherent in this approach. With Ann Stahl (Reference Stahl2001b, 33), I am doubtful that adding more evidence can always erase historical silences, in that these silences are also created by the assumptions and questions that frame historical inquiry (Trouillot Reference Trouillot1995). Nor does it necessarily result in more ‘complete’ accounts of the past (in that they often continue to reproduce the gaps on which they are built). And so rather than stitching sources together, as if assembling a puzzle, we might be better off starting from the premise of their ‘productive tensions’ (Stahl Reference Stahl2001b) – that is, confronting sources of different kinds to attend to the construction of history, historical evidence and historical representations, and think of new ways to imagine the past.

Horning places deserved stress on thicker discussions of archaeological evidence as a way to get at ‘substantive considerations of meaning and materiality’ and catalyse artefacts’ ‘expostulatory and explanatory power’ (p. 28). Without disagreeing with her, my sole concern is that such emphasis can, in its own inadvertent way, lead to an essentialization of archaeological information (data fetishism rather than fascism) and drive us back to reductionist debates pitting artefacts against texts (or other word-based sources) – thus replaying historical archaeology's long history of disciplinary anxiety and fear of subordination to history. Rather, I would suggest that what gives archaeological information its epistemic vantage is its ‘tense relationship’ with a broader repertoire of evidence, the fact that archaeological evidence's autonomy is a relation of heteronomy. Put differently, historical archaeology's strength, to me, does not simply lie in its access to artefacts and their raw concreteness, but in a particular sensibility to the material world, one predicated on bringing different lines of evidence to bear on an understanding of materiality in the past (and present). Strong illustrations of this approach can be found in recent archaeological accounts that craft astute histories of colonial materiality from an engagement with archival and documentary records (e.g. Hall Reference Hall2000; Dawdy Reference Dawdy2008). If part of the archaeological thinness of my account is undoubtedly a matter of empirical shortness, it is also because I have chosen to frame it as a historical anthropology, one that privileges materiality and the relationality of sources, rather than according de facto superiority to artefacts.

Narratives and publics

Horning identifies a tension perhaps not entirely resolved in the article when she asks who the audience is for this critique – one which also impacts the mode of presentation. In effect, who we are writing for also determines the form and content of our narratives, and perhaps also some of the reactions they generate (Stahl, Mann and di Paolo Loren Reference Stahl, Mann and di Paolo Loren2004). Again, writers, texts and publics are inextricably connected: publics come into being through the circulation of texts, by virtue of being addressed, while the existence of publics and of the fields of discourse they inhabit creates the conditions of possibility for academic texts (Warner Reference Warner2002). Part of the success of a discourse lies in meeting the expectations of one's interlocutors (Bakhtin Reference Bakhtin1981) – another feature of hospitable thought. Striking ‘the right note’, in turn, often implies a balancing act: knowing what to retain and what to sacrifice to make the argument intelligible, anticipating how one's ideas will be, imagining how much context audiences need and what they can live without . . .

While the intended audience was never quite specified, the respondents have no trouble identifying the various domains of ideas that informed the article: historical archaeology (Croucher), postcolonial studies (Van Dommelen), African archaeology (Insoll), and Atlantic world studies (Horning). Archaeological dialogues, rather than a more topical venue, provides an interface for addressing these different audiences and literatures. The public I am imagining is one conversant with archaeological theory, and concerned with conceptual interventions, which imparts shape to my writing. For this reason, while I recognize the mutual constitution of data and theory, the article is less an exposition of empirical evidence than an attempt to draw on such evidence to make a conceptual point about the relationship between history, discourse and politics – an argument which I think has relevance beyond Senegal or African archaeology. The imagined public I write for is explicitly not africanist, revealing another set of motivations left unspoken in the article, namely my feeling that African archaeology enjoys a somewhat provincial status in global archaeology, but that the continent's material pasts have important things to say about materiality, colonialism and historical production in other parts of the world at different points in time.

