Integrating Ethnicity Into Political Democracy in Africa. Interview with Michel Cahen – 18 August 2024

Paul Martial – When discussing Africa, the notion of ethnicity is frequently mentioned, both in general media and among social science researchers. How would you define this concept?

Michel Cahen – In France during the 1930s, people began using the word ’ethnicity’ as a euphemism for ’race’. But even this word didn’t quite have the same meaning as it’s reduced to today. For example, Ernest Renan, the great 19th-century theorist of the French nation who wrote “What is a Nation?”, frequently spoke of the ’French race’. This meant community, or nation, but with the idea that culture was in the blood. It’s a kind of essentialism. The Communist International used to speak of ’Negro’, which we no longer do today.

In the 1930s, people began using the word ethnicity in an equally essentialist way. In French social sciences, the two foundational moments for the more modern discussion of the word and concept of ethnicity are Jean-Loup Amselle and Elikia M’Bokolo’s book “Au cœur de l’ethnie” (At the Heart of Ethnicity), first published in 1985, and then a few years later Jean-Pierre Chrétien and Gérard Prunier’s “Les ethnies ont une histoire” (Ethnicities Have a History), published in 1989. Their thesis is that ethnicity isn’t simply a manipulation by colonisers, that Africans didn’t wait for Europeans to feel identities – which we can call ethnicity or nation – that are social constructions and therefore fluctuating. There are centuries-old identities in Africa; I’m thinking, for example, of the Kongo nation, initially the product of a political construction. The Kongo kingdom existed for two centuries before the Portuguese arrived in the region in 1482, and the Kongo identity still exists today, although colonial borders have divided it into five pieces: western Congo Brazzaville, western ’democratic’ Congo, the far south of Gabon, the Cabinda enclave belonging to Angola, and that country’s two northern provinces which are actually called Congo and Zaire. It’s an entity that continues to exist, with its own language, its own rites, its king as a cultural dignitary – even if he has a minor role – and whose capital is Mbanza-Kongo in Angola.

Can you clarify the notion of ethnicity?

As a Marxist, what has always struck me is the difficulty in grasping the ethnic phenomenon. There was a similar debate in Europe about the national question. We remember that Engels supported independence for Ireland and Poland, while Rosa Luxemburg opposed Polish independence because she believed it would divide the proletariat that had been forcibly unified within the German Empire. Engels explained that for a proletarian to enter into struggle, they first needed to know what territory they were walking on, what belonged to them: for an Irish proletarian to ally with an English proletarian, the national question had to be resolved. Engels gave importance to Ireland and Poland because they were colonised by England and Germany respectively, two major industrial countries. He had what I would call a somewhat instrumental sympathy. And the same Engels had appalling things to say about the Balkan states, which were small and had peoples “without history”. As these weren’t industrial countries, he considered them backward. Yet from a materialist viewpoint, ethnicity is a subjective social formation that expresses societies according to their own identity trajectories and must be integrated into our thinking.

Let’s take the case of two very different countries, the Cape Verde islands and Mozambique. Mozambique is a country in the southern cone of Africa, bordering the Indian Ocean, whose population is part of the great Bantu family. Cape Verde is a Creole archipelago located in the Atlantic Ocean, 500 km west of Dakar, which was uninhabited when the Portuguese arrived. It was populated entirely by slaves brought from different parts of Africa, who didn’t share the same identities, religions, or languages, which is why they had to forge the Creole language. Thus a new identity emerged, the Creole identity, territorially circumscribed by the archipelago. We can say that historically there has been the formation of a nation quite comparable to our nations in Europe. There is no major problem of identification between the poorest Cape Verdean and the Cape Verdean state.

For Mozambique, it’s different, as we’re in continental Africa. The country was also colonised by the Portuguese at the very beginning of the 17th century, although most of the territory wasn’t occupied until the end of the 19th century. It’s Bantu Africa with its lineages, its traditional chiefdoms, its pre-colonial African nations, great states that were militarily defeated by the Portuguese. There were African identities, but they weren’t necessarily nation-states: the nkosi (king/chief) of one of the main political formations in southern Mozambique, the Gaza Empire, was a Zulu immigrant linked to the Mfecane (great movements of Zulu migrations from the end of the 18th century). It was an entirely slave-trading and violent state that partially “Zulu-ified” these populations. Even so, its population was far from homogeneous; it wasn’t a pre-colonial nation-state. But other political entities were related to much more homogeneous populations. Why not call them nations?

