|1-2(9-10)/2001, p. 89
Jadwiga Sławińska, Barbara Widera
Trends in contemporary architecture. Problems with its separation
Methodology, used in the history of architecture, as well as in many fields of science, was developed mainly in the period between the First and Second World War. It has respected the rules of traditional logic: specifying strict definitions and correct (i.e. dis-joint and exhaustive) divisions.
However, objects existing in the real world are not usually assembled into delimited sets, with clearly defined boundaries. Using the rules of traditional logic, in many dynamically developing branches of science (especially natural and technical sciences), has been confronted, in the last several years, with some difficulties. The solution appeared with fuzzy logic (competitive to traditional logic) coming into existence.
Some of the methodological assumptions, verified in those disciplines, may also be applied to researches on the history of the newest architecture. An assumption can be made that the trend in architecture makes a fuzzy cluster of the elements – architectural works. Affiliation to the trend, as well as to many other fuzzy clusters, is a graduative feature. To determine the level of affiliation, for a particular building, we use a model. As the model, we understand a group of all the features typical of the trend. The complete set of features is established on the basis of the cluster of all the architectural works, belonging to the trend treated as a whole.
The particular trend differs from any other, just in such a cluster of typical features, although these features do not appear in the full set, neither do they in each element (i.e. an architectural work) of the cluster, nor even in most of the elements. It seldom happens to find even one building which would present all the features distinguished as typical of the model.
The model is a very useful research tool. It allows us to over-come the limits of radical empiricism – collecting of only separate facts and avoiding any attempts at generalization. On the other hand it may prevent a too unrestricted segregation of trends too far off from reality.