Navigation – Plan du site
Débats et controverses

“Who Speaks for Africa?”

Patrick Chabal
p. 979-988

Texte intégral

  • 1 Henceforth referred to as Sardan.
  • 2 “Le culturalisme traditionaliste africaniste: analyse d’une idéologie scientifique”, Cahiers d’Étud (...)
  • 3 Patrick Chabal & Jean-Pascal Daloz, Africa Works: Disorder as Political Instrument (Oxford: James C (...)

1Under the guise of an article on culture and public service in Africa, Jean-Pierre Oliver de Sardan1 has produced a withering attack on what he deems to be a pernicious “scientific ideology”: Africanist traditionalist culturalism (atc)2.  His article is one of the sharpest critiques of the use of culture for political analysis by Africanists.  As one of the targets of his ire, I am naturally concerned to take his assessment seriously but the aim of this article is not to defend the two volumes he most censures3.  Rather, I would like to take up some of the intellectual and analytical issues raised by his considered critique to try to move the debate forward.  My ambition, therefore, rests on the hope that students of Africa — here an anthropologist and political scientist — can have a dialogue that will serve what I take to be our common purpose: understanding contemporary Africa better.

2Sardan’s basic argument is that the study of public services in Africa is not best served by reference to “African culture”, for two fundamental reasons.  First, because there is no such thing as “African culture”, or “tradition”, which can usefully be marshalled to provide a locally plausible explanation of how public services operate.  The whole notion is either too ahistorical or too vague.  For him, it is the type of lazy explanation given by those who do not trouble themselves with the empirical research required for meaningful explanation.  Second, the very use of a concept such as “African culture” is nothing other than a resort to a biased, Western, and patronising view of Africa.  As though those who seek to explain the continent’s present condition had nothing better to do than to fall back on the tired, indeed racist, clichés about the “heart of darkness”.  For Sardan, such an approach condemns us to ask the wrong questions and thus to misunderstand Africa’s “modernity”.

3Although I think that Sardan’s work shows that he is not as reluctant as he claims to be to take into account cultural factors, a point to which I return below, I shall try to delineate as plainly as I can the assumptions upon which his argument is constructed.  I hope in this way that some clarity in what appears to differentiate us may serve to move the debate constructively forward.  My response is in four parts.

What Is an Africanist or the Curse of Original Sin?

4At the heart of Sardan’s critique lie the two ur-questions about Africa or what I would call the elephants in the room: is Africa different and what is an Africanist?  The critique that consists in saying of some scholars that they resort to tired old clichés about the continent or that they are obsessed with Africa’s traditions are in effect code words for these two questions. Whether we like it or not, when it comes to the study of Africa there seems to be a prior (historical or existential) issue that comes into play: can non-Africans (and if so, which?) legitimately speak about the continent given the twisted nature of the relationship between the continent and the outside (chiefly Western) world?  I dub this the curse of the original sin.

  • 4 With the single exception of Axelle Kabou and Tarik Dahou.  Axelle Kabou, Et si l’Afrique refusait (...)
  • 5 Such as, for example, Achille Mbembe, whose work is always stimulating.  See among others Achille M (...)

5The assault on atc is clearly a reference to the use of cultural interpretation by non-African Africanists, since (with one unconvincing exception)4 Sardan does not mention any African scholar5 who might also invoke culture as a relevant explanatory factor.  So the question arises as to what is an Africanist or, to put it more provocatively, who is entitled to write on Africa.  The implication here is either that African scholars do not resort to atc because they know it is perniciously irrelevant, and demeaning, or that non-African scholars do so because they take the view that Africa is somewhat different from the rest of the world.  Sardan will probably dispute that such is his intention but his article leaves little doubt that he has strong views on the question of what is a legitimate Africanist approach.

6For the sake of a more open debate, I will try to outline the disagreement here.  The nub of the matter, to put it starkly, is whether there are pre-conditions to being an Africanist — either of origin, standpoint, ideology or approach.  In other words, should any argument about Africa be disqualified on account of the Africanist’s credentials?