In the centre of the country there was a very different phenomenon, the ’prazos’. These were former Portuguese feudal estates that had become largely Africanised without reverting to traditionalism. Black or Goan chiefs owned lands in the name of the Portuguese king. These political structures were superimposed on previously existing identities. These entities were clans or lineages, sometimes very marked identities, in the north of the country, notably among the Makonde and Makua peoples. When the Portuguese occupied the entire territory at the end of the 19th century, they didn’t transform the population into “black Portuguese”; people naturally continued to be African. The Mozambique Liberation Front (Frelimo) [a national liberation movement that later became the ruling party of independent Mozambique], taking power in 1975 after ten years of armed struggle, refused to take into account the existence of pre-colonial African nations, considering them collectively as “tribalism”, and did not promote their culture and language. But neither did it succeed in being a social state for the 80% of the population that was rural, which could have led all these groups to identify with “Mozambique”, this new territorial space officially made a nation. Instead, “For the nation to live, the tribe must die” was Frelimo’s policy. The use of the word tribe was highly questionable, and this anti-ethnic policy had practical consequences, such as literacy campaigns conducted exclusively in Portuguese – with a gigantic failure rate – along with the banning of traditional chiefs, rain rituals, etc. It was, in my view, a kind of attempt at “Portugalisation” or “Lusophonisation” of the country with the idea of the New Man, borrowing somewhat Maoist jargon.

In France, something similar happened, with very strong ethnic repression; Napoleon, then Napoleon III, and especially the Third Republic Frenchified France: everyone remembers the signs “it is forbidden to spit and speak Basque or Breton during recreation”. But this French state that repressed ethnicities simultaneously created compulsory public schools, hospitals, roads, bridges, it brought progress, and thus there was a political identification with the French social state, which gradually became a national identification. This exchange between social progress and ethnic repression – I’m not saying it was good – could function.

The capitalist state of the periphery or ultra-periphery, as African states are, is not, with rare exceptions, a social state; it’s a neocolonial, kleptocratic state [1] that oppresses socially, economically but also ethnically, although it has ethno-clientelist practices. Ministers build the road that goes to their village by misappropriating their ministry’s budget, but this isn’t at all a policy of combining African identities to build a nation of nations.

If we take the case of Great Britain, it isn’t the federation of England, Wales, Scotland and Northern Ireland. It’s not a federal state; there’s a British supra-identity. It’s a singular identity of plural identities. A Scot can accept being British, but won’t appreciate being confused with an Englishman. African states haven’t followed this model of nested identities, and they have opposed pre-colonial African nations to the new nation that was meant to be a break rather than this coagulation bound together by a social state.

The system in Ethiopia takes into account different ethnicities, at least officially.

It’s almost an exception, and it hasn’t worked at all. In this country, you have identity-based federalism. In principle, each constituent nation has a territory with an autonomous province, but the country is ruled by a dictatorship and the autonomies have never been respected. What was good in the Constitution thus hasn’t been materialised.

I’m not saying federalism is needed everywhere within each African country. Internal federalism in African countries risks leading to the definition of mono-ethnic provinces. Yet in Africa, mono-ethnic provinces are very rare. In the Makua region, there are also Makondes, Yao, a majority people and minority peoples. The question isn’t to create federalism; we should rather look to Evo Morales’s Bolivia, which in 2009 proclaimed the constitution of the Plurinational Unitary State of Bolivia.

