Defining “Portuguese-speaking” Africa?

However, our study does not ignore cultural features and starts by posing this initial question: what is “Lusophony?” Included within “Lusophony” we find Portugal, Brazil, five independent African countries, and East Timor. “Lusophony” is, first, an inter-state organization of those eight countries, the Comunidade dos Paises de língua portuguesa (CPLP), set up in Lisbon on 17 July 1996.^{Footnote 10} This immediately poses the interesting problem of how to translate that into English, since the most obvious option - “Community of Portuguese-speaking countries” - suggests that everybody in that “community” actually speaks Portuguese, which is false. “Lusophony” cuts across the varying social status of the language in the different locales. Portuguese may be the mother tongue of a huge majority of the population, i.e. the identity language. That is the case in Portugal and Brazil, and also for certain specific social milieus in Africa (old Luanda Creole elites and, more and more often, the city-born sons of the rapidly burgeoning urban population borne by the waves of refugees engendered by the long civil war). Portuguese may also be a “second language” - not the traditional language or the mother tongue - but one whose mastery is generalized or is, at least, very common. This characterization applies to the Mozambican elite and, but differently, to Cape Verde and the São Tomé and Príncipe archipelagos, countries in which, although nobody speaks Portuguese, everybody speaks a Portuguese-based Creole which makes Portuguese itself easy to understand. Last but not least, Portuguese may be a completely foreign language, as in the hinterland of Mozambique and Angola, for the majority of the Bissau-Guinean and East Timorese population, and for the whole of the population of Goa state. In this third case, Portuguese is not an identity language at all, but merely a tool - understood to a greater or lesser degree - by small minorities.^{Footnote 11}

The social status of the language is closely bound up with its social usefulness: why, after “five hundred years of colonization,” does nobody in Goa speak Portuguese? Probably because, for the Braham Konkani elite in Goa, speaking Portuguese was not an essential social or professional tool - many of them having emigrated since the nineteenth century to Bombay and elsewhere in the British empire (in particular East Africa), where they were appreciated as ship stewards and for their mastery of the Latin alphabet. On the contrary, although the Portuguese remained only sixty years in Malacca (Malaysia) during the sixteenth and seventeenth century, there is a “Portuguese settlement” (district) in which a Portuguese-based Creole (called kristang) is still spoken by old people as an identity language for Christian fishermen in this highly Islamized country.^{Footnote 12} Doubtless, in East Timor, the combination of Catholicism and Portuguese language as tools of resistance against Indonesian invasion was important in increasing the number of more-or-less Lusophone people in the country (circa one fourth).^{Footnote 13} We see, therefore, a great variety of situations, but which share one common feature: all the states which are members of CPLP speak (or, better, write in) Portuguese.

That is why I believe that the best definition given for “Lusophony” was that of Eduardo Lourenço, a Portuguese philosopher and writer. For him, Lusophony is not an “area where Portuguese is spoken,” but a “peculiar area of intersection with other identities” (European, Indian, Bantu, Muslim, Christian, Jewish, etc.).^{Footnote 14} What is important, whatever the social status of the language, is the presence of a certain “weight” of Portuguese expansion. Certain social scientists remain dissatisfied with retaining the concept of “Lusophony,” even with such a definition, since in some areas nobody speaks Portuguese, even as a second or foreign language. Such scholars prefer the more restrictive concept of “Lusotopy,”^{Footnote 15} a term chosen in order to express the topoi (places) where Portuguese expansion has exerted an influence, even if the Portuguese language has completely disappeared - as, for example, in the Agudas community of Ghana, Benin, Togo and Nigeria, or the Goanese. But, whatever the term, Lusophony or Lusotopy, we should bear in mind that what characterizes the areas denoted by the concept is not a “common feature,” but a peculiarity of intersection with other identities or features: we will, therefore, combine the Whitehead and Lourenço approaches.

Here, then, we find a first major difference between “Luso-Latin” Africa and Latin America, even Brazil. In Latin America, although some Indians peoples do not have Spanish or Portuguese as their mother tongue, or do not speak those neo-Latin languages, most of the population undoubtedly has those languages as identity tongues, and those tongues are founding features for those countries. That is not - at least, not yet - the case in former Portuguese Africa.

Geographical Continent versus Political Archipelago

A second key difference is explicit in the observation that whilst Latin America is a continent, Portuguese-speaking Africa is formed of separate pieces within a continent. It must be admitted that the fact of being a continent does not prevent Latin America from being heterogeneous.^{Footnote 16} But that is even more true in the case of Portuguese-speaking Africa. That set of countries may be divided in several different ways: geographically, with west African countries (Cape Verde, Guinea-Bissau), central Africa (São Tomé and Príncipe) and southern Africa (Angola and Mozambique); from the social formation viewpoint, with Creole countries (Cape Verde and São Tomé)^{Footnote 17} and continental African countries (Guinea-Bissau, Angola, Mozambique); culturally, by grouping countries within the west African (Cape Verde, Guinea-Bissau, São Tomé) and Bantu areas (Angola and Mozambique). Other modes of classification could also prove plausible (by main religious areas, by size,^{Footnote 18} for example).

In spite of that substantial heterogeneity, “Os Cinco” (“The Five”) maintained a separate inter-state organization (Cimeira dos Cinco/The Five Summits) throughout the whole “radical” period (1975-1988/90). It was not the Portuguese language per se which held the Five together, but the political situation produced by a common enemy and late decolonization. Indeed, the Cimeira dos Cinco stemmed from the colonial time CONCP (Conferênça das Organizações nacionalistas das Colónias portuguesas/Conference of the Nationalist Organizations in Portuguese Colonies). CONCP (with its headquarters first in Rabat and then in Alger) was a founding mould for a political generation of movements which, though claiming to be either “Marxist-Leninist” (Angola, Mozambique) or adepts of “revolutionary democracy” (Cape Verde, Guinea-Bissau, São Tomé), established very similar single party systems. The war against apartheid South Africa maintained that political solidarity and identity, but the creation of the CPLP in 1996 weakened that “solely African” organization.^{Footnote 19} Despite being weakened, however, the organization has not completely disappeared and meets at summits from time to time.^{Footnote 20} But it was not just the creation of the CPLP; a tendency toward the weakening of any political identity of the “Five” was doubtless precipitated by their membership in the African Union, the integration of Cape Verde and Guinea-Bissau in ECOWAS (Economic Community of West African States), of Angola and Mozambique in SADC (Southern African Development Community), and of Cape Verde, Guinea-Bissau and São Tomé within Francophony and Mozambique in the Commonwealth. We will see, nevertheless, that the colonial legacy is still an important common factor, as regards social relevance.

Myth and Reality

It could be said that everything argued here about the status of the languages might equally be said about “Anglophony,” “Hispanophony,” and “Francophony,” and it would be true. It is not at all a question of “-phony,” but has to do with historically constructed social formation. One important specificity is to be found here in the case of Portugal: the longevity of its colonial presence. That Portuguese longevity is found in “Latin America” as well as in Africa (even if differently), which is not the case for Spain. But we should be more precise, since that longevity is both myth and reality.