These background conditions framed my assumptions about readers’ expectations and how to meet them. This led to editorial decisions as to how much information to leave in or take out. In combining with my respondents’ interests and concerns, these calculations (or miscalculations) elicited different responses. Where Horning would like to see more precise engagements with data, Van Dommelen finds the superficial discussion of archaeological information less alarming, and in some sense justified by the level of argument to which it is directed (e.g. noting presences and absences). Croucher, whose teaching responsibilities involve a joint appointment in gender studies, asks rightfully about the absence of gender (and other subject positions) in my account, while Insoll, who has written extensively about African Islam, would like to hear more about how religious expressions mediated political-economic encounters in west-central Senegal. These topics are important but, for reasons of space, empirical resolution and intention, had to be left unaddressed (but see brief notes below).

Unfortunately, in this process of textual construction, formative parts of the story get elided: its context of emergence; the twists, turns, false starts and hiccups in its development; the surprises, discoveries and hesitations afforded by the confrontation of ideas and evidence; the various voices and conversations shaping its arguments; and so forth . . . By the time of a text's publication, the subterranean histories, which are essential to understanding its significance, have all but been erased, eclipsed as they are by the authority of printed word on the page. Oftentimes, in fact, the prehistory of our writing has little place in academic settings, where intellectual value (and its recognition) is structured by certain performances of authority, regimes of truth and modalities of certainty.

I would suggest, as others have (Joyce Reference Joyce2002; Stahl, Mann and di Paolo Loren Reference Stahl, Mann and di Paolo Loren2004), that part of accepting our answerability for the histories we paint, theories we choose and different publics we touch lies in making the process of narrative construction clearer; that is, in developing modes of writing and forums where one can both deconstruct historical discourses and simultaneously unpack the construction of our narratives – and thus formulate different criteria of evaluation. Attending to these issues would have perhaps helped to clarify why African voices were largely dropped from the analysis or why a discussion of local ceramics contained in an earlier version of the article had to be removed for lack of space, or examined how contradictions emerging ‘on the ground’ in the confrontation with local senses of history seminally shaped the argument made in the article . . . These could have helped to craft a richer context for the article and its reception, by explaining why the narrative took the shape that it did, and ultimately providing more solid grounding for the political claims of the piece.

Voicing: ‘other’ ways of imagining histories

If, as indicated above, the article and its mode of address were calibrated to an academic audience and were engaging academic discourses, the dynamics of problem-spaces and archaeological research have farther-reaching implications that trickle beyond academia (Richard Reference Richard2009, 117–21). The respondents notice an omission in the ‘Recharting’ piece, namely the absence of African voices. In doing so, they underscore some of the potential dangers or ironies of postcolonial theory, which, despite its theoretical advocacy of ‘other histories’ and petits récits, sometimes contradict or obliterate the very subaltern memories which it seeks to recuperate or celebrate (Liebmann and Rizvi Reference Liebmann and Rizvi2008; Loomba et al. Reference Loomba, Kaul, Bunzl, Burton and Esty2005). As I point out in the article, and as others have noted, every attempt to ‘revise’ particular discursive regimes and their ‘truth-value’ is itself situated within particular regimes of knowledge, which contain their own blindspots, silences and partialities. The idea of ‘problem-space’ is not innocent or objectively emancipating; it is itself conditioned by, and producing of, forms of authority and inequality.Footnote ¹ For, as Horning notes, what are the consequences of a de-privileging of the Atlantic slave trade (and colonialism), when local knowledges, historical understandings, and identities may well be built around these very motifs? Clearly, it would take another article to address this question in full, but let me make some quick points of context.

As Insoll indicates, perceptions about the past in Africa are often ‘vividly embedded’ and contested in a complex assortment of local discourses and historical memories framed in the present. Scholarship on the continent has shown that there are many ‘routes of remembrance’ of Atlantic encounters in Africa (Holsey Reference Holsey2008), and that different societies express, commemorate or forget the traumas of slavery and colonial violence in different fashions (Cole Reference Cole2001; Piot Reference Piot1996). Senegal is no exception. On Gorée Island, for instance, one of the iconic sites of the Black Atlantic (Ebron Reference Ebron1999), transatlantic slavery is centrally featured in official historical transcripts, no doubt because of the economic realities of heritage tourism. These memories, however, are not uncontested. For example, archaeological research on the island has downplayed the historical importance of global slavery and racial politics, by illuminating another side of the Gorean story, that of a vibrant multiethnic community, where domestic slavery and cohabitation between free and unfree residents actively shaped local experience (Thiaw Reference Thiaw, Sansone, Soumonni and Barry2008). In turn, this portrait stands at odds with the narrative of displacement linking black diasporic subjectivities, and local discourses intent on suppressing traces of indigenous slavery.