7Put in this way, Sardan would most likely say “of course, not!” But let us probe.  Sardan’s argument that atc only refers to “un passé sans histoire […] un passé essentialisé” (p. 422) is clearly an attack on those who, for some (unstated) reason, would have cause to fail properly to historicise the African past — as opposed to the ways this has been done successfully about the past of other parts of the world (in for example the “histoire des mentalités”).  But why would these Africanists so obviously “discriminate” against Africa in their scholarly work if not because they are not qualified to be Africanists?

  • 6 Suffice it here to refer to Meagher’s very personalised review — K. Meagher, “Cultural Primordialis (...)

8It seems to me — but again this is not an issue that is ever discussed explicitly — that the study of Africa is marred by this curse of original sin. What I mean concretely is that it is virtually impossible to debate questions of interpretation without resorting to implicit judgement about the motives and qualifications of those who claim to be Africanists.  Leaving aside racist and politically correct attacks, which are becoming less frequent, there is in operation a form of self-censorship that dare not speak its name.  And this is most frequently expressed in the critique made of those who use concepts, or theories, that appear to refer to African specificities6.  Sardan’s attack on what he calls atc, although it is a perfectly reasoned argument, is also a warning that some approaches to the question of the relationship between culture and politics in Africa are simply beyond the pale.  But why should that be the case?  As Sardan quite rightly points out, just because one may stress the relevance of, say, the occult in Africa (among many other factors), does not mean that one is not aware of the relevance of similar processes in other parts of the world.

9And is the notion that Africa needs to be “protected” in this way not an inverted form of paternalism — since it is premised on the assumption that Africa is after all different?  And who, then, be the “guardians” of Africanist scholarship — or who is a bona fide Africanist?  This is an uncomfortable point.  I shall not take it any further.  But I wanted to seize this opportunity, finally, to bring it to the surface because I take Sardan’s piece to be judicious rather than prejudiced.  We must now move on.

What is Culturalism or Who Has Authority to Speak About Culture?

  • 7 “Cette sensibilité à la culture comme façon de penser la réalité des autres en sa spécificité est p (...)

10To be fair to Sardan, he grants my cultural approach some validity7.  This makes it clear that for him at least “culturalism” is not simply a question of identity but chiefly of scholarship.  So, the issue for debate is, to put it in its broadest terms, how best to study the relationship between politics and culture in the African context.  Having dismissed atc largely in terms of a critique of the ways some political scientists have (mis)used the concept of culture when writing about contemporary Africa, Sardan details his objections by way of a review of how the concept has been used in the anthropological literature — with which he is quite naturally more familiar.

  • 8 One might regret here that Sardan did not provide the historical context for the use of the concept (...)
  • 9 Sardan might have mentioned here that Culture Troubles offers a systematic critique of Parsons and (...)

11Following a useful historical overview of the genealogy of the concept of culture in anthropology, Sardan makes four important points8.  The first is that the original cultural anthropologists brought a welcome counterpoint to the teleological and developmentalist assumptions of the early ethnography, which marked a useful epistemological break.  The second is that Talcott Parsons was responsible for an analytical revolution that had two fateful consequences: it severed the study of culture from its multi-dimensional empirical base and it turned it into “une abstraction holiste” (p. 436).  Third, the “piège idéologique du culturalisme” was extended beyond anthropology, and all too readily embraced by other social scientists, seduced into using Geertz’s vague definition of culture as a shared system of meanings.  Lastly, Parsons made possible the “traditionalisation” of the notion of culture, leading thereby to the “great divide” between the so-called modern and traditional and to the generalisation of ahistorical studies9.

12So, what is “culturalism” beyond a codeword for “doing” culture the wrong way and why is it so insidious to the understanding of contemporary Africa?  According to Sardan, the two main effects of “culturalism” are to make culture an arbitrarily coherent whole and to allow for the unwarranted explanatory shortcut of traditionalisation.  In a word, it boils down to the objection that it results in an interpretation based on an “invented” notion of tradition that is completely divorced from empirical reality.  But, clearly, all the authors cited as guilty of atc would agree with Sardan’s objections here — including Bayart, whom Sardan exonerates of the worst excesses of “culturalism”.  So, if we are to move beyond these apparently irreconcilable differences, how can we approach this question beyond the crude charge of “culturalism”?