In order to explain the “exceptionality” of its African empire, the authoritarian Portuguese regime (1926-1974) disseminated the idea that it was a peculiar case, since the Portuguese had been there for 450 years, arriving even before some African peoples whose migration took place only later. Portugal did not colonize for economic reasons, but to advance the cause of Christianity and civilization, inter alia.^{Footnote 21} This was bolstered by claims about miscegenation and the capacity of the Portuguese to integrate themselves in the Tropics - completely disregarding the slavery context of pre-capitalist miscegenation - and confusing the Brazilian and African situations.^{Footnote 22} The power of this theory resided in the fact that it was not only a “fascist” but also a national Portuguese concept, one which was fully endorsed within even the most strongly opposed of republican circles. Portugal had never been racist, and the Third Empire in Africa (after the first one in Asia and the second in Brazil) simply pursued the path of the Discoveries: Portugal had given “New Worlds to the World” and, thanks to its Empire, was regaining its lost importance. That kind of feeling is still very much alive in Portugal today, with Lusophony being confused with merely diluted Lusitanity.^{Footnote 23}

In fact, the places in Africa where Portugal could claim centuries of colonial presence were somewhat limited: the island of Santiago in the Cape Verde archipelago, Cacheu in present-day Guinea-Bissau, São Tomé and Príncipe, which were first populated with exiled Jewish children, Luanda, and a few other places in what is now called Angola or the Congos, the Islands of Moçambique^{Footnote 24} and Ibo, the peninsula of Mossuril, as well as Quelimane and Tete in Mozambique.^{Footnote 25} Other old sites, like Mombasa in today’s Kenya, disappeared.^{Footnote 26} This means that 95% of what is today Portuguese-speaking Africa was conquered between 1895 and 1914, at more or less the same time as the other European states were engaged in their scramble for Africa. Throughout the whole breadth of Portuguese-speaking Africa, Portuguese colonization is no older than that of the other European imperialist conquests.

One very important point must be stressed: from that historical point of view, it seems to me incorrect to speak about the “pre-colonial” period before the effective conquest of the land by Portugal (and other European countries) at the end of the nineteenth century. That period was certainly one of a new age of colonization, the turn to colonial capitalism, which imposed the re-organization of the African land in order to be able to develop the direct exploitation of African manpower. But capitalism should not be reduced to the capitalist mode of production. Most of the capitalist process was characterized by non-capitalist forms of capitalist exploitation (such as slavery, or unequal or forced trade).^{Footnote 27} The first age of colonization did not require the whole land to be conquered, but managed to prove quite efficient in amassing profits, thanks in part to the slave trade, and that age was particularly important in the Portuguese areas. Huge caravans of pombeiros penetrated far into the hinterland of Angola, Zambia and Moçambique.^{Footnote 28} The same phenomenon happened with the lançados in Guinea-Bissau^{Footnote 29} - to be compared with the bandeirantes of seventeenth-eighteenth century Brazil. This first age of colonization profoundly marked local societies. In Mozambique, the specific phenomenon of the prazos - a kind of Portuguese, although gradually becoming blacker and blacker, “Luso-Africanized” senhorio (a kind of seigneurality) in Africa - allowed the indirect but long-term rule of “something Portuguese” in vast areas.^{Footnote 30} It is worth noting that those Portuguese pombeiros, lançados or prazeiros were not necessarily white, and might be mixed-race or completely black, but were still “obviously” Portuguese to the African population all around them (significantly known in Guinea as Kriston - ‘Christian’–, which was a synonym).^{Footnote 31} Additionally, in a few places of permanent Portuguese presence (already mentioned, like Ilha de Moçambique, Tete, Luanda), the so-called Terras firmes (dry land), there were actually several centuries of colonization.^{Footnote 32} Although very tiny land areas, they were of considerable social importance and historical projection, producing specific social milieus, non-capitalist but truly colonial milieus, white, black, or of differently mixed-races (African-Portuguese, Swahili-Portuguese, Indian-Portuguese, Indian-African-Portuguese) which were nonetheless undoubted vectors, or at least brokers, of Portuguese presence during first age colonization.^{Footnote 33}

Rupture between the Ages of Colonization

Those milieus were heavily involved in the slave trade until the middle of the nineteenth century in Angola^{Footnote 34} and even later in Mozambique,^{Footnote 35} and that was probably the main factor preventing them from adopting the turn to capitalism and not producing “New Brazils.”^{Footnote 36} Being exclusively (or overwhelmingly) slave exporters and not fazendeiros (slave estate owners), their situation was completely different from that of the Brazilian, Caribbean, Cuban or Southern United States estates: there was no plantation complex in continental Portuguese Africa,^{Footnote 37} and they would have had to create such fazendas when slave trading became impossible, and could not simply adapt their slave manpower to a more or less wage-labour manpower in the already existing fazendas. Culturally and “capitalistically,” they were incapable of producing such a turn.^{Footnote 38} Their strategy was simply to try to continue the slave trade in clandestine form increasingly repressed by the British Navy and Portuguese authorities.^{Footnote 39} Those fazendas were to emerge with the introduction of forced labour (not to be confused with slavery), installed later, in the twentieth century, by metropolitan or foreign capitalist companies. “Official Portugal” had to wage war against those first-age Portuguese colonizers (in particular against the prazeiros), but waging war against an early Portuguese social milieu produces a completely different result than doing so against an African state. Very often - but not always - when an African state had been militarily defeated, its elite disappeared. The Portuguese or Afro-Portuguese elites of the first age of colonialism might be economically opposed to the Portuguese state, and wage wars to indicate their will to preserve the first age of colonization against the new colonial capitalism, but they nonetheless continued to belong to the colonial society. That society was certainly highly fragmented, but they remained within it, not being pushed back towards the colonized society. By various kinds of family links, it was always possible to reintegrate the new colonial world, albeit in a position of subalternity. Therefore, although marginalized, those elites generally survived. That fact is capital and can still be observed even to the present day.

