Abstract
Analysis is made on the frequencies of the appearance of ten million key words of 1.2 million chemical articles and of 3.6 thousand terms of the typical texts of the Muromachi Renga. The former represents the literature resource of the modern science and the latter that of the classical Japanese culture. The analysis has evidenced that a logarithmic relation holds between the frequencies of the appearance of the top terms and the order of their magnitudes.
By the nature of the logarithmic relation, the top few terms of higher frequencies dominate the total appearance of all other terms, implying that those few terms represet the whole system. The top few terms could be referred to as those of the general nature and the bottom ones as those specific. After we classify the terms into two types of the general and the specific natures, we will be able to evaluate the character of any document by counting the relative ratio of the numbers of the general terms to the specific ones. By definition, any document dominated by the terms of the general nature could be referred to as the 'soft' one.
As further analyses are made with respect to correlations among the terms of the top, and the bottom groups of the appearance, separately, another evidence of correlation has been noticed on both groups. This finding will refer to the fact that the knowledge forms another structure of the higher order. Analogy is referred to for the structures of the universe or those of the atoms.