The ethnic fact in Africa isn’t an enemy for us Marxists. It’s simply something that exists in society, which we must be careful not to essentialise. These are identities that may or may not become nations, but sometimes social discontent will express itself along ethnic alignments. Generally, there is never a civil war whose characteristic would be solely inter-ethnic. For example, in the case of Rwanda, the Hutus and Tutsis are not two ethnicities. If we must give them a name, they are rather castes, two groups having the same language, the same origin myths, the same kingdom, but some were considered professionally as farmers and others as herders. Everyone will agree that colonial manipulation bore fruit, but it wasn’t an ethnic war.

When there’s an ethnic conflict, it’s often because there are social problems. Currently, in northern Mozambique, there’s a jihadist guerrilla movement. A group that previously existed as a religious sect has militarised and then affiliated itself with the Islamic State. It recruits among the coastal Mwani group, and among the Makuas – a large group that was quite badly treated by the Portuguese colonisers then by Frelimo. Finally, there’s the Makonde group at the Tanzanian border. That’s where the liberation war began in 1964. The Makonde group, although a minority in the region, was extremely important in the anticolonial liberation war [2]. As its members were major actors in the liberation war, they monopolised important leadership roles. From generals in the guerrilla movement, they became ministers. Although very clearly a minority in the country (2% of the population, and about 10% in the northern Cabo Delgado province), they have monopolised most positions that allow one to become wealthy. Today, there’s an ethnic expression of social discontent against the Makondes from the Mwanis or Makuas, but this is due to the inequality caused by state power being monopolised by a particular ethnicity due to historical circumstances.

The difficulty in Africa is that we are on the periphery of capitalism. The States are not social States, but prebendary States, comprador States, States that manipulate ethnic clientelisms, which often promote a single ethnicity. In Senegal at the moment, there is an accentuated ‘ouolofisation’ and the other African languages are in decline and could disappear in the future. This has led to endemic guerrilla warfare in Casamance [3] and little by little there could be other revolts (not necessarily in the same form), especially if development remains very uneven across the country. Behind all this, there are always material and social conditions, and it’s not economism to say this: identity never comes on its own, it’s the expression of positions taken in the face of changes that are felt to be aggressive or worrying.

I’ve seen this first-hand in Mozambique: during the colonial era – that is, until 1975 – anthropologists were able to identify a large area in the north of Mozambique where people spoke a language family called Makhuwa-lómwè. After Frelimo came to power, its policies mainly benefited the southerners, the capital and the big cities. The people felt oppressed by this authoritarian modernising state and the rebellion supported by South Africa was to gain a lot of weight in these areas. People started calling themselves Makuas in response. They really felt it, and it doesn’t help to say that it was a ‘false consciousness’.

How can both ethnic and social problems be addressed when ethnic issues are completely manipulated and overshadow all social subjects? Certain Nigerian Trotskyist groups seem to underestimate the ethnicity question in my view.

There are political entrepreneurs who openly manipulate identity, and not necessarily ethnic identities. They can manipulate black identities in a country where there are mixed-race people. You can manipulate anything if that thing exists. However, it’s clear that social problems have ethnic effects. I gave the example earlier of northern Mozambique, where the jihadist guerrilla movement has no trouble recruiting young men against Frelimo power. Even though this naturally doesn’t concern the entire Makonde population, they, although a minority in the province, have better access to state revenue. This question of socio-economic inequality then expresses itself ethnically: the Mwanis say “we have nothing, the Makondes eat everything”. This isn’t ethnic manipulation; it’s the ethnic expression of social inequality.

This reminds me of the famous debate that Trotsky had with his rare supporters, already expelled from the Communist Party, in South Africa. The already Stalinised CP and Third International defended the slogan of Black Republic and the young South African Trotskyists were for a colourless republic, if not red. This was quite abstract universalism on their part because the rule of the majority meant a black republic. It didn’t mean that Whites had to leave, but they must lose their white privilege, and Trotsky had defended the Black Republic slogan.

We Marxists must understand what the famous expression “The history of all hitherto existing society is the history of class struggles” means. It doesn’t mean there is only class struggle, only class consciousness. A person can be white, black or mixed-race, they can be male or female, they can be left-wing or right-wing, they can like cider or beer, they can prefer rugby to football, they can have lots of identities, and the only place where all identities mix is within themselves, in the individual, the only place indivisible without death. At a given moment, it’s not necessarily the class question that will be most important for moving this person to action: it might be being Muslim, because the mosque has been burned down by racists, that moves them to action, not as a proletarian of Muslim religion but as a Muslim person full stop.