Certainly, the situations in Angola and Mozambique, or in Guinea-Bissau, are different. In Angola, Luanda has never ceased to be the capital city of the country. That fact allowed the long-established old Creole families to resist the “whitening” of the colony by metropolitan immigration in the twentieth century, especially after the Second World War, and to secure certain good social positions, though their increasing difficulty to do so became the basis of their frequent turn to anticolonialism.^{Footnote 40} Similar cases can be found in Malange (in the hinterland of Luanda) or in such other old colonial cities as Benguela. So, great mixed-race Creole families first developed an assimilationist anticolonialism against discrimination, asking - from the 1920s to the end of the 1940s - for equal treatment as “Portuguese” citizens, since they were “Angolans,” i.e. Portuguese from Angola and not indígenas (approximately “natives”). The next generation, reinforced by a tiny minority of Angolan-born Whites, turned to separatist anticolonialism, and that generation proved crucial in the process of formation of the MPLA (Movimento Popular de Libertação de Angola/Angolan People Liberation Movement). The historical division of the Angolan anti-colonial movement relies, at least in part, on that specific process: one of the main movements is not the product of “traditional” African society, or of more recent modern social milieus,^{Footnote 41} but of a specific old urban colonial history and of its closed hinterland (mainly of Mbundu ethnicity). Undoubtedly, things are never simple, and the MPLA was able to open up to a variety of other social and ethnic milieus and - to be clear - I am not at all suggesting that the MPLA was not a “truly African” movement. I simply suggest that although that process occurred within African society, that society itself was highly fragmented by endogenous and colonial factors. The other two main movements, FNLA (Frente Nacional de Libertação de Angola/Angolan National Liberation Front) and Unita (União Nacional para a Independência Total de Angola/National Union for the Total Independence of Angola) emerged from socially different kinds of trajectories. The first movement emanated from a branch of the Kongo royal family, with its support stemming mainly from the Bakongo people, and its social milieus, urbanized in Kinshasa after the emigration waves set in motion by coffee plantation expansion in the 1950s, and the subsequent dispossession of African land by the Portuguese fazendeiros. ^{Footnote 42} The latter came from a split in the FNLA, led by some Umbundu (South-Central Angola) and Cabinda leaders.^{Footnote 43} It should be added that if Catholicism was the predominant religion within all three leaderships, the forms of Protestantism differed (more Evangelist for MPLA, Baptist for FNLA, and Congregationalist for Unita). Socially, ethnically, religiously, linguistically, those trajectories had very few features in common within an area that was not a nation. They felt that the two others were aliens, and hence illegitimate to rule Angola.^{Footnote 44}

In Mozambique, the turn to colonial capitalism resulted in different features. “Old Mozambique,” which dated from the first age of colonization, was located in the north-eastern part of today’s country, from the river Zambeze (Chinde, Tete, Zumbo) and lower Zambezia (Quelimane) to the former capital city, Moçambique Island, and as far as Ibo Island in the far North-East, all areas globally integrated into the Swahili world and the economic sphere of the Indian Ocean. At the end of the nineteenth century, Portugal decided to “plug” its colony into the far more developed capitalist economy of South Africa, transferring the capital city to the far South, at Lourenço Marques (present-day Maputo), which was barely even a small town at the time. That allowed Portuguese sovereignty on Delagoa Bay to be secured and money made from the railway and the port receiving South African coal. The two main ports of colonial Mozambique (Lourenço Marques and Beira, in the south-centre part of the country) were the pure product of the new age of colonization, without any prior urban tradition. The old elite nucleus of the northern part of the colony, completely marginalized by the geopolitical aspect of the economic turn to capitalism, never regained its importance and nourished a variety of discontents, from assimilationist anticolonialism - in particular in Zambezia province from the 1920s to the end of 1950s^{Footnote 45} - to support for the anti-Frelimo rebellion from the end of the 1970s to 1992.^{Footnote 46} Consequently, Frelimo (Frente de libertação de Moçambique/Mozambique Liberation Front) has not undergone the same history as the MPLA: Frelimo, it could be said, is the product of what Messiant,^{Footnote 47} talking about Luanda, called the “new assimilated” social milieus of Lourenço Marques, with very weak participation from the old and pre-capitalist colonial milieus. That does not mean that Renamo (Resistência Nacional de Moçambique/National Resistance of Mozambique), the rebellion guerrillas sustained by Rhodesia and apartheid South Africa against “Marxist-Leninist” Frelimo, was the natural expression of this old elite nucleus; but doubtless many of its members were sympathizers, or even gave direct support to that movement.

What is very important to note is that the same social (and ethnic) milieus which had been marginalized by colonial Portugal during the turn to capitalism were again marginalized by the new independent modern African state. Only one part of the old elite succeeded, very often by migrating towards the South (first from Moçambique Island or Quelimane to Beira, and then to Maputo), in “entering” the main elite, the “nation,” and thence disappearing as a specific social milieu. One major aspect of present-day Mozambique’s regional imbalance remains linked to that history: “old Mozambique” areas have the worst indices of human development, and the old Zambezian elite in Quelimane or the Afro-Luso-Indians in Moçambique Island remain convinced that Mozambique was stolen by southern foreigners, since their regions were the first to be colonized and, therefore, the most civilized.^{Footnote 48}

In Guinea-Bissau, the situation is somewhat different, as the Cape Verdean factor is historically important: the colonial administration used to employ Cape Verdeans at the low and mid levels of state colonial administration from the first age of colonization right through to the new one.^{Footnote 49} Guinean (not Cape Verdean) mixed-race people, all mulattoes, whether of Guinean or Cape Verde origin, were (and are) designated as “Burmedjus” (i.e. “vermelhos” in standard Portuguese, the “red colour skin” persons). This social distinction was exploited by the Portuguese against the PAIGC (Partido africano da independência de Guiné e Cabo Verde/African Party for the independence of Guinea and Cape Verde) whose main leaders were Cape Verdeans, whilst the guerrillas were black Africans. Anti-Burmedju demagogy was reactivated too at the time of the coup d’état of Nino Vieira in 1979, which put an end to unity with Cape Verde, causing the first president, Luís Cabral (brother of PAIGC founder, Amílcar Cabral and Cape Verdean) to flee the country.^{Footnote 50} Clearly, the problem ran deeper than that of mere opposition between black Africans and Burmedjus. But behind the demagogy against the Burmedju social milieu was the huge difference between an African society, with its ethnicities, lineages, clans, age classes and so on, and a completely colonially-produced Creole society in Cape Verde, easily comparable to Caribbean social formations.

In São Tomé, too, the scenario is different. Even if practically the whole of the black population in the nineteenth century was of slave origin, the decline of the sugar cycle within the small archipelago meant that the local Africans were able to leave the estates and were increasingly able to avoid working on agricultural tasks. They were thus able to go to the towns, where they could mix with the also-declining mixed-race Creole milieu, thereby shaping a new and broader Creole social group, the Filhos da Terra, or Forros (“Sons of the land”),^{Footnote 51} whose non-agricultural social life became a key identity matter.^{Footnote 52} Furthermore, when native laws were passed between 1878 and 1926, Forros were never considered as indígenas, but as a kind of assimilated people (like the Cape Verdeans). When the cocoa cycle began and “exploded,” the Portuguese administration was unable to put those Forros back to work again at the roças (or - only rarely - as foremen) and used large-scale slave imports from Angola and, later, forced-labour manpower from Angola and Mozambique.^{Footnote 53} Cape Verdeans were imported too, for foreman tasks, thanks to the harsh famines of the 1940s. The result was that when MLSTP (Movimento para a Libertação de São Tomé e Príncipe/Liberation Movement of São Tomé and Príncipe) came to power in 1975, if the roças had been given to the people working in them, the main wealth of the country would have gone to “foreigners” (even if they were often of the second or third generation). MLSTP then nationalized the roças, which was explained in a more or less Marxist and anti-colonial narrative, but may be interpreted as a completely conservative measure in order not to give the roças to their workers and to maintain, via the state, the command of the Forros over the land they did not work. Whatever the case, this led to a complete decline of the roças, with the workers being easily able to compare the late period of colonialism (when their situation had improved) with the new and harsh situation of independence, and therefore trying more frequently to escape the roças and to develop their own micro-farming concerns.^{Footnote 54}

Last but not least, the situation in Cape Verde itself has taken on completely separate characteristics. So far, I have spoken of creolity in terms of social milieus, more or less synonymous with a non-white elite.^{Footnote 55} But, in Cape Verde, the peculiarity of the long decline of slavery and the departure of most of the white Portuguese towards Brazil or Portugal produced a situation in which creolity is not a social milieu but the whole nation. That does not mean that there is no distinction between the “lightest” and the “darkest” Cape Verdeans,^{Footnote 56} but that, in that case, the distinction does not produce a contrast between Creoles and non-Creoles.