13With all due respect to Sardan, and he is not alone here, there is a contradiction at the heart of his argument.  On the one hand, he admits that culture — but only if used in a very circumspect and circumscribed manner — is relevant to the understanding of present-day Africa.  On the other, he denounces the use of the concept because it raises issues that are, somehow, uncomfortable.  I will not speculate further as to why such issues should be more uncomfortable in respect of Africa than of other parts of the world but will move on to see how Sardan attempts to square the circle, both conceptually and in practice.

  • 10 Patrick Chabal, Africa: the Politics of Suffering and Smiling (London: Zed, 2009), pp. 178-180.

14Conceptually, he wants to place two clear limits to the usefulness of the concept.  One is that it has always been more relevant to the study of early (so-called primitive) societies that were isolated and socially bounded, such as no longer exist in “modern” Africa.  Secondly, it can only be heuristically relevant if it is seen through the empirical lens of what the actors themselves actually do in locally identifiable situations — never in general or in the abstract.  Hence, his objection to Bayart’s notion of the “politique du ventre”.  Of course, this is in effect an argument against generalisation, which I rehearse it in some detail in my last book10.  I suppose one must accept that some of us are allergic to generalisation but the question is whether we need to rule out those who see merit in it.

  • 11 See Jean-Pierre Olivier de Sardan, “A Moral Economy of Corruption”, Journal of Modern African Studi (...)
  • 12 Sardan summarises the situation as follows: “Nous avons été nous-mêmes, à notre grande stupéfaction (...)
  • 13 Tarik Dahou, “Déculturaliser la corruption”, Les Temps Modernes, 620-621, pp. 289-311.
  • 14 John Iliffe, Honour in African History (Cambridge: Cambridge University Press, 2004).

15In his writings, Sardan is much more convincing, and here I will touch briefly on the (no doubt) painful charge he faced when he wrote very illuminatingly about corruption11.  His argument is quite revealing and takes us to the heart of the issue of who has authority to speak about culture, and how12.  Stripped of the shock that he should have been taken to be culturalist, his exculpation comes close to saying that there is only one way of “doing” culture when proposing a credible account of a phenomenon like corruption, which is not specific to Africa.  The interpretation of what Sardan calls the “cultural logics” of corruption forces recognition of local factors, of which culture is necessarily a part.  And there seems to be no cast-iron method whereby one can both incorporate culture and avoid the charge of culturalism.  If that is the case, then we are all faced with the same dilemma when we try to make sense of what we observe.  But Sardan’s summary dismissal of Dahou’s critique reads like arbitrary adjudication13.  Indeed, what qualifies Sardan to say that the recourse to what Dahour labels “la logique traditionnelle de l’honneur dans le Waalo” is culturalist?  After all, respectable historians of Africa have written about the place of honour with great insight14.

What Is Public Service or What’s Culture Got to Do With It?

16Returning to Sardan’s article on atc, the origin of the argument lies with how he sees the debate about the performance of public services in contemporary Africa.  His claim is that those interpretations that rely on cultural (or, in his words, traditionalist) explanations utterly fail to understand the full complexities of the factors that have a bearing on the organisation and functioning of public service delivery.  By diverting attention to vague and mostly “imagined” considerations of African specificities that are inimical to the workings of these services, they over-interpret the difficulties faced and confine all efforts to teleological failure.  They tar Africans with a brush of ahistorical incapacity, which merely confirms pre-existing prejudices about their inability to function as “modern” men and women.

17Sardan’s argument is powerful and the large body of research he has conducted on public services provides concrete evidence of the varying dynamics involved.  In his work he is mindful of the need to study a certain type of “culture” — such as, for instance, “la culture professionnelle des douaniers ou celle des magistrates” or “la culture électorale prévalente lors des scrutins municipaux” (p. 445).  In other words, what is known in the social sciences as the culture of organisations or processes.  What he excludes is any reference to a more general cultural context, especially one that might be thought to be “African”.  But this is odd, especially for an anthropologist.  The relentless focus on the “situations africaines contemporaines”, which atc is supposed to ignore, is itself bereft of the historical depth and attention to the slow movements of socio-political evolution, which characterise all societies and the knowledge of which is a necessary pre-condition to the assessment of their current situation.  For example, it would be odd indeed to try to make sense of the Japanese post-War bureaucratic culture without some reference to the country’s history and, indeed, its deep “culture”.  So why should it be different for Senegal or Mali?

