Liszka József (szerk.): Az Etnológiai Központ Évkönyve 2000-2001 - Acta Ethnologica Danubiana 2-3. (Dunaszerdahely-Komárom, 2001)

1. Tanulmányok - Borsos Balázs: A magyar nyelvterület számítógép segítségével meghatározott kulturális régiói

Unfortunately, the atlas itself and the database set up on the basis of the data of the atlas have many problems. Some weaken the correctness of cultural regions defined by computer and some are to be solved before the analysis. As I have dealt in detail with these problems elsewhere (Borsos 2000), now I just summarise them briefly:

The atlas, though it shows the regional distribution of around 634 cultural phenomena, does not represent Hungarian folk culture completely. Due to the characteristics of ethnocartography, elements of material culture are overrepresented and elements of social life and folklore are underrepresented on the maps. Games, dance and music are totally neglected, as well as (with few exceptions) the texts of folklore. Branches of culture shown in the atlas are a bit unilateral, too. Some phenomena (for example guiding and ordering animals, or the occurrence of certain first names in the population) are also overrepresented, compared to their cultural importance.

The atlas contains data from 417 settlements, comprising only 3,33 % of the settlements of Hungary around 1900. However, the choice of settlements for collection was determined so that they had to represent both the culture of their neighbourhood, and had to cover the whole Hungarian-speaking area (Barabás 1987-92. Introduction.) A more serious problem is that due to the character of the cultural phenomenon shown and due to problems of collecting data, many of settlements do not have data on all maps.

Although each map is to be considered equal in the following analysis, the cultural phenomena shown on them are not. Different variables of some phenomena are shown in two or more maps, some maps represent more than one phenomenon, and several maps show only details of cultural elements presented on another map.

Some maps deal with the changing of a certain cultural phenomena through time. These cannot represent folk culture around 1900.

These problems can be solved by sorting out settlements and maps not suitable for later analysis. But, due to the law of great numbers, the result will not be influenced excessively whether we take them into consideration or leave them out. The one-sidedness of the atlas itself is more problematic. This can be solved only with comparative investigations of the territorial distribution of cultural phenomena not shown in the atlas.

Problems to be solved before the analysis

As computer elaboration was not taken into consideration during the editing of the atlas, there are inconsistencies in the atlas that cannot be dealt with using a general program for databases. For this reason we had to write a separate program, and one for the analysis as well. This work was done by programmer László Gábor Breiner, whose field of specialisation is cluster-analysis.

According to the principle of cluster-analysis, every object has to be put in an n dimensional virtual space, according to the values of the n variables characterising them, and the distance between these virtual points has to be investigated. The arrangement and the investigation are done by computer. Points closest to each other get into the same cluster or group. But cluster analysis can also be viewed as a process whereby at one end of the analysis every point forms a single cluster, and at the other end all points are in the same cluster. The small-53