Geography is a “field science” in the sense that it needs field work. In spite of it, the methodology on field work is not yet satisfactory through various branches of science including geography. Present article is a rough sketch and comments on the personal history of my inquiry into the methodology of field work, especially in the fields of geography and ethnology. In particular, a method of idea generation named KJ Method which was generated by me was explained in detail.
For the purpose of recording field data, I devised a form of data card (abbreviated as DC; cf. Fig. 1). In order to classify a file of DC, I tried to adopt the classification table of HRAF (Human Relations Area Files). Soon, however, I understood that classification only was much unsatisfactory for a field worker who sought for true integration of data. Quite different from classification and analysis, another unknown methodology for the purpose of “Facts speak for themselves” must exist.
In around 1951, I obtained a first hint for this purpose. And my work “Ethno-Geographical Observations on the Nepal Himalaya” (in Peoples of Nepal Himalaya, ed. by H. Kihara) became the first output along this new method of data processing. Later on, this methodology was greatly improved by myself and named KJ-M. in 1965 by various men. It was a nickname in origin. The first book systematically written on this methodology “Hassôhô” (Abduction) was published in 1967. (cf. References.) This method was welcomed very much, firstly in the fields of company management, business and engineering and gradually in the field of education and science.
In the basic KJ-M, there are four essential steps: a) label making, b) label grouping, c) chart making, d) explanation. Label making is to record one concept on a label usually in the form of sentence. (Rarely in the form of any picture.) Surrounding a theme, as rich variety of labels as possible are collected. Label grouping is attained through the repetition of the steps of making teams of labels and title making. Through this process, a number of labels are organized, not by the classification based on some pre-conceived ideas but according to the appeals of original labels. In the step of chart making, these organized labels are spreaded spacially on a sheet according to the recognition of natural relationship among the titles or labels. At last a multi-layered complex relationship between the labels is presented in a chart. Then the last step “explanation” is applied to this chart, i.e. explanatory story making connecting all labels in writings or by verbal explanation.
Using the basic KJ-M. repeatedly, we can challenge highly complex problem solvings. A full process of the basic KJ-M. was named “a round”. Multi-round application of this method is called the Cumulative KJ Method (C-KJ). When a C-KJ is applied along the course of problem solving shown in the W-shaped chart (Cf. Fig. 3 and Fig. 7), this method is most fruitful.
Using the steps of field work→recording DC→using KJ-M. and C-KJ, chaotic field data can be organized heuristically. The course of data processing is clear and open to anyone who wants to know it. Plenty of hints, hypotheses and generalization arise on the way and in conclusions. These suggestions are not harmful because of subjective judgement but welcomed because of stimuli to readers, as the grounds and processes which bore these suggestions are offered to them frankly. The readers may agree with the author or oppose him. In both cases, both parties will receive desirable stimuli through the dialogue. Thus a basic recognition that data are never facts and the means of observation and data processing intervene between the two leads to the truly scientific and charming geography or ethnography.

View full abstract

The aim of this paper is to discover the underlying hierarchical structure of the Nagoya Metropolitan Area, and to delimitate a group of smaller nodal regions, i.e., metropolitan sub-areas, which are included in the metropolitan area. First of all, the fringe of the metropolitan area is delimitated by means of several statistical data concerning the daily commuting movement, person's trips, origin-destination studies of road traffics, etc., which reveal the functional linkage between the metropolis Nagoya and its hinterland. This area, occupying the greater part of Nobi plain, has relatively few topographical obstacles and a fairly developed transportation network with Nagoya as its main starting point.
The sub-areas are delimitated by two methods, theoretical and empirical. The procedure of the former method, based on the gravity model well known as the Leilly's law, is as follows. If in a co-ordination system, the origin is placed at the one point (place A, See fig. 5) and the other point (B) is chosen as (d, o), then the condition for equal effect in (x, y) is expressed as follows, C_A/x²+y²=C_B/(x-d)²+y² which gives (x-C_A·d/C_A-C_B)²+y²=(d√C_A·C_B/C_A-C_B)² which means a circle with radius r=d√C_A·C_B/|C_A-C_B| and the center point m=C_A·d/C_A-C_B. The measurement of mass (Ci) used to characterize each of the centers in this study area, where the topographical and transportational conditions are favorable for the application of the method mentioned above, is an estimate of “basic” retail annual sale and employment possessed by it. Through this method, we can delimitate five sub-areas with Ichinomiya, Handa, Seto, Tsushima, and Komaki as the centers.
The delimitation of the sub-areas by means of the empirical method is based on several kinds of empirical data including daily commuting movement, medical service flow, consumer behavior, etc. These data, which show the spatial interactions between the dominant cities except for Nagoya and the smaller centers, enable us to obtain a delimitation of socio-economic areas and of administrative ones. As to the former, the study area is divided into six socioeconomic areas and five of the centers of these areas are identical with those derived through the theoretical method. As to the latter, we have twelve patterns, slightly different from one another, of administrative areas according to the difference of the data utilized for the delimitation.
When we compare the theoretical sub-areas on the basis of the retail sales with the actual sub-areas based on the empirical data concerning many kinds of spatial interaction, we may say that they coincide with each other to a relatively high degree. A more detailed comparison, however, reveals a slight discrepancy between the two patterns, especially in the viciniy of the metropolis Nagoya, which is excluded expediently from the study area in order to eliminate its influence over the wider area. This discrepancy might be attributed to the fact that the index of the centrality used in the theoretical method is limited to the economic function and that the non-central function of industrial or dormitory cities in the area concerned deviates the spatial interaction performed by central functions.
After all, in this study area, there are five metropolitan sub-areas with the centers situated almost at regular intervals

View full abstract