As we have seen, the situations in each of the Portuguese-speaking African countries are fairly or even completely different. But there is one fundamental common point: however diversified the consequences, we always find what I have called the (more or less) “Creole issue,” i.e. today’s presence of colonial social milieus produced by, or stemming from, the first age of colonization. We have here an illustration of what Fernand Braudel called “le temps long dans le temps court” (“the long term within the short one”). That is the main aspect of a specific combination of factors evoked at the beginning of the paper, and a peculiar legacy of the long-declining Portuguese first age of colonization.

At this point we can return to Latin America and, more broadly, to the decolonization process.

Decolonization?

Would the main difference between Latin-American states and the African ones really lie in the fact that the first gained independence hundred and fifty years before the second? That is an obvious factor of differentiation, but one which has probably been overestimated. Moreover, between the first independences - the earliest was that of Paraguay in 1811, and the latest that of Cuba in 1902 - it is not obvious that the “date factor” is the best way of understanding the difference between the new states and nations. And that is without mentioning the slave revolution, which gave atypical independence to Haiti (1804). On the other hand, the early independence of African countries - Liberia in 1847, and Ethiopia (never completely colonized except for a short while by Italians, 1935-1941) - does not mean per se that they are in a better situation. It seems that the process by which a colony becomes a state is far more important.

Former colonies have experienced three kinds of development in relation to their former “mother countries:” (1) independence without decolonization, (2) decolonization without independence, and (3) independence with decolonization.

Independence without Decolonization

The first category is most widespread in the former Spanish and Portuguese Americas. Independence was a political rupture, but in no way an anticolonial movement. The “anti-peninsularism,” which gained force within the settler class and other colonial milieus at the time of the Napoleonic invasion of the Iberian peninsula and led to their questioning their allegiance to the crown, should not be confused with anti-colonialism. It is true that the revolts were not led by the settler class alone, and also included the colonial society as a whole, with all its social milieus, including black slaves: but all these, however, were in fact the social milieus of the colonial society, not of the colonized society.^{Footnote 57} In such a case, the new country is independent but remains a colony, an auto-centred colony. That is what Aníbal Quijano calls the coloniality of power, which seems quite different from the mainstream postcolonial approach to Latin-American matters.^{Footnote 58} I use such a concept slightly differently than he does.

Anibal Quijano defends the concept of the coloniality of power of the independent states, but does not characterize as colonial states the independent Latin-American countries, even if he (and Wallerstein too) equally argues that independence did not destroy the colonial situation: “(...) even when formal colonial status would end, coloniality would not. It continues in the form of a socio-cultural hierarchy. (...) It was decolonization that fixed the stateness of the decolonized states.”^{Footnote 59} That is quite puzzling, even on the basis of their own statements: how can a state which has remained totally within the socio-cultural colonial situation be decolonized? That would make “decolonization” merely synonymous with a political split from the former empire. Moreover, one of the supporters of Quijano’s thesis, Ramón Grosfoguel,^{Footnote 60} has always criticized the idea that decolonization actually took place. But there is another somewhat huge problem with the latter, because of his obvious Latinocentrism: he seems to speak in general terms, that is to say including African and Asian decolonizations, when decolonization actually took place in those countries, even if their policies were neo-colonial. Other radical postcolonial social scientists, like Sanjinés C.,^{Footnote 61} speak too of the need for decolonization, and one therefore does not understand why they do not characterize the independent Latin-American state as a colonial state, only defending the thesis of the ‘internal colonialism’ of the independent state.^{Footnote 62} Here I speak about the coloniality of power, but also about the coloniality of the state itself, of the nation, of the very country and its delimitation.

Before independence, there was an imperial state apparatus (Spanish vice-royalties and colonies, or Portuguese Empire) but no state of the colonies, no colonial state. A state, even when completely neo-colonial, always has the function of representing its own land (if not the whole population of that land), and not being the representative of a foreign power in that land. The foreign power is a state quite clearly, the imperial (and colonial) state, but its state is that of its mother-land, not of the colony. This European imperial state develops what I call estates within the colonies, but in my view it is completely wrong to consider that what existed in a former imperial time (foreign state estates in the colonies) was a locally based colonial state, the state of the colony. We will return infra to the African case, but for the American (north and south) one, I consider that colonial states were born with their very independence and have, at least in part, remained as such.

This process occurred at its most extreme in Portuguese America, as it was the metropolitan state which decided to stay on in its colony, finding refuge there during the Napoleonic wars, but afterwards refusing to return to its mother country. The “Grito de Ipiranga”^{Footnote 63} never intended for the state to stop being Portuguese, but from then on it would build the Império not from Lisbon but from Rio.^{Footnote 64} What once had been a colony of Portugal, was to become a Portuguese auto-centred colony. Indeed, the very birth of Brazil in 1889 (the advent of the First Republic) did not mean at the time that the country was no longer a colony, but that it was no longer a Portuguese auto-centred colony – it had become a Brazilian colony. The nation formation process had slowly changed the imagining of the nation, from a mere split within the Portuguese nation itself to two nations, but had not changed its social colonial structure. Indeed, the abolition of slavery by Princess Isabel in 1888 was one direct cause of the proclamation of the Republic in 1889, a very conservative act indeed.^{Footnote 65}