These Nigerian comrades are making efforts to overcome divisions but it’s all the more abstract precisely because, in Nigeria’s history, the boundary between the Muslim zone and the animist zone – more Christianised because Christian missions only succeeded in non-Muslim territory – corresponds to the former Sokoto emirate, the great pre-colonial African state. These divisions weren’t invented by the coloniser; they are historically produced. Nigeria is an artificial construction like many postcolonial states but that doesn’t mean it can’t function: if it were a social state highly respectful of the different ethnic identities historically produced on the territory of present-day Nigeria. Nigeria is actually a federal state but this doesn’t in itself mean better social, economic and cultural respect for populations due to the existence of the peripheral capitalist state and the oil catastrophe.

The “revolution” of 1959 in Rwanda, where the absolute power of the Tutsi elite was overthrown, instead of taking a social trajectory, took an ethnicist trajectory with the dramatic consequences we know.

A social mobilisation can express itself through ethnic (rather castist in my opinion in this case) polarisation because these are the most available lines of understanding for people. This ethnic/castist feeling then takes on its own autonomy: even if the social problem from which it comes is resolved, the ethnic question doesn’t magically disappear. If an identity is massively felt, the social problem that nourished it may disappear, this may avoid massacres, but it won’t prevent the perpetuation of this identity over several generations and political democracy should take this into account.

The case of Somalia is interesting because it’s an ethnically homogeneous nation-state, but two regions are demanding independence, Somaliland and Puntland.

Indeed, it’s not just the ethnic question; in Somalia there are what are called “clans” which correspond to what we would call tribes in the Arab world, for example. These are political structures, generally not ethnic identities. But we shouldn’t have a static vision of ethnicity. People could feel Somali before and no longer feel Somali tomorrow. Somaliland claims independence, and has in fact obtained it. It’s a state that isn’t recognised by anyone but it’s the part of Somalia that functions best! There have even been elections that were monitored by international observers. Colonial domination also had identity effects. I’ll take up my example of Northern Mozambique with the Makondes, this group that was so important in the anti-colonial struggle and which monopolised positions of power. There are Makondes on both sides of the border: north of the Rovuma River, you’re in Tanzania and south in Mozambique. 120 years of English colonisation on one side, and Portuguese colonisation on the other, have had identity effects. Today, even if they recognise that they are cousins, the southern Makondes know very well that they are no longer quite identical to the northern Makondes.

In Somalia, ethnologists may well speak of a single country, but this won’t prevent internal contradictions that mean certain regions will demand independence. But this search for independence isn’t necessarily ethno-national; it can be motivated by the lack of functioning of the state, which isn’t democratic, which doesn’t bring social progress or which has been monopolised by one clan when there are a good fifteen of them, etc.

Somalia shows two things. Firstly, it’s not because you have an identity, an ethnic homogeneity, that everything will go well, because there are other problems. Secondly, identity changes according to trajectories that can cause disparities within the population. Identity is only a community of people who feel a certain way at a moment in their identity trajectory.

And what about tribes and clans?

We can perfectly well use the word tribe without colonial paternalism. A tribe is a political organisation of a fraction of the population, with a chieftaincy, delegate chiefs in different regions. They exist in the Arab-Berber world, in Somalia (under the name of clans).

In Mozambique for example, there are many ethnicities but there are no tribes because they were broken by the Portuguese coloniser. Unlike the English, the Portuguese practised direct administration, they didn’t reinstate powerful traditional chiefs who had been defeated, now docile and local managers of the European imperial state.

The clan is an imaginary organisation (at least in the territories I know). A certain category of the population, based on animal myths, says they descend from the turtle, or the monkey. We shouldn’t forget that the word “Bantu”, before designating a family of African civilisations, simply meant “human being” (as opposed to the animal kingdom). These mythical animal origins imply food taboos, for example not eating turtle if you descend from the turtle.

