Cultural Manifold Analysis is a statistical method grown into a subdiscipline of mathematical statistics, and Ryozo Yoshino is one of its most prominent representatives. In this book he explains its philosophy and the role it played and continues to play in Japanese social survey research carried out by anthropologists, psychologists, and sociologists. He begins by retracing the history of the method and how the Japanese state was involved in its genesis. The Institute of Statistical Mathematics (ISM) where it originated was established under the control of the Ministry of Education in 1944 and was intended to assist the war efforts.
After the war, Japanese researchers, as Yoshino puts it, “took off their military uniforms” and under guidance of the US-led occupation developed instruments for statistical marketing surveys and “a scientific opinion poll system for democracy” (p. 3). This is a point of major interest. Making scientific research subservient to war is something many scientists find objectionable, even unethical. However, to assume that the world will be all right again once they have taken off their military uniform would be naïve.
The first post-war survey with nationwide statistical random sampling, a literacy survey executed in 1948, provides good evidence for the risk of researcher bias. Assessing the literacy rate of the Japanese was deemed necessary by some members of the occupation’s Civic Information and Education Agency who were convinced that the relatively complex Japanese writing system was incompatible with democracy and that, therefore, a writing reform, i.e. romanization, was in order. Yet, the survey results were taken to indicate that the Japanese people commanded literacy levels that were sufficient to develop democracy. Thus, no writing reform was necessary, and that was that.
In retrospect, it is not the outcome which is of interest about this project, but rather the fact, which is still relevant for us today, that the investigator’s point of view plays a decisive role in social science research even where quantitative methods are applied that rely on seemingly objective numbers only. In the event, the main question underlying the survey was as absurd as the “result”, for there was not the faintest shadow of evidence that any relationship existed between forms of government and kinds of writing systems, or literacy levels. (Just think of the celebrated “largest democracy in the world” and India’s notoriously low literacy rates.)
After this dubious start, the ISM embarked on a long and laborious journey to develop highly sophisticated tools of multi-layered statistical random sampling for investigating Japan’s “national character”. Aware of the problematic nature of this notion, Yoshino calls it a “nickname” for the ambitious endeavour to uncover patterns of people’s consciousness [Shakai-ishki ni kansuru chōsa]. On this basis, the Japanese National Character Survey (JNCS) came into existence in 1953 and has been carried out every five years since. It has examined people’s attitudes and preferences concerning life, health, self, children, family, wealth, love, work, state, society, trust of government and legal system, among others. Along the way, new items and questions on various topics were added, as one might expect, but despite changes in format and contents, it evolved into a longitudinal survey. Documented on the ISM’s homepage,Footnote1 the wealth of data collected over seven decades are a treasure trove for empirical research about Japanese society and beyond.
As Yoshino shows in this book, the JNCS data is instructive both with regard to studying elaborate quantitative methods and comparing Japan with other societies. International comparisons are always difficult, not least because of the researcher standpoint-problem mentioned above; however, reliable survey data gathered by means of clearly indicated and transparent methods are a most valuable first step.
International comparisons can be comprehensive, such as the World Value Survey, or more limited in scope. For instance, within the framework of JNCS, the salience and importance of Confucian teachings in Japan has been compared with a number of relevant populations in China, Taiwan, and Singapore. Longitudinal data enable, or at least invite, interesting considerations about societal change in time, for example, how social variabilities, such as life expectancy, fertility rates and other demographic dynamics interact with ideological precepts and beliefs, e.g. that the eldest son should look after his ageing parents.
The difficulties of cross-national and longitudinal comparability have not been entirely resolved. For “people’s values about life, religion, leadership, nature, etc. are not independent, but intricately mixed with each other. It is necessary to deepen mutual understanding of the world by taking a cross-national comparative approach rather than simply comparing survey data superficially” (p. 163f.). Here one can only agree with Yoshino, and it is encouraging to meet in this book a number cruncher who is keenly aware of these problems and has consistently dedicated his work to the goal of combining quantitative and qualitative methods of social research with one another instead of making them mutually exclusive. This is particularly important because, as Yoshino points out repeatedly, the research at the ISM in Tokyo portrayed in this book has never lost sight of its applications for social planning and evidence-based policymaking.
In sum, for readers involved in social science research about Japan this is an instructive and readable book. An index would have been helpful, and the book’s English makes no pretence to be anywhere near perfect, but that may be an issue more for the publisher than the author.
Additional information
Notes on contributors
Florian Coulmas
Florian Coulmas is Senior Professor of Japanese Society and Sociolinguistics at the Institute of East Asian Studies (IN-EAST), University of Duisburg-Essen. He has published widely about Japanese society and culture, as well as other societal issues. His most recent books are Identity. A very short introduction, Oxford University Press, 2019, and Das Zeitalter der Identität. Zur Kritik eines Schlüsselbegriffs unserer Zeit, 2020, Universitätsverlag Wintrer Heidelberg.