Ever since Wilhelm Max Wundt (1832-1920) first studied national character scientifically and published Elemente der Volkerpsychologie (Leipzig: Alfred Kroner, 1911-1920), the only systematic approaches to the study appeared in nonacademic publications and they could rarely be considered scientific by contemporary standards, being content to compare percentages, for example. Indeed, scholars intentionally avoided the term "national character." In the 1950s, Alex Inkeles began studying national character and using the term. The Institute of Statistical Mathematics started quantitative survey studies of the Japanese national character in 1953 (Research Committee on Study of the Japanese National Character, 1992).
Two major foundations of the study of national character are its interchronological and international comparative aspects (Hayashi, 1992e; 1993a; C. Hayashi and F. Hayashi, 1995). Turning to the first of these two elements, the interchronological aspect, enables clarification of the enduring and changing features of national character. Here, time-series data are indispensable, and sampling and questionnaires must be strictly comparable. That is, one must use the identical sample design and identical data collection methods.
In market research, the panel survey, with its longitudinal aspect, works well; however, this method is not practical for attitude surveys because (a) finding the same respondents after more than five years is very difficult; (b) administering an attitude survey yields aftereffects because it does not measure actual states of affairs; and (c) problems of privacy arise. For attitude surveys, then, the cross-sectional method is adopted as a better one. As for the questionnaire, one must use identical question and response wordings over time. (We will discuss problems specific to the questionnaire itself later in this paper.) Above all else, strict quality control in the continuing survey is paramount.
Turning to the international comparative aspect, the common and differing features of national character can only be made clear by comparing nations. Here, too, there are many methodological problems, which we will discuss in turn. Interestingly, apart from interpretive and speculative studies based on existing literature and observation, one finds few quantitative comparative studies. In earlier work, the quantitative studies simply compared data in surveys of peoples from various cultural spheres. A few careful researchers did heed the importance of comparability in translated questionnaires.
Other researchers have given fundamental consideration to cross-societal and cross-cultural studies. On theoretical and methodological studies, one can find useful discussions in Berting (1988), Inkeles and Levinson (1969), Inkeles (1991), Scheuch (1968; 1970; 1989; 1990), Szalai and Scheuch (1972), Szalai and Petrella (1977), and Research Committee (1991). Recently, several scholars have published interesting quantitative comparative studies mainly based on hypothesis-testing and theory construction (Inglehart, 1977; 1990; Schuman, 1989; Stoetzel, 1983).
My approach to the quantitative study of national character is different from what it used to be. In this paper, I will describe this new approach, as well as the underlying methodology, aided by heuristic examples, with an emphasis on the Japanese national character.
Fundamental Scheme of Study
What is National Character?
I define national character operationally as a collective character encompassing belief systems, ways of thinking, emotional attitudes, feelings and sentiments. Individuals receive various stimuli from the outside. By surveying individuals, we identify individual response patterns, which integrate into a collective through social communication. Thus a collective or national character, and in some cases an ethnic character, forms beyond individuals. To put it another way, a social environment emerges; a cultural climate is created, all out of a merging and commingling of social norms, customs, paradigms, education, contemporary thought and art, religious feelings and so on. …