18The crux of the matter, therefore, is that Sardan’s objections are arbitrarily set to exclude a range of cultural considerations about the possible relevance of the general cultural context within which public services operate. No-one would deny the merit of the closely circumscribed research that offers a fine-grained picture of what is happening to health workers in a district of Niger.  But why should one rule out a priori consideration of cultural practices that may impact on the efficacy or durability of a public service that is not working as it should — especially if there is good empirical evidence that such is indeed the case?  The question surely lies in what ‘tempirical’ means, a point to which I return below.

19For now, suffice it to take note of the fact that Sardan’s critique is in effect an attempt to disqualify a particular type of work on “culture” on what I would suggest are particularistic grounds.  In other words, the objection rests on an intellectual sleight of hand: lump all cultural interpretations under the label of atc so as to disqualify the heuristic relevance of cultural factors in the assessment of public services in Africa.  Yet, Sardan’s excellent work on the moral economy of corruption is proof of what needs to be done as soon as one is concerned to provide a less parochial explanation of certain recurrent processes (could they be tendances lourdes?).  So one can only conclude that his attack on atc, and that of many who share his views, is prey to a personal, subjective, contradiction between the need to protect Africa from culturalism and the knowledge of the relevance of culture to the understanding of its present condition.

20So let us now turn from the critique to the prescription.

What Is To Be Done or the Fetishism of the Empirical?

21Sardan’s argument is clear cut: the only legitimate means of studying public services is empirical.  And by implication, therefore, the only legitimate way of taking “culture” into account is to refer to the empirical reality of the organisational culture with which one is dealing.  Other contextual factors, insofar as they are considered, must also be determined by the specific demands of the empirical research being undertaken.  So, for instance, issues of ethnicity would only be addressed if the research showed that they were empirically of some relevance — not as background or historical context.  Thus, one should refrain from making any assumption about non-parochial cultural considerations one might expect would matter from other empirical research about that particular society.  But it is difficult to understand how such a rigid distinction could be made in practice.

22Furthermore, it is also questionable why such a separation should be made in the first place.  Is it really desirable that empirical research should be defined in such narrow terms that it prevents us from seeing the whole elephant — especially when we are dealing with the complex reality of public services, the operation of which is quite obviously the result of intertwined and fluid logics that relate to all aspects of society and politics?  The truth is that we would not advocate such thinness of approach if we were to study the bureaucratic culture of our own public services.  So why be so severe in our definition of the empirical when it comes to Africa?  Are we not in danger of limiting the value of our conclusions for eminently subjective reasons?  Do we really need a specific African type of empirical research?

23Taken to its logical conclusions, Sardan’s instructions risk leading us into what might be called “Africanist Empirical Fetishism” (aef).  Let me explain.  Above and beyond the call for an empirical — as opposed to, say, theoretical or conceptual — approach, which is a perfectly legitimate demand in the social sciences, there is a danger here that one might construct, or invent, a type of empiricism that is especially tailored to the study of Africa.  In other words, the well-meaning intention to “protect” the continent from the prejudiced damage caused by atc, might in the end result in an equally injurious form of paternalism: aef.  If we are to have a debate about the superiority of empirical research one must avoid doing so on the dubious grounds of African specificity, as no doubt Sardan would agree.

  • 15 See here particularly, Jean-François Bayart, L’État en Afrique (Paris: Fayard, 1989).

24Moving the discussion forward, then, I would suggest the atc label is not a constructive way of dismissing the work of scholars who have made the effort of trying to explain how “culture” may be relevant to the understanding of observable socio-political phenomena.  Sardan’s critique of Bayart’s work15, well-taken as it is, reveals an a priori rejection that flies in the face of well-identified empirical observations based on in-depth research in the field and detailed literature review of other studies.  The charge that political scientists are not anthropologists — implying they lack the skill and dedication required of “real” empirical research — is both facetious and prejudiced, as I am sure Sardan would agree.  Social scientists work differently but there is no reason to suppose that only some of them are concerned with understanding reality “as it really is”.

