Today, national and ethnic identities (and, for that matter, all other collective identities) are routinely interpreted as constructions fashioned in order to create or confirm ‘imagined communities', to use Benedict Anderson's now famous term (l991). Authors such as Hobsbawm (1990) and Urban and Sherzer (1992) have developed the argument by showing how political elites have used and manipulated popular beliefs and customs to further their nationalist projects. Still, it may be asked whether this emphasis on the constructed nature of social ideologies is, in the cnd, satisfying. It leaves two problems unsolved.

First, the symbols and historical interpretations which are chosen to bolster ethnic or national identities are not completely arbitrary, nor is their emotional appeal. It may be true that these symbols are distorted, exaggerated, sometimes invented, but even in this latter case, such inventions do not fall from the sky. They originate in the history or the culture of a given group of people and are only accepted when they do not deviate too far from existing cultural perceptions and social memories (Fentress and Wick ham, 1992). These memories are not necessarily true themselves, but they are social facts at the moment of their general acceptance.

Second, the selection and manipulation of symbols do not go uncontested. Actors in the social arena have different interests, which llre reflected in the WhY ideologies are received and interpreted. This process is well-documented for intellectual and political elites, but often ignored in the Case of the common people. Superficially, the latter often show II more or less passive acccptance of the constructions of dominant groups, politicians or ethnic leaders, but this is not to say that they necessarily believe all their ideas or accept all the consequences of those ideas. The emphllsis on the constructed nature of national and ethnic identities should not close our eyes to the lack of success of many of these constructions

In this context, Harry Hoetink's The Two Variants in Carihhean Race Relations, first published in Dutch ovcr thirty years ago, still provides important insights.