Bantu languages have been categorized by scholars into a number of sets of related tongues. Some of the languages in any set may be more or less mutually intelligible, especially in the areas where speakers of a dialect of one language have had sustained contact with speakers of a dialect of another language. Given the mobility and interpenetration of communities of Bantu speakers over the centuries, transitional languages--for example, those that share characteristics of two tongues--developed in areas between these communities. Frequently, the languages of a set, particularly those with many widely distributed speakers, would be divided into several dialects. In principle, dialects of the same language are considered mutually intelligible, although they are not always so in fact.

Language alone does not define an ethnic group. On the one hand, a set of communities lacking mutually intelligible dialects may for one reason or another come to share a sense of identity in any given historical period. On the other hand, groups sharing a common language or mutually intelligible ones do not necessarily constitute a single group. Thus the Suku--most of them in Zaire but some in Angola -- had a language mutually intelligible with at least some dialects of the Bakongo. However, their historical experience, including a period of domination by Lunda speakers, made the Suku a separate group.

Although common language and culture do not automatically make a common identity, they provide a framework within which such an identity can be forged, given other historical experience. Insofar as common culture implies a set of common perceptions of the way the world works, it permits individuals and groups sharing it to communicate more easily with one another than with those who lack that culture. However, most Angolan groups had, as part of that common culture, the experience and expectation of political fragmentation and intergroup rivalry. That is, because one community shared language and culture with another, political unity or even neutrality did not follow, nor did either community assume that it should. With the exception of the Bakongo and the Lunda, no group had experienced a political cohesion that transcended smaller political units (chiefdoms or, at best, small kingdoms). In the Bakongo case, the early Kongo Kingdom, encompassing most Kikongospeaking communities, had given way by the eighteenth century to politically fragmented entities. In the Lunda case, the empire had been so far-flung and internal conflict had become so great by the nineteenth century that political cohesion was limited (see Kongo Kingdom Lunda and Chokwe Kingdoms , ch. 1).

Very often, the name by which a people has come to be known was given them by outsiders. For example, the name "Mbundu" was first used by the Bakongo. Until such naming, and sometimes long after, the various communities or sections of a set sharing a language and culture were likely to call themselves by othefdb her terms, and even when they came to use the all-encompassing name, they tended to reserve it for a limited number of situations. In virtually all colonial territories, Angola included, the naming process and the tendency to treat the named people as a discrete entity distinct from all others became pervasive. The process was carried out by the colonial authorities--sometimes with the help of scholars and missionaries--as part of the effort to understand, deal with, and control local populations. Among other things, the Portuguese tended to treat smaller, essentially autonomous groups as parts of larger entities. As time went on, these populations, particularly the more educated among them, seized upon these names and the communities presumably covered by them as a basis for organizing to improve their status and later for nationalist agitation. Among the first to do so were mestiços in the Luanda area. Although most spoke Portuguese and had a Portuguese male ancestor in their genealogies, the mestiços often spoke Kimbundu as a home language. It is they who, in time, initiated the development of a common Mbundu identity.

In general, then, the development of ethnic consciousness in a group encompassing a large number of communities reflected shifts from the identification of individuals with small-scale units to at least partial identification with larger entities and from relatively porous boundaries between such entities to less permeable ones. But the fact that these larger groups were the precipitates of relatively recent historical conditions suggests that they were not permanently fixed. Changes in these conditions could lead to the dissolution of the boundaries and to group formation on bases other than ethnicity.

In any case, ethnic identities are rarely exclusive identification with other entities, new or old, also occurs in certain situations because not all sections of a large ethnic group have identical interests. It remained likely that earlier identities would be appealed to in some situations or that new cleavages would surface in others. For example, descent groups or local communities were often involved in competitive relations in the precolonial or colonial eras, and the conditions similar to those giving rise to such competition might still prevail in some areas. In other contexts, younger members of an ethnic group may consider their interests to be different from those of their elders, or a split between urban and rural sections of an ethnic entity may become salient.

In Angola, the displacement of hundreds of thousands of people, especially in the late 1980s, had significant repercussions on ethnic identification. For example, many of those forced to abandon rural areas and traditional ethnic communities for urban dwellings no longer engaged in agricultural activities and the small town life that defined their communities. Instead, they were forced to become urban laborers in ethnically mixed surroundings. Many were compelled by their new circumstances to learn new languages and give up traditional life-styles in order to survive in their new environment.

Data as of February 1989