In Siin, by contrast, the Atlantic economy has left surprisingly few traces in historical memories. While Senegalese historiography has underscored the disastrous impact of Atlantic forces on northern Senegal, contemporary discourses of historical trauma in Siin revolve around the violence of colonialism and, more recently, that of the postcolonial state. Clearly, these discursive silences do not mean that Atlantic processes had no bearing on the structure of historical experience (Holsey Reference Holsey2008). There are many West African instances where the wounds of transatlantic slavery find little expression in language or memory, but quietly linger in other, more embodied realms of cultural production, such as religious performances, witchcraft and the ritual landscape, as noted by Insoll (Baum Reference Baum1999; Shaw Reference Shaw2002). The point, rather, is that the revisionist account presented in ‘Recharting’ emerged precisely from an on-the-ground engagement with the disconnect between academic and folk historical discourses, a ‘synaesthetic and largely pre-theoretical’ (Appadurai Reference Appadurai1996, 1) grappling with local historical imaginations, which was given theoretical shape by later readings in postcolonial scholarship. African voices thus echo throughout my discussion of ‘problem-space’, though they are not given sufficient recognition in the narrative.

Of course, there is no easy correspondence between the ‘histories’ emerging from my research and those told and remembered by people in Siin, in part because the politics of memory in the region are complex, and historical representations fluid in time and space (Galvan Reference Galvan2004). Our research has been at times welcomed by elders, teachers and interested parties, at others fiercely contested by historical stakeholders (descendants of the nobility, especially), but more generally flatly ignored by most villagers as something superfluous which tubabs (‘white people’) do and which does not concern them (Richard Reference Richard2007, 7–15). This phenomenon is not unusual in Senegal, where archaeology has a poor history of engagement with broader publics (Thiaw Reference Thiaw2003), and imagination of the nation places much greater attention on precolonial kingdoms, colonialism and Islam than on the Atlantic slave trade. This has perhaps been one of the most disappointing aspects of the project, whose objective was to unfold in dialogue with local memories to forge the kinds of alternative history evoked by Croucher. Clearly, success in this endeavour will not arise from the circulation of texts, and must involve modes of engagement suited to Siin's fairly dispersed village landscape, its mosaic discursive communities and the uneven literacy (in French or other national languages) of its rural populations.

Materialities

Many of the commentators’ questions concern issues of materiality (broadly construed). Notwithstanding the analysis's preliminary character and evidential limitations, I would like to think through some of their suggestions and outline how they can help to refine portrayals of Siin's Atlantic world. I agree wholeheartedly with Van Dommelen, for example, that, while the article aims to do justice to Siin's complexity, it does on some level maintain a separation between ‘colonizers’ and ‘colonized’. In effect, the hard lines blur quickly when we realize that for people in Siin during the period of observation, especially along the coast, European merchants had long been a familiar sight, that Africans developed complex relations with foreigners that ranged beyond trade or barter, and that these cultural interactions spurred the growth of a thriving community of Luso-Africans that endured well into the 19th century. By all accounts, the conditions for ambiguity and hybridity had long been established in the region by the onset of formal colonization in the 1860s, thus underscoring the need for more nuanced analyses (Bhabha Reference Bhabha1994; also Van Dommelen Reference Van Dommelen, Lyons and Papadopoulos2002). Likewise, as Croucher indicates (and as I lament in the article), the current state of artefact assemblages only enables us to scratch the surface of Siin's layered sociology, which far exceeded the elite/commoner distinction presented in the article. Historians of Senegambian kingdoms generally distinguish a complex architecture of social positions, differently ranked in relation to wealth and status. The four major social orders (aristocracy, freemen, slaves, socio-professional ‘castes’) showed numerous internal divisions that conditioned the kinds of political, economic, social and spiritual actions which cultural agents were able to exercise. What is interesting, of course, is the extent to which these political hierarchies appear to have been laterally tied by object circulation and redistribution, which index a complex, cross-cutting network of obligations, labour exchanges and social relations. By contrast, archaeological information for Gorée Island is showing a much different history of local entanglements with the outside world, marked by different experiences of class formation, creolization and social intimacy between European, mixed-race and free/enslaved African dwellers (Thiaw Reference Thiaw, Sansone, Soumonni and Barry2008; also Hinchman Reference Hinchman2000).

