THE CENTRAL PLACE SYSTEM IN KOREA

Abstract

The purpose of this paper is to classify objectively central places in Korea into hierarchical orders. In order to clarify the Korean urban hierarchy, three methods (i.e. Davies', Beavon's and the factor analysis) were employed in this study. Though these three methods have their own merits and demerits, the writer would like to point out that the results based on Beavon's relative homogeneity (H'q) seem to be best suitable to classify central places in Korea into hierarchical orders. Based on the results of Beavon's H'q, the hierarchical orders of central places were compared with such previously verified characteristics as the orders of nodal region and the integrated charateristics of Korean cities to understand central place system. The data for 157 central places and 57 central establishments in the “Municipal Year Book of Korea in 1975” were analyzed. The main results are summarized as follows: 1. Central places having similar Davies' functional index are found in various hierarchy cal levels due to the difference in the composition of business type. Peavon's frequency modulated relative homogeneity (H'qm) tends to overestimate a central place which has some specialized central functions in the process of weighing the central function, whereas his H'q does not emphasize the difference by deviation of the score of a specialized central function. In addition to these, his method does not emphasize the particular scores of those central places but the similarity among central places. Rn the other hand, factor analysis using incidence matrix tends to disregard qualitive characteristics of central functions, so that there is a great loss of details. However, in the factor analysis using ratio data, the difference of structure and the size of central functions are confused in the process of analyzing correlation matrix, which causes difficulties in its applications. 2. The hierarchical orders of 157 central places were analyzed and classified in six orders using Beavon's H'q and the Ward's cluster analysis (Table 2, Figs. 4 and 5). It was proved that the central place hierarchy in Korea corresponds in no way to Walter Christaller's three principles in terms of number of central places. The orders of nodal region based on long distance telephone calls of 35 cities are identified by three orders (Table 3 and Fig. 6). Seoul dominates all the central places not only in the hierarchical orders of central places but also in the orders of nodal region. The second and third order central places, such as Busan and Daegu, have their sphere of influence in the city level, but are subordinated to the sphere of influence of Seoul. Central places below the fourth order do not have their own sphere of influence. This means that with the modernization of transportation system, the sphere of one day's journey from Seoul covers whole Korea, which results in the subordination of all the other centers to Seoul. 3. Population size, the number of establishment as well as population density and population growth rate have shown tendency of increase in proportion to their hierarchical orders (Table 4). The first order, Seoul, is with the highest level in population size, functional establishments and population density. Busan and Daegu are next to Seoul. The three higher order centers hold more than 60% and 50% in urban population and all functional establishments, respectively. The fourth order centers, Incheon, Gwangju, Daejon and Masan, are placed in the medium order. Through the comparison of the hierarchical orders of central places and the results of integrated characteristics in Korean cities (Table 5), it has been proved that Seoul is characterized by its primacy. The six largest centers and other specialized industrial centers of Masan, Pohang and Ulsan have played an important role in the process of high economic growth in Korea.