25The question, then, is not so much to prescribe a certain type of approach but to be open to the reasons why we might choose one way rather than another or why we might wish to use a combination of approaches rather than arbitrarily to restrict ourselves to a single method.  So the real dilemma is…

Can we Ever Conceptualise the Relationship Between Culture and Politics?

  • 16 Those who are interested can turn to Culture Troubles, which provides a systematic discussion of wh (...)

26It seems to me that we can only make progress on this if we accept that we must cease to view Africa through Africanist lenses.  Yes, we are all Africanists if we study Africa but that is all.  And many of us, comparativists, are not confined to one continent.  The attack on atc is therefore a diversion from what should be a more general concern about how best to conceptualise cultural factors in social science research.  There is place for debate here but I would want to highlight some important considerations, which I believe are relevant to a more open approach to this contentious subject. There is no scope here to develop, so I will confine myself to two broad remarks16.  The first has to do with standpoint; the second with the use of induction.

27The problem with Sardan’s critique is not so much the vehemence of the attack on atc, which is widely shared in some circles, but the absence of perspective.  One way to avoid needless disagreement is to make clear where one stands and why.  We must all of us incorporate into a presentation of our epistemological and methodological preferences a discussion of why we come to the question the way we do — as individual, scholar and practitioner — and thereby make explicit what we believe to be the strengths and limits of our undertaking.  No-one is above personal, academic or intellectual “suspicion” and no-one should pretend to stay above the fray, least of all when passing judgement on the credentials of other scholars.

28But the question of standpoint goes beyond these considerations.  It is also imperative to situate oneself within the various traditions that are relevant to our approach.  This means not only identifying oneself as an anthropologist, historian or political scientist but also explaining why one favours a particular type of anthropology, history or political science and why.  There is no consensus on how best to study human affairs and it is pointless, as well as restrictive, to imply otherwise.  Finally, it is important to justify one’s interests for a more or less comparative focus and to explain why a certain level of generalisation, or empirical narrowness, is desirable. One cannot simply take it for granted that one is, inherently, superior to the other.

  • 17 See Culture Troubles, Chapter 7.

29On the question of induction, which I have developed elsewhere17, I would argue that Sardan’s a priori definition of what is the “right” kind of empirical research is not as heuristically fruitful as may appear at first sight.  The difficulty in assessing the significance of culture in the study of African societies lies less in the definition of the concept, as Sardan seems to think, than in working out how to find out about it.  And here it would seem more profitable to work in steps, going back and forth between the apparently neutral empirical observations of the case study Sardan advocates and a pointed querying of the relevance of more general knowledge about important cultural factors known to influence agency in similar African societies, as some of us would recommend.

30This, in any event, is what most of us would favour when thinking about how best to understand the cultural factors that affect the behaviour of our own politicians, public servants and the working logics of our own institutions.  Political scientists have worked on political culture for some considerable time and, however controversial that approach has been, it is a perfectly respectable way of suggesting ways of factoring culture in the study of politics.  How relevant it may be in specific instances is precisely the sort of question that induction ought to help answer.  So, to repeat myself, it seems strange that we should not want to be as inductively flexible about Africa as we are inclined to be about what we know of our own societies.  In other words, there seems to be no good reason to rule out the benefits of this inductive method, which has a very long pedigree in the history of the social and human sciences.

Haut de page


1 Henceforth referred to as Sardan.

2 “Le culturalisme traditionaliste africaniste: analyse d’une idéologie scientifique”, Cahiers d’Études africaines, L (2-3-4), 198-199-200 (2010), pp. 419-453.

3 Patrick Chabal & Jean-Pascal Daloz, Africa Works: Disorder as Political Instrument (Oxford: James Currey, 1999) and Patrick Chabal & Jean-Pascal Daloz, Culture Troubles: Politics and the Interpretation of Meaning (Chicago: Chicago University Press, 2006).  It is strange, and a little troubling, that Sardan omits to say that Culture Troubles is not an Africanist book but a book on comparative politics from a cultural perspective.  So, why portray it as a “culturalist” interpretation of Africa?