Van Dommelen is also correct when he notes that my reading of materiality pays little attention to the physical properties of artefacts. In effect, objects’ material qualities are always already implicated in their ability to carry or convey meanings, and thus in their capacity to signify value (Keane Reference Keane2007). To illustrate, I have long suspected that bead shape and colour were particularly important to African buyers along the coast, and may help to identify differences in consumption patterns, aesthetics and cultural preference between regions. Perhaps, as Van Dommelen suggests, similar considerations apply to other artefact classes (glass bottles, for instance), where form/texture/colour, rather than function/content, could have guided local practices of acquisition, consumption and reuse.

Another important dimension of local assemblages, which could not be included in the article, are locally produced ceramics. As argued by Croucher and Van Dommelen, this omission, and thus the primacy placed on more easily dated and better-contextualized trade imports, may lend inadvertent support to narratives stressing the triumph of global agencies over local ones. This is an astute point, and indeed a look at local ceramic assemblages suggests otherwise – or lends greater subtlety to the picture of entwinements between local crafts, the sociality and collective values they channelled, and broader economic circulation. Unlike other parts of West Africa (e.g. DeCorse Reference DeCorse2001a), where one finds over time the adoption of imported ceramic traits into local pottery repertoires, ceramic vessels in Siin show no incorporation whatsoever of European forms or styles prior to the 20th century. In fact, this lack of popularity applies to imported ceramics, which never quite found their way into local social and culinary worlds. Not only did Siin potters continue to meet local domestic needs until cheap, mass-produced enamelled metalware and cast-iron cooking pots began to flood African markets in the 20th century, but European ceramic forms were simply ill-suited to local practices of food preparation, serving and consumption, and not readily translatable into local systems of symbols. Today in Siin, food and the items associated with its collection, production and preparation, though heavily gendered, evoke an ethos of sharing, commensality, communal experience, and social belonging rooted in matrilineal identities. While ethnographic analogies can be misleading, it is likely that 18th- and 19th-century imported ceramic sets or vessel forms, with their focus on individual consumption and partition, may have projected an image of ‘asociality’ clashing with local conceptions of eating as a conduit of solidarity-building. At the same time, Van Dommelen does well to remind us that local ceramic assemblages, and the social practices they encode, did not remain static in time. Local vessel forms, types, and decorations become less variable during the Atlantic era. Likewise, the appearance of ‘steamers’ (perforated ceramic pans) suggests that the practice of steam-cooking millet couscous (which continues to this day) emerged during the era of oceanic commerce, though its origins probably lie somewhere else, perhaps Mauritania.

In tandem with the artefact trajectories mapped in the article, this schematic history of local pottery vessels supports Croucher's intuition that studies of consumption have much to tell us about colonial entanglements (Stahl Reference Stahl2002) – so long, of course, as studies of consumptive practices remain paired with concurrent analyses of production and exchange, at different scales. In effect, production, exchange and consumption are always co-constituted over time and space. Such a conjoint look ensures that our considerations of African consumption remain situated in broader fields of power, involving consumer markets, spheres of circulation and apparatuses of production in Europe, the Americas and Asia, and the local systems of labour, modes of economic organization and politics of gender raised by Croucher.

Extending previous arguments, Insoll is absolutely correct to note that placing too much weight on the transformative gravities of Atlantic forces, by itself, reproduces certain Eurocentric narratives (Feierman Reference Feierman, Bates, Mudimbe and O'Barr1993), while disregarding the fact that early Atlantic circuits grafted themselves onto pre-existing trading spheres and coexisted with them for centuries, that imported commodities filled no demand which local industries did not already satisfy, and that Africans for many decades controlled the terms of exchange. What needs to be examined in greater detail is this process of grafting, or, more broadly, the complex questions of transitions occurring between political- economic spheres over time (Stahl Reference Stahl2004b), and occluded behind catchy phrases like ‘the caravel replacing the caravan’. This demands a much closer archaeological engagement with the period directly preceding ‘the Atlantic moment’, so as better to contextualize changes that occurred during the era of European expansion. Equally relevant here are the temporalities of the long term and multiple durations of experience encoded in archaeological assemblages, which can help us to question and ultimately revisit historical periodizations for the continent, which have been so influenced by European historiography/modernity. For instance, understanding cultural contexts of alcohol consumption before 1500 (for which there is virtually no information) would provide much-needed baselines to assess the practices of drinking that developed during the Atlantic era.