It must be stressed that this characterization of new Latin-American states as colonial states applies not only to the ruling (settler/Creole)^{Footnote 66} class, but to the whole of colonial society. I am not speaking restrictively of settler states which typify one possible form of colonial state, the Creole oligarchic state. For example, one may concede that while the Bolivarian or Brazilian independences had a strong elite aspect, the Cuban revolution of 1898-1901 was a popular revolt, by no means restricted to an elite uprising. US military intervention against Spain did not change the nature of that social event. It was the colonial society (not just the colonial ruling class) which took power, expelling the Spanish state, within a framework shaped entirely by colonial history after the disappearance of the Indian population. So, when Brazilian colleagues say “We were colonised by Portugal,” I answer, “Begging your pardon, no, you were the colonisers for Portugal,” because most of today’s Brazilian population comes from colonial society. ^{Footnote 67} In Latin America, to a greater (Brazil, Paraguay, Argentina, Chile, most of Mexico) or lesser (Peru, Bolivia, Ecuador, part of Mexico) extent, it was the coloniser who fashioned the society, the state, the nation, the country. The very name of “Latin America,” which is not at all a peaceful one, signifies the maintenance of a colonial structure (Mignolo Reference Mignolo2005).^{Footnote 68} That does not mean that Laurence Whitehead is wrong when he writes that “‘Latin America’ exists as much in the minds of outside observers as it does in the minds of Latin Americans themselves - both among elites and within popular culture”^{Footnote 69} - firstly because obviously a lot of Latin-Americans, including those of low class, are of Hispanic ancestry, and secondly because of the well-known phenomenon of cultural hegemony. Nevertheless, one could question such a statement for the regions where indigenous populations have remained the majority.^{Footnote 70} To this objection, it might again be counter-objected that choosing the name of “Latin-America” has its own history and that it was not a given designation from the very beginning. But, firstly, one can remark that the Bolivarian revolutionaries never used the words “decolonization” or “national liberation” to designate their split with Spain, not to mention imperial Brazil. Secondly, if there were a lot of hesitations when deciding how to name the region (such as “South-America,” “Latin-America,” “Ibero-America,” “Hispanic-America,”^{Footnote 71} and “Spanish America”) there were, within the intelligentsia and statesmen, two designations which were resolutely opposed, except within a tiny minority, the young Víctor Raúl Haya de la Torre’s idea of “Indoamérica”^{Footnote 72} and José Carlos Mariátegui’s “América Indo-Ibérica,” two Peruvians living, it should be noted, in a country with a very big Indian population.^{Footnote 73} But, as Leslie Bethel points out, “whether the preferred expression was ‘Indoamérica,’ ‘América Indo-Ibérica’ or ‘América Latina,’ Brazil remained, for the most part, excluded.”^{Footnote 74} That exclusion may have a twofold origin: firstly, republican opposition to including an empire within the modernity of Latin-America but, secondly, the wish to be culturally Spanish in spite of the political split. Bethel shows how the history of Spain continued to be exalted throughout the nineteenth century by the new nation builders. More generally, with the few exceptions quoted, the nation builders need to be “Latin,” both for the legitimacy of their power, as for the very imagining of their country (not that of the Indians ).

The question is not only to accept that those states developed an “internal colonialism,” as Mariategui and, later, certain social scientists have denounced.^{Footnote 75} There are at least two main reasons for that. The first is that part of the American states (North and South) were colonialist, not only “internally” but equally “externally,” just like imperial Russia and western European countries, expanding their territories, at more or less the same period, that of imperialist expansion. Therefore, they were not only “colonial states” in the sense of states of auto-centred colonies born with their independences (as mentioned above), they were also “colonial states” just as European metropolitan and imperial states were. The fact that their conquests were, most of the time, in territorial continuity with the original land does not alter the fact they were colonial conquests. That is clear in the case of United States of America, which conquered the “West” by exterminating the Indians, and half of Mexico, not to speak about Panama, Philippines and Guam Island. But it has not been sufficiently highlighted that Latin-American states continued after their independences to occupy Indian lands they had never previously dominated: the best example could be the most recent one, that of the Chaco War, 1932-1935, which was the occupation war for the last free Indian lands in America.^{Footnote 76}

The second reason is that “internal colonialism” may describe the policy of a government or even of the state over a longer term, but not the socially-rooted imagining of the very nation, of the country, which is “Latin” and not Indian. However, if we characterize Latin-American states as colonial states, we have to envisage the question of the duration of those colonial states as such. Those colonial states were to change a lot, from the narrow Creole or oligarchic states of the beginnings to the “modernizing states” which appeared between the 1920s and the late 1950s,^{Footnote 77} and again after the democratization process of the 1980s. But my own view is that that is not sufficient to observe such a mutation: a Creole or oligarchic state is a state in which only the upper class of the colonial society obtains the state’s power. Its social basis may grow within the colonial society, to express the upper-middle milieus or the middle-class milieus, or even, in some populist cases, the lower milieus, but it is still a coalition of social milieus within the colonial society. “Colonial state” is not a class characterization as are, for example, in Marxist concepts, the bourgeois or worker states, but a societal concept: the state is the state of the colonial society, and not that of the colonized society. The country itself is imagined as “Latin,” and not indigenous/Indian. Indeed, class and societal characterizations quite clearly combine.

Understanding the nature of present-day Latin American states as colonial states allows us to see the Indians’ struggles not only as community struggles but as ongoing decolonization processes. Quite clearly, decolonization of Latin America could occur only in the countries where the “natives” (often with close mestizos milieus) have remained the majority. From this point of view, whilst Hugo Chavez’s so-called “Bolivarian revolution” remains a colonial (Hispanic) revolution, Evo Morales’ Bolivian politics may be analyzed as an attempt to decolonize, to “deLatinize” and even to “deBolivarize” Bolivia, to change the very nature of the state in order to have an Indian Republic in the same way that in Africa one has African Republics. In the other countries where Indians are only a minority, sometimes even a very small, marginalized one, and where full decolonization of the state will never occur, the Indian aspiration for decolonization has often taken the form, apart from struggles for social justice and respect for their land, of claims for “indigenous autonomy,” as in Chile with the Mapuche nation today or, on several occasions during the twentieth century, in Mexico or even Peru.

The colonial state, even the modernizing one of recent times knows very well who the natives are, and the very feeling of that demonstrates the nature of a state which manages two (obviously interconnected) societies within the same territory. When the natives have virtually disappeared, as in the United States of America, the coloniality of the state is not only expressed by the existence of a community of black slaves descendants, but primarily by the fact that its capitalism has been able to expand internally, within the country, in lands now rendered “empty.” Space is a fundamental marker of coloniality (why is there “space” in the New World?). Laurence Whitehead - who does not adopt the concept of coloniality of power - quotes a striking example, that of the architectural projects at Mexico City for the Centenary of Independence (1910). The model was Haussmann’s Paris but, “in contrast to Paris, in Mexico City it was not necessary to destroy urban sectors or to relocate largely inhabited zones in order to build these new avenues and boulevards.”^{Footnote 78} And, quoting Mauricio Tenorio Trillo:

It is certain that the frontier expansion was felt as an expansion over “nature,” since the society existing within that “nature” was simply not relevant, and thus did not exist for the imagining of the nation’s capital city.