Lineages are the organisation of kinship – patrilineal if descent is through the father, and matrilineal through the mother. In the latter case, this doesn’t designate matriarchal power but a social organisation in which it’s not the woman’s husband who has power but the woman’s brother. The lineage is defined by the mother, a bit like in classical Judaism.

Perhaps a word in conclusion?

For us Marxists, it’s high time to think about integrating ethnicity into political democracy. Certainly, there aren’t only struggles for democracy, there are also social struggles, class struggles of course, but the latter need democracy and political democracy needs ethnicity to be integrated rather than fought against. It’s not about defending tradition, that’s not the question. If things are good in tradition, we defend them and, if things are bad in it, we fight them. But be careful not to designate false culprits: for example female circumcision doesn’t come from Islam, it existed long before. And we can only fight against this “tradition” with people, not against them.

Behind the right to identity, there’s the right to equality. I have the right to be Yoruba, to be Makua or other, I have the right for my children to be made literate in this language at school, for my province’s territory to be drawn according to where people who speak like me are in the majority, I have the right for the state to be locally bilingual. The state can be English-speaking, Swahili or Portuguese but there must be official bilingualism. Civil servants appointed don’t necessarily have to be of the locality’s ethnicity but must be able to speak its language for a public service that respects people.

For Marxists, I think this is a very important issue because of Africa’s socio-economic evolution. The latter is currently experiencing galloping urbanisation without proletarianisation. People who can no longer live in the countryside come to the city but generally can’t enter the capitalist mode of production. They can’t become workers, employees. To become civil servants, you need ethno-clientelist connections… These people then need, for their social survival, to preserve horizontal solidarity ties like ethnicity. It’s only later perhaps that they’ll feel vertical solidarity ties, that is class against class, proletariat against bourgeoisie. But the vast majority of the poor in Africa aren’t part of the proletariat.

Indeed, the proletariat is far from being the majority there (nor is it necessarily the most deprived social environment), compared to the urban plebs. The plebs isn’t a class, it’s an unstable social formation of people perfectly useless for capitalism since they barely represent a market [4]. They can die from AIDS, Covid or in a civil war, it’s not a problem for capitalism. But these are people that Marxists must defend. Often, the main question in Africa isn’t proletarian but plebeian and it’s not easy to define transitional demands for this kind of population. We don’t have a real political tradition to defend these people but we’ll have to invent it. Current political developments in West Africa, for example (the electoral tidal wave of PASTEF in the 2024 Senegalese elections, the “anti-French” coups in Mali, Burkina Faso, Niger with, initially, undeniable popular support, etc.) are the indirect expression of the plebeianisation of the population, which is moreover extremely young.

Footnotes

[1] A kleptocracy is a term designating a political system in which one or several people at the head of a country practice corruption on a very large scale, often with their close associates and family members.

[2] I personally don’t call it national but anticolonial liberation, since there wasn’t strictly speaking a pre-existing nation before the liberation war.

[3] Casamance, sometimes called casa-di-mansa (“the land of kings”), is a historical and natural region of Senegal, located in the south of the country and bordering the Casamance River.

[4] I don’t confuse the plebs and the so-called “informal sector”. The informal sector is a classification that covers a large part of the population whose economic activity is not “legalised” in a legal-juridical framework. This informal sector covers various social classes and formations (proletariat of small informal enterprises themselves, plebs, artisanal circles, small and medium traders…). By plebs I mean the mainly urban population that is no longer part of the domestic mode of production in the countryside but cannot integrate into the capitalist mode of production due to the peripheral nature of capitalism in these countries.

Michel Cahen is a historian of Portuguese colonisation and a political analyst of Portuguese-speaking Africa. His latest book is ‘Colonialité. Plaidoyer pour la précision d’un concept’, Paris, Karthala.

This is an English translation, by Adam Novak, of the short version of an interview originally conducted (in French) by Paul Martial, host of the ‘Afriques en lutte’ website, on 18 August 2024. This short version was originally published in Inprecor No. 723, November 2024.

Views: 1