At the moment, my discussion of the liquor trade raises the respondents’ concern. Horning's (Reference Horning and Shackel2002; Reference Horning2007b) important work on cultural stereotypy in Ireland and Appalachia makes her particularly aware of the unsuspected dangers lurking in certain portrayals of the past. She legitimately worries whether stressing the negative impact of the liquor trade in Siin might not have unforeseen consequences. Interestingly, I have in fact long struggled with this problem, or at least with the interpretations arising from my reading of both archaeological and historical evidence, and their broader implications. Discussing the disruptive effects of the liquor trade in Siin is not only precarious because it seems to suggest a certain passivity in the face of global commodities; more importantly, perhaps, it also runs the risk of reinforcing certain contemporary stereotypes about Siin's Serer populations. Because of the region's long history of resistance to Islam and its rural character, popular imagination in Senegal tends to regard the Serer as being out of sync with the country's Muslim, urban modernity. Although the Siin is today an epicentre of rural exodus and boasts a large Muslim majority, its inhabitants continue to be derided as culturally conservative and animistic, and thus heavy consumers of alcohol.

I have repeatedly shared these concerns with some of my informants, and they have consistently surprised me by downplaying them. Again, the reasons are numerous: as mentioned above, the Atlantic era is often accorded limited valence in the public sphere; archaeological texts have little circulation beyond academic circles; Muslim informants relegate excessive alcoholic consumption to a pre-Muslim past from which they feel disconnected; Christians and practitioners of local religion lay claim to alcohol use as a central feature of their non-Muslim identity, social practices and religious rituals . . .

Ultimately, I have chosen to retain this element, because omitting it would be omitting an important part of the story of Atlantic exchanges: that they were complex entanglements, that their effects were multisided and ambiguous, and that practical logics cannot be abstracted from larger structural forces. On one level, the introduction of trade alcohol triggered new cultural productions: it became an important part of different rites of passage, libations to ancestors and offerings to spirit shrines, celebrations of life and death, social occasions, labour parties and collective work. On another level, kings and their political entourage pillaged, fought and enslaved to obtain liquor, weapons and ammunitions, which escalated levels of social violence in Senegambia. Later, alcohol became an important medium of commerce (serving as currency at times), and played a key role in the fashioning of a space of colonial regulation structured by commodification, taxation and cash-cropping. This was not confined to the Siin only. With the usual caveats about European representations of cultural others, historical archives show that trade alcohol was appreciated by political elites in all coastal kingdoms, whether Muslim or not. As Horning anticipates, economic records provide empirical glimpses into the liquor trade. For instance, commercial documents for the French Compagnie des Indes, which serviced northern Senegambia, show that in 1718, ‘brandy’ or ‘spirits’ represented 25.4 per cent of all imports (in value) into the Siin, easily outdistancing the second-greatest importer of alcohol at 11 per cent. More generally, estimates computed for commercial records between 1718 and 1838 show a steep increase in the volume of alcohol (surpassed only by imported textiles) matched by a comparable decline in prices (Curtin Reference Curtin1975, 316–17, 322).

While I agree with Horning that material evidence should not be downplayed and that quantification should have a place in archaeological analysis, I am not sure that raw bottle-sherd counts can really tell us much at this point, because we are not sure what they represent socially. Again, here is a problem of commensurability across archaeological contexts. Unlike other parts of the world, depositional processes and the absence of permanent architecture in Siin make it difficult to ‘read’ spatial patterns on the level of the site and secure the social context of deposits, thus confining archaeological analysis to the burden of site-wide aggregate statistics. Combined with the fact that excavations have so far been small-scale, on sites occupied at different points in time over the past 1,500 years, there simply is not the critical mass of data and quality of context for the Atlantic period sufficient to delve into the kinds of subtle readings which Horning advocates. I have good faith that these readings will become possible in time, particularly as robust assemblages are collected on residential sites in other parts of the coast. Once a more solid comparative framework develops, it will enable us to delineate archaeological patterns in time and space, and compare them with commercial figures (made somewhat unreliable by the French companies’ archival negligence) and documentary accounts. There, quantification will surely offer an important window into social processes of alcohol consumption and assessments of the culturally mediated violence of the liquor trade. I am particularly interested, for instance, in the prospects of comparing Siin's archaeological materials with those from different loci on Gorée Island, where different observers during the 18th and 19th centuries also lament the abundant consumption of alcohol associated with company employees, traders and workers. Still, as we await vaster campaigns of excavation, one of the upsides of the article is that limited archaeological baselines may still allow us to ‘say something’.