Independence with Decolonization

In sub-Saharan Africa, the situation is completely different and belongs to the third category: independence with decolonization. African societies were defeated and conquered, exploited and humiliated, partially acculturated and socially “reshaped,” but those African societies have continued to exist as such, with their ethnicities, lineages, clans, age classes and sometimes matrilinearity, as the vast majority of the whole populace. The coloniser did not produce society, and only represents a small minority. As in Latin America, the imperial powers had not built local “states” but imperial administrations of the occupied lands (the foreign state local estates already mentioned). The domestic mode of production was not destroyed, but was subjected to capitalism. But, unlike in Latin America, the independences did not create colonial states. The new states of today may be perfectly “neo-colonial” - but “neo-colonialism” relies on the kind of policy and relationship with the former imperial power maintained by governments, not on the very nature of the states and countries - they are no longer colonies in which states of maintained colonies could develop, meaning that decolonization did in fact take place: they are African states. Indeed, there were very few colonial states in Africa: South Africa from 1910 to 1991, Rhodesia from 1965 to 1980, and, in very specific forms, Liberia from its creation as an independent colony in 1847 to 1980, and Sierra Leone from its independence in 1961 to 1991, until the coups d’état and civil wars which destroyed their exogenous Creole states.^{Footnote 80}

African Coloniality of Power

But would this mean that African (decolonized) states do not suffer from any coloniality of power? Once again, it must be stressed here that I am not speaking about the politics of their regimes, which may be perfectly “neo-colonialist,” but about the very nature of states and countries. When Africans (natives) could clearly not see and consider the European imperial administration as their state and as a state for the whole population of the colony, the African-ness of the new states gave them the minimal legitimacy to be felt as the states of Africans. In Mozambique, people said that, at independence, they were glad to have “um governo da nossa cor” (“a government of our [skin] color”). African independent states are therefore not colonial states as in Latin America, but states with a deep dimension of coloniality (in Quijano’s sense). I propose to qualify them as colonialistic states, deriving from the Hellenistic states stemming from Alexander the Great’s conquests in Asia, with a rapid indigenization of their former Greek elites but also maintenance of part of their legacy. This means that they are truly African states and felt as such by the population, but that their very birth had not been thought out to represent African societies: they had to accept the legitimacy of the colony’s borders, even if those cut right through long-established African nations.

This often induced an overwhelming preponderance of the small elite living in the capital city and socially produced by the late imperial apparatus (without preventing, on the contrary, ethno-patronage links with the former indigenous population). This elite has had to quickly legitimize the country’s absolutely incredible boundaries and space and to build up the “nation” by denying the first nations (consequently called “tribes,” “ethnic groups,” and so on: “For the nation to live, the tribe must die,” said Samora Machel, Mozambique’s first president). The logical consequence was a certain idea of modernity, with literacy generally being developed in the colonial language, with city buildings being made of concrete (and not from local materials), and with the import of European consumer goods for the elite serving as a model for the whole society.^{Footnote 81} This does not only mean that colonial borders were “artificial”^{Footnote 82} - some colonial borders had respected the pre-established African ones, but that was far from being the general case -, but that, even if African peoples doubtless wanted independence and “governo da nossa cor,” they were never consulted about the areas of the new countries, the idea of the new nation, which had to be the former colony. Anyone who wanted to change that was therefore rejected as “tribalist.” All this means that most African states today were structurally not founded as representative of the peoples totally or partially embedded within their borders, but as states for managing “spaces” and their relationship with the capitalist world-system – what Frederick Cooper calls “gate-keeper states.”^{Footnote 83} The intangible fixation of the colonial boundaries is one of the main “markers” of the coloniality of power in Africa. Contesting such boundaries would doubtless provoke trouble and violence, but the historian cannot but note that their imposition has also provoked social and economic difficulties and bloody problems since the independences. Will the independence of Southern Sudan signal a new epoch? It is too soon to know if it will remain an isolated exception, or signal a new approach to nation formation. But the first hypothesis is more probable.

It may be said that, in Latin America, several wars opposed the new independent countries as regard their borders during the nineteenth and even twentieth century. That is true. But even if the soldiers could be mixed-race or blacks or Indians, those wars were waged by the independent colonial state elites in order to define how their countries would be. Naturally the colonized Indians were not consulted, but it was not a matter for their state (they did not have one) but for the state of their Hispanic/Luso-Brazilian masters. In Africa, in spite of the coloniality of power within the new states, it was a matter of African states, of their states and they were not authorized to discuss (or to wage wars) to define the new spaces. Everything took place in Africa as if, in Latin America, the Hispanic elites had not had the right to discuss, or to wage wars one against another, in order to define what the very country and its boundaries would be.

Coloniality of Power in Portuguese-Speaking Africa

That is the global contrast between Latin America and Africa. But what about Portuguese-speaking Africa? We have already noted one of its peculiarities - the diversified but ever present Creole issue - which is nothing other than the survival of the first age of colonization within the new one. That feature has a geopolitical dimension. Considering French colonization, for example, there was a clear difference between the “Vieilles colonies” (“Old Colonies”) in the Caribbean (since India and Canada had been lost) and the “Nouvelles Colonies” (“New Colonies”) in what is now called sub-Saharan Africa.^{Footnote 84} Great Britain, too, had lost the main part of its “Old Colonies” (United-States of America) but retained some Caribbean islands. India, for France (Pondicherry) and for Portugal (Goa), included too “Old Colonies,” but very modest in area and in economic relevance. For Great Britain, after the Treaty of Paris in 1763, India became a kind of “early New Colony,” already with characteristics of the new ones, i.e. the conquest of colonized societies and not the production of colonial societies. For Spain, it was the same, with the difference that its “Old colonies” were now independent states, and that the new ones in Africa were modest (Equatorial Guinea; north of the Rif, Ifni; Seguia el-Hamra and Rio de Oro, present-day Western Sahara). It was a question of colonial ages, that is above all a question of social formation processes. The “Old Colonies” in the Americas had been entirely shaped by the colonizer,^{Footnote 85} since the native populations had been eradicated or deeply marginalized. In the French Antilles, the whole of the population (the white “Békés” as well as the black slaves and the free coloured people) was colonial, being integrated into a colonial (and not colonized) society. The same may be said about the British West Indies and Spanish islands. It was more complicated in the Portuguese Estado da Índia (Goa), where there was a strong distinction between the “Old Conquests” which were profoundly “Portugalized” as and from the sixteenth century (including by way of the Inquisition) and the “New Conquests” of the eighteenth century which have maintained their largely Indian society. In Pondicherry and other small French “comptoirs,” a significant number of “assimilated” Indians, who were made French citizens, as in the Caribbeans, then emerged as a relevant social category.

The social formation within the new wave of colonization in Africa was completely different, with the massive presence of colonized societies (Africa can, from this point of view, be compared with Asia) and a reduced number of colonizers. Even in South Africa or Algeria, the presence of the colonized societies remained massive. In the French case, the peculiarity of the social formation was later illustrated by the complete integration of the “Vieilles colonies” in the motherland (decolonization without independence), which means that assimilationist anti-colonialism had a huge social basis. In Mozambican old Zambezia (the earliest and most deeply colonized province of Mozambique, land of the prazos), the same feeling (the will to be, or to stay, “truly Portuguese”) remained an urban isolated will in the midst of the rural colonized society.^{Footnote 86}

But the peculiarity of Portugal is that there were at the same time and place “Old Colonies” and “New Colonies” in Africa: on the psychological level, Portugal, ever since the glory of the Discoveries, had thought that large parts of Africa were already part of its sovereignty (such as the Kingdom of Kongo, the present-day hinterland of Zimbabwe and Zambia) and, at the end of the nineteenth century went to “submit,” but not to conquer, its “own” territories.^{Footnote 87} The most important aspects, nevertheless, are geopolitical and social: Portugal went to “submit/conquer” its hinterland from the position of its old colonies within the same area. It was in the same continent and there were no ruptures in the land continuum, as if it were merely a dilatation of the old colony. Even if, as we have seen above, processes of marginalization of the old Creole nucleus occurred (differently in Angola and Mozambique), the fact remains that when a “new colony” is conquered from the “old one,” the social, political, psychological and cultural weight of the latter is bigger. We have already quoted Braudel’s “temps long dans le temps court,” the concatenation in the same colony of two ages of colonization. In Angola, this led to the huge division of anticolonial movements, with three movements stemming from three social trajectories of elites, each one powerful enough to imagine itself as being legitimate for the ownership of the whole of the nation.^{Footnote 88}

