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

tural features shown in the Atlas, the settlements here are more similar to those of the Highlands. The border between the northern part and the rest runs clearly along the line that divides the Nyírség and the Hajdúság. There seems to be a transitional region between the Plains and Transylvania, in which the culture of settlements seems to be also similar to the Highlands. In most cases, however, (as we shall see at the ninth step) it is not so, and the villages, especially in the middle and in the southern part of the zone are, in fact, the ones characterised by a lot of 0 values. At this phase of the analysis they still go with the northern region. With an absence of data on 502 maps Kecskemét has to be ordered into this special group, too.

Between step four and five we can find only a very small decrease in the value of variance (Map 5). But it still divides the settlements of this most problematic northern-transitional zone into two clusters: the real Highlands (-) and the rest (+). But there are some differences in the borderline drawn by previous research and by the computer. The southern and the eastern borders are the same: the line where the Plains meet the mountains, and the region of the Eperjes-Tokaj Mountains. Although traditionally the Highlands extend to the west as far as the River Garam (Kosa 1998. 246.) or even as far as the River Vág (Paládi-Kovács 1994, 21.), the computer, basing itself on the data of the Atlas, finds that the settlements along the Garam and the lower Ipoly valley are more similar to the villages of West-Hungary. At the western border of this region we can find another exception, Óvár, but as in the case of Kecskemét, here we lack a sufficient amount of data (405) again.

When we form 6 clusters (Map 6) we can find a less typical distribution due to the limited decrease of variance (6,61-5,97). The new cluster defined at this step is that of Southern Transdanubia ($). This cluster contains the settlements of Somogy, Baranya and Slavonia. Traditionally Southern Transdanubia is bordered by the River Zala, Lake Balaton, the Rivers Sió and Dráva, but former research emphasized the clear difference between Göcsej in the west and Sárköz in the east (Kosa 1998, 218.). The computer, having elaborated the data of the Atlas, draws attention to this division, by not counting either Göcsej or Sárköz in this cluster.

The most dramatic decrease in variance (5,64-4,13) in the middle phase of the analysis happens when 9 clusters are formed (Map 7). This distribution of cultural regions can also be viewed as typical, but the dramatic decrease happens at the step when a cluster of a very special character is formed. At step 7 the Hajdúság and the Nagykunság (1) are separated from the rest of the Plains (*), and at step 8 the westernmost part (+) of West Hungary forms a different group. Step 9 calls for the special cluster mentioned above. Contrary to the others, this cluster is a dispersed one. The villages that belong to it, are those of a lot of 0 values, which means that a lot of data are absent (Table 2).

Unfortunately, they are mainly in the transition area between Transylvania and the Plains, which hinders us from drawing a clear and correct border in this important territory. It is worth mentioning that from the first 22 settlements, where data are absent in more than 380 maps, only 18 form this cluster (without a sign in brackets in Table 2). The other four are all villages from Transylvania. With two other Transylvanian settlements they form a separate cluster only at step 39 (sign !). There are four villages (sign §) left of the group characterised by a lack of more than 50 % of data, which are not members of any of these clusters even at this stage. This means that the data characterising these (mainly Transylvanian and Moldavian) settlements in the database of the Atlas, are very similar to the data of villages around them.

57