4 With the single exception of Axelle Kabou and Tarik Dahou.  Axelle Kabou, Et si l’Afrique refusait le développement (Paris: L’Harmattan, 1991), is commonly considered to be the most exaggerated cultural explanation of Africa’s underdevelopment and the author is often labelled the equivalent of the “self-hating Jew”.  I refer to Dahou’s critique of Sardan below.

5 Such as, for example, Achille Mbembe, whose work is always stimulating.  See among others Achille Mbembe, De la postcolonie: essai sur l’imagination politique dans l’Afrique contemporaine (Paris: Karthala, 2000) and Sortir de la grande nuit (Paris: La Découverte, 2010).

6 Suffice it here to refer to Meagher’s very personalised review — K. Meagher, “Cultural Primordialism and the Post-structuralist Imaginaire: plus ça change […]”, Africa, 76 (4), 2006, pp. 590-597 — which Sardan himself considers over the top.

7 “Cette sensibilité à la culture comme façon de penser la réalité des autres en sa spécificité est plus qu’honorable: elle est elle-même au fondement des sciences sociales empiriques, et nous la partageons tous.  C’est le contenu stéréotypé et traditionnaliste qui se faufile en contrebande derrière ces bonnes intentions qui pose problème”, Sardan, op. cit., p. 433.

8 One might regret here that Sardan did not provide the historical context for the use of the concept of culture by other social scientists, including the founders of modern sociology and politics (e.g. Weber).  He might have found more connections than he is prepared to see with today’s Africanist scholarship.

9 Sardan might have mentioned here that Culture Troubles offers a systematic critique of Parsons and Parsons’ legacy in comparative political science, which rehearses many of the very points he makes.  And by providing a comparative analysis of the understanding of the State in, respectively, France, Nigeria and Sweden, it made possible the assessment of our approach above and beyond the Africanist perspective — which readers of his article would not have surmised.

10 Patrick Chabal, Africa: the Politics of Suffering and Smiling (London: Zed, 2009), pp. 178-180.

11 See Jean-Pierre Olivier de Sardan, “A Moral Economy of Corruption”, Journal of Modern African Studies, 37 (1).  See also Giorgio Blundo & Jean-Pierre Olivier de Sardan (dir.), État et corruption en Afrique (Paris: Karthala, 2007).

12 Sardan summarises the situation as follows: “Nous avons été nous-mêmes, à notre grande stupéfaction, accusés de ‘culturalisme’ (Dahou 2002) à propos d’une tentative antérieure de dégager quelques ‘logiques culturelles’ jouant un rôle de ‘facilitateur’ dans les légitimations des pratiques corruptives (Olivier de Sardan, 1999).  Dans cet article sur l’économie morale de la corruption, nous avions pourtant pris soin d’utiliser le terme de logiques culturelles en précisant: ‘All these logics are syncretic, none is ‘traditional’, none is coming directly from a so-called pre-colonial culture’ (ibid., p. 44).  Le paradoxe, dans le cas du procès que nous a fait Dahou, est que sa propre interprétation de la corruption est elle-même profondément culturaliste (Blundo 2007: 46-47): il impute la banalisation de la corruption dans le delta du fleuve Sénégal à la logique traditionnelle dans le Walo!”.

13 Tarik Dahou, “Déculturaliser la corruption”, Les Temps Modernes, 620-621, pp. 289-311.

14 John Iliffe, Honour in African History (Cambridge: Cambridge University Press, 2004).

15 See here particularly, Jean-François Bayart, L’État en Afrique (Paris: Fayard, 1989).

16 Those who are interested can turn to Culture Troubles, which provides a systematic discussion of what such an approach might entail for comparative politics. I take up the question of the limits of Western social sciences in my next book: Patrick Chabal, The End of Conceit: Western rationality after postcolonialism (London: Zed, forthcoming).

17 See Culture Troubles, Chapter 7.

Haut de page

Pour citer cet article

Référence électronique

Patrick Chabal, « “Who Speaks for Africa?” », Cahiers d’études africaines [En ligne], 204 | 2011, mis en ligne le 06 janvier 2014, consulté le 22 mars 2017. URL :

Haut de page


Patrick Chabal

Department of History, King’s College London.

Haut de page

Droits d'auteur

© Cahiers d’Études africaines

Haut de page