In Mozambique, the marginalization process of the old colonial nucleus of the northern part of the country led to the birth of modern anticolonialism in the far South, in the capital city Lourenço Marques, a city entirely produced by the new colonial capitalism and without any trace of first age colonization. The very tiny African elite of Lourenço Marques had, therefore, no links with the old nuclei which were unable to forge their own anticolonial expression. Consequently, the uniqueness of Frelimo does not express the capacity of unifying several branches of anticolonialism (as official and sympathizer academic historiography have argued). On the contrary, it expresses the incapacity of other milieus to develop that form of expression - with the exception of very minute groups. Those milieus felt very uncomfortable within Frelimo, seldom joined it and when they did, were quickly expelled for “tribalism” or for taking a “counter-revolutionary line.” Those imbalances stemming from the colonization were therefore reproduced during the armed struggle and in the new independent state. I have written that civil war in Mozambique did not begin in 1977 with the Renamo uprising supported by Rhodesia, but in 1898, with the change of capital city from Ilha de Moçambique to Lourenço Marques.^{Footnote 89} The very historical core of the colonial country was denied. In Angola, the MPLA elite had to fight against other elite nuclei but in Mozambique, Frelimo has taken up the whole of national imagining; it was a single state-party but still very much a single nation-party.^{Footnote 90} In Angola, Unita and FNLA wanted to defeat the MPLA state in order to replace it. It is worth noting the slogan of Jonas Savimbi (leader of Unita) in the electoral race in 1991: “É a nossa vez” (“It’s our time [to get power]”).^{Footnote 91} Renamo (and its apartheid supporters) never wanted to beat Frelimo, but to oblige it to change. Renamo did not want to build another nation, as FNLA and Unita might dream about in Angola against the MPLA nation, but wanted to enter the nation of its model, of its enemy. It was not rivalry between elites. It was rancour between the historically “lumpenized” nuclei of an old elite able to gather about them certain sections of marginalized peasantry, and modern Frelimo elites.^{Footnote 92} Renamo was a strange alliance between certain old elite nuclei, who have been socially close to those who sustained the birth of the MPLA if they had not been marginalized by the whole of twentieth century history, and peasant sections of society, against the new African elite of the new colonial city, Lourenço-Marques, the social basis of Frelimo. In Angola, the “temps long dans le temps court” brought to power the MPLA, which historically stemmed from the old elite nuclei. In Mozambique, the same concatenation led to the deep marginalization of its own old elite nuclei. Clearly, today, with huge urbanization and concentration in the capital city (in particular in Angola), the social formations are once again disrupted. Nevertheless, one cannot understand the policies in those countries without bearing in mind those “old stories” from history.

The Strength of the Independent Colonial State...

Independent states in Africa are African states, even if they are deeply rooted in coloniality. Latin American states are colonial states. Even if Indian and close mixed-race population are in the majority, the state has remained structurally colonial, based on a strong minority of colonial people. But saying that they are colonial states does not mean that they are weak.

On the contrary, that is one reason for their relatively strong, historically built, stability (even if that can be questioned, as in Bolivia). Indeed, in that kind of state, the nature of the state corresponds to its own population: both they are colonial, with their colonized fringes being very tiny or indeed not able (even now) to challenge that founding couple. Even if the capitalist mode of production had not been implemented before the end of the nineteenth century, the whole merchant economy, slavery, etc., was integrated from the very beginning into the capitalist world-system.^{Footnote 93} Coloniality was not only brought by capitalism, it was constitutive of the capitalist world-system in the sixteenth century. There are few, or in any case far fewer than in Africa, of what Marxist anthropologists have called the “articulation of modes of production,” i.e. the global submission of a surviving domestic mode of production to the indirect rule of capitalism through imbalanced trade, slavery, forced labour, etc.^{Footnote 94} A great part of the economy is “integrated” and the imagined community of the nation works within this link between the nature of the population and the nature of the state. Even if the Indian question has remained important in some countries, in most of them the state is stable, historically linked to its (colonial) people. That does not mean at all that governments will be stable, as a huge number of coups d’état has shown, expressing very often the social formation exacerbated between a tiny rich Hispano-Creole elite and a mass of poor Hispanic and mixed-raced plebe.^{Footnote 95} But there is a basis on which to build a country - there is a people, there is a nation. That nation-building has been greatly facilitated by the formation of the colonial state. The Latin-American state successfully fulfils its tasks. Objectively, that is a favourable context for democratic life, aspirations, claims and struggles, which does not mean that democracy will always win, but that extreme violence would be needed as military coups (specially in Chile, 1973, and in Argentina, 1976) have shown, if democracy is to be thwarted.

... and the Weakness of Colonialistic African States

In Africa, too, there has been a lot of violence, but democratic issues have not necessarily been the cause. Very often, the issue is bound up primarily with the very nature of the country. Paradoxically, the fact that independent African states are not colonial states is a cause of weakness. Indeed, they are African states - unlike Latin-American states which are not Indian states - but they have been profoundly shaped by coloniality: peoples have not even been consulted about the very countries to be built. The colonial ideological hegemony has been accepted by the new elites, as we have seen. Even if large majorities of Africans easily recognize that they are “Senegalese,” “Mozambican,” “Gambian” or “Namibian,” etc., that expresses the minimal level of political acceptance to be the member of a state obedient to its chief, and that does not suffice, however, to form a historically rooted imagined community. The “nation” still remains in posse. The state must manage the link between the local area and the world-economy, but implementation of capitalism (as a mode of production) is not yet generalized and the domestic mode of production still survives, in spite of galloping urbanization and rural exodus. The state has a problem with its own population. The ‘gate-keeper state’ is definitely not a nation-state (in spite of its “stato-nationalist” ideology). Its elite must “build the nation” very quickly in order to legitimate itself, which sets off a permanent top-down development imagining.