Ethics, politics, ‘others’

I would like to conclude by returning to questions of ethics and politics. In starting with Sarkozy and ending on the Black Atlantic, via problem-spaces and object trajectories, the ‘Recharting’ article sought to underscore the inevitably political nature of discourses about Africa's Atlantic past, and archaeology's ability to create corrective, open-ended narratives that can reformulate historical understanding while opening the possibility for different or new political imaginations. By virtue of my own positionality in different fields of power/discourse/knowledge, this argument inevitably generates certain partialities, which my respondents have brought into clearer view. In doing so, they remind us that our choice of theory has real consequences for which we hold responsibility. This is so because our writings are given life and meaning by ‘others’ (people to whom we talk or listen, with whom we work or interact, with/for/against whom we write), whose engagement and responses we try to provoke; because our texts can travel across multiple realms of discourse, and sometimes pick up unintended messages along the way; and because not all ‘others’ are equal under the sun – some are indeed more precariously exposed or impacted by the ‘authority’ of what we say. Horning is thus right to alert us that making light of certain issues (data presentation, alcohol consumption) can be heavy with implications, that archaeology's lightness can be harmless fun just as it can be unbearable (politically).

In this light, I am grateful that, in inviting and challenging me to respond and clarify, the commentators also pushed me to become more accountable to the varied publics and audiences implicated in the problem-spaces I engage: those that may read this article (but do not get to respond), my informants, the people whose past I research, Senegal's public sphere, my students, people at conference, and so on . . . Relationally, the hyphen that connects self and others, that defines authors and audiences, that makes ethics and politics each the condition of possibility for the other, can also guarantee the productive articulation between different communities of memory across uneven terrains of power. As I have argued above, while local historical imaginations in Siin make few direct references to the Atlantic world and while academic visions of Africa's past are saturated by it, the kind of narrative I propose chooses to unpack the ‘time-knots’ entwining global history and local histories, and bring different discourses into conversation with each other, without one eclipsing the others (Chakrabarty Reference Chakrabarty2008). This perspective can help us to pry open our historical imaginations by vernacularizing narratives of global processes and ‘Atlanticizing’ African oral memories, to recognize that the past five centuries of African cultural experience remain unintelligible without the ‘Atlantic moment’ and that Atlantic modernity is unimaginable outside the archipelago of cultural worlds that constituted it.

Finally, recognizing the mutual embeddedness of different ways of thinking, telling or remembering history also poses the basic conditions of an ethics of writing – not simply a philosophical ethics concerned with fixed moral positions and ideas of right or wrong, but a relational ethics, sensitive to establishing a particular quality of relationship (Derrida Reference Derrida2002) between different pasts and knowledges about the past, unevenly situated in shifting worlds of power (cf. Pels and Meskell Reference Pels and Meskell2005; Hamilakis and Duke Reference Hamilakis and Duke2007). In effect, if ethics is the fact that ‘obligation happens’, then politics is the space in which the exercise of this obligation acquires its justification, meaning, shape, and effects (Simmons Reference Simmons2006, 89). The state of tension between ethics, politics and histories preserves the possibility of dialogue, the openness of our conclusions, and our accountability in the face of changing times, discourses and political realities. This might involve, for instance, appreciating the legitimacy of indigenous claims, sensibilities and views of the past, and making space for them, even as they may depart from what archaeological analysis seems to say. Or thinking through the consequences of our interpretations for descendent communities. Or doing a better job of discussing the multiple voices, tensions, and preoccupations entering the knowledges we produce (Stahl, Mann and di Paolo Loren Reference Stahl, Mann and di Paolo Loren2004).

If I have come any closer to fulfilling the conditions of such ethics in this response, it is, of course, thanks to my respondents, guardians of my integrity. And, although I get the final word here, the problem-space I have opened remains for others to expand, reframe, ignore or close.