抄録
1. Noticing the fact that the same design appears quite different to the different persons at different times and places, I searched for the statistical regularity in this phenomenon. 2. Making on purpose six kinds of spoons by combining parts of three kinds of spoons so as to have common characteristics of shape partially, I carried on an experiment in order to test the similarity and difference of those spoons. One hundred subjects were chosen for the experiment from the university students in Tokyo. The result shows that there is much difference in the judgment of similarity and difference of shapes among the subjects and even the same person changed his judgement looking at a pair of design at different times. The average frequency of different judgments comes to 3.1 per 10 times. Though this change of judgements is ordinarily considered as clue to the observation error, when I investigated correlation of each test I found that there were little or no cases where correlation coefficient is 0 or nearly 0. In most cases there were observed positive or negative correlation. The fact shows that each subject does not always judge a form synthetically from a fixed point of view but particular parts or particular characteristics of the form from various view points. 3. In addition to this, I examined the consumers' estimate and recognition of the characteristics of designs by Semantic Differencial method, taking eight kinds of designs of spoons and forks on the market. The subjects consisted of one hundred persons of the Industrial Art Institute and about seventy consumers living in Tokyo. The result shows that there are remarkable differences among the distributions of estimates of designs and the variance of a distributions is large in the case of new and strange design, while the variance of a distribution is small in the case of familiar design. 4. The reason why I put the subtitle "Statistical Approach to Morphology" to this study is that, apart from the traditional morphology based on unchangeable objective forms, I consider this study as an initial step to the attempt to characterize forms with the patterns of distributions of observers' responses to them.