The Latin-American “Flavor” of Portuguese-Speaking Africa

Portuguese-speaking Africa does not escape from such a situation. But we find here what I will call a “Latin-American flavor.” I mean here that, even if former Portuguese Africa has not experienced a kind of Latin-American colonization, some common factors do exist, with diversified consequences. It is not a case of language, but of “longue durée” social formations. At the end of the nineteenth century and for the first third of the twentieth, to be “Angolan” was not at all to be African, but a Portuguese (white, mixed-race or black) from Angola, versus the gentio (the native). This elite, as we have seen, was a non-capitalist colonial elite, and it took power in 1975. But, unlike in Liberia or Sierra Leone, those Creoles were not “imported” but historically produced in loco. Angola is the country where this phenomenon has been the most important: the colonial census of 1950 still contains traces of this, with fewer than 5,000 assimilados in Mozambique and more than 30,000 in Angola, whilst the first colony had a far bigger population. In Angola, this indigenous Creole elite, with its skilled manpower, could be an important factor for democracy. Civil society was a possibility, far more so than in Mozambique or Guinea-Bissau. Indeed, the late colonial period had also created what Christine Messiant called the “new assimilated,” thanks to literacy, social improvements, etc. This new social urban milieu, with part of the old elite, provided the social basis for the attempted leftist coup d’état by Nito Alves in May 1977, a coup from within the MPLA, expressing social tensions between different trajectories. The attempted coup killed fifteen persons, but its repression led to some 15,000 deaths.^{Footnote 96} It was the generalized operation of a frightened elite to secure its power in a country it did not recognize. A few months after the fierce repression of the 1977 coup, the MLPA proclaimed itself as “Marxist-Leninist,” thereby expressing its will to gain complete control of the state. The coup’s repression and the civil war years built a strong state which has been able to replace the single-party by an ultra-hegemonic one.^{Footnote 97} Here, too, as one can see, a globally favourable context for democracy imposed the need to resort to massive violence to impose a single party. Nevertheless, in the long term, the globally favourable context remains and will necessarily empower fights for more democracy. It is not just a matter of chance that the Arab revolutions fill some youth people in Luanda with enthusiasm - which is not the case in all African countries.

In Mozambique, the Latin-American “flavor” is weaker, except in “old Zambezia,” Moçambique Island and other small places. Nevertheless, huge imbalances between the regions are not only a question of today’s social conditions, but of the “longue durée” trajectories of the northern parts of the country. The regions where the first age of colonization had been the most marked are still marginalized. The political weakness of Renamo leadership prevented its profiting from this situation, but that was not just by chance: once again, “old Mozambique” has been unable to express by itself, to build its own (non-ethnic but surely regional) political expression. “1898” has not yet been overcome, but it remains a permanent factor that Frelimo must deal with.

In São Tomé and Príncipe, the small size of the country is a very significant factor:^{Footnote 98} the Creole (Forro) issue remains, along with its permanent rural problem, but the political disputes all take place within a few old families. Victory or defeat of one or other party does not mean, unlike in Cape Verde, a change in the political culture,^{Footnote 99} but that is a factor of permanent “soft” instability (the coups d’état in São Tomé have been quite mild and quickly resolved). In that light, the discovery of oil is not necessarily a good piece of news.

Guinea-Bissau is becoming more and more a Narco-State and clearly that phenomenon is not linked with a Creole issue. Nevertheless, the importance of this social milieu,^{Footnote 100} and its virtual absence within the army, continues to express the imbalanced history of the armed liberation struggle.^{Footnote 101}

In the Cape Verde islands, as we have seen, the situation is completely different, since creolity is not a social milieu, but the whole of the nation. Cape Verde is undoubtedly the most “Latin-American” country of Portuguese-speaking Africa, since it has been entirely produced by the imperial state. Its society is - as in the French Caribbean or in the Cono Sur Latin-America, where Indian people are very tiny communities - mainly or almost exclusively a colonial society, and has never been an African colonized society. But there is a huge difference with the Latin-American cases since, because of the departure of the white Portuguese during the long decline (eighteenth and nineteenth century), the lowest class within the colonially-produced population remained the vast majority of this population: the slaves (including mixed-race slaves), as well as a few mixed-race foremen and free coloured people. It is indeed a strange case of colonial society, almost entirely made up of the descendants of the lowest class within the colonial world, the slaves! It could be compared with Haiti, if miscegenation had not been a major factor in Cape Verde, which did not experience a slave plantation complex as in Saint-Domingue. The Creole factor is so massive in Cape Verde that it has diversified itself, not only because of old trajectories^{Footnote 102} and insularisms, but because of modern globalization: the massive emigration towards Portugal, France and Africa on the one hand, and towards the Americas or Holland on the other hand, has globally shaped two different political cultures. One of those is more “Africanist,” as is clear from the very name of the PAIGC/PAICV (Partido Africano...) and its flags (with the very frequent colours of African flags: red, yellow, green); another is more “Euro-American” (as expressed in the new flag after the 1991 opposition victory - blue with twelve stars which, without its tiny red-white strip, would look amazingly like the European Union flag). Whatever the case, conditions for a democratic life are certainly better in Cape Verde. Even during the single-party time, very few people were tortured and people were never shot down by PAIGC.^{Footnote 103} The same party in power in Guinea-Bissau committed such acts on a large scale...

Establishing a comparison between “Latin” America and Portuguese-speaking Africa has not been useful merely to compare per se. But, by avoiding any culturalist approach, it has allowed me to highlight some features, such as the evidence of certain common historical trends, the long existence of a non-capitalist colonization and the production of non-capitalist colonial local elites who wanted to colonize their own country – the very presence of the “Old colony” within the “New colony”^{Footnote 104} – all of which explains the Latin-American “flavor” of former Portuguese Africa. Clearly, the differences are huge, including what concerns the “Creole issue.”

Even if I strongly stress this factor for a better understanding of Portuguese colonial history in Africa and today’s Paises africanos de lingual oficial portuguesa, it seems clear that creolity has been far more relevant in Latin America than anywhere in Africa, including former Portuguese Africa. But this demographic “modesty” does not prevent its having been a key long-term historical factor in the state formation process, far more so in Portuguese Africa than in the other Africas.

The geopolitical and geo-economic urban aspect of the turn to colonial capitalism was diversified: from Cacheu to Bissau in 1942, from Moçambique to Lourenço Marques in 1898 as capital cities; from Benguela and Moçâmedes to Luanda as the main city, even if this latter had always remained the capital city; the final marginalization of other small towns – all Cape Verdean towns, and São João Baptista de Ajuda, present-day Ouidah in Benin. But such a turn may be found, too, in Latin America: from La Paz to Sucre in Bolivia, from Oaxaca to Mexico City in Mexico, from Quito to Guayaquil in Ecuador, etc.^{Footnote 105} But probably the “weight” of such a urban transfer was less in South America than in Africa: the fact of becoming the capital city (or even the main economic city) is of tremendous importance in Africa, deriving from the neo-patrimonial ownership of the elite over the gate-keeper state – capital city transfer means the very displacement of the state itself, and a huge change in the imagining of the whole country. Furthermore, changes were politically decided within a very short space of time in Guinea-Bissau and, even more so, for external reasons in Mozambique, whereas change in the Americas rather expresses the result of its slow former evolution.^{Footnote 106} Be that it may, changing the capital or main economic cities has led to change within the ruling social group. This may have confirmed the weight of the old elite (Angola) or, on the contrary, signaled its weakening – but this weakening has become a permanent problem for the country (Mozambique).

No direct political consequence should be drawn from such features, as we have seen that the same factor (a Creole elite, for example) may have different consequences according to the period, contexts and colonies. But this kind of long term “gravity,” a distinctive feature of former Portuguese Africa within Africa, is one factor for democratic building, and this factor is, at least in part, the same as in Latin America.