[Special Feature: Yukichi Fukuzawa and Statistics] Yukichi Fukuzawa's Theory of Civilization and Theory of Statistics / Tadao Miyakawa

2020/06/05

Image: "An Outline of a Theory of Civilization" (Collection of the Keio University Fukuzawa Memorial Center for Modern Japanese Studies)

Everyone's life has infinite possibilities at the starting point, but which of them becomes reality depends on encounters with various people and books. This is the second time I, a non-Keio University alumni, have written for this magazine, and I am very happy and honored to be given this opportunity, as I have benefited from encounters with many Keio University alumni. My previous contribution was a book review of "Introduction to Econometrics" (Nippon Hyoron Sha, 1972) by my colleague Professor Keiichiro Obi.

The time when one decides which academic field to major in upon entering university is a major turning point, and I had an encounter there. When I was in my third year of junior high school (under the old system), I became interested in differential calculus after being introduced to it in a mathematics class. By chance, I came across "Principles of Economics for Beginners" (Keio Publishing, first edition 1946) by Shinzo Koizumi, who was the President of Keio University and also the tutor for the then Crown Prince Akihito. There, I learned Gossen's law of satiation of wants, and discovered that the concept of marginal utility in the Austrian School of Economics corresponds to differentiation, and that the law of diminishing marginal utility can be explained by the second derivative being negative. I was moved by the relationship between economics and mathematics.

After majoring in theoretical economics at university, I majored in econometrics, a fusion of economics and statistics, in graduate school. Then, at Harvard University in the United States, where I studied abroad, I received guidance from Professor Etaro Tsujimura, a brilliant young professor from the Juku, and in seminars by Professor Leontief on input-output analysis in the Littauer Building. In this way, statistics became one of my subjects at the university, but it was through my encounter with "Gakumon no susume (An Encouragement of Learning)," published between 1872 and 1876, and Fukuzawa's theory on the relationship between scholarship and politics in "An Outline of a Theory of Civilization" (hereinafter abbreviated as "Theory of Civilization"), published in 1875, that statistics became linked with Yukichi Fukuzawa. From then on, I became strongly interested in policy science in the 1970s. Fukuzawa's argument in "Gakumon no susume," famous for its opening line "Heaven does not create one person above another, nor one person below another," states that even among naturally equal human beings, the difference between becoming a noble or wealthy person versus a lowly or poor person, and between those who govern a civilized country and decide its direction (policy) versus others, arises from the presence or absence of learning. This felt like a highly persuasive argument emphasizing the importance of the academic field of policy science.

Regarding this policy science, Professor Hiroshi Kato, a flagship professor of the Juku, was a leader in Japan. As the president of the Japan Association for Planning and Public Administration, he awarded me the society's prize in 1993 for a paper on policy science I contributed to the society's journal. He also gave my book, "Foundations of Policy Science" (Toyo Keizai Shinpo-sha, 1994), the overly generous review of being a "monumental achievement" in Japanese policy management studies.

As an economist, the professor possessed broad and deep knowledge spanning politics, administration, and finance, as well as rare executive power, making significant contributions to national administration, such as his efforts in realizing the privatization of the Japanese National Railways into JR. He was also a promoter of the founding of the Shonan Fujisawa Campus (SFC) in 1990. He led the era by gathering diverse talent at SFC, including Professor Jun Murai, known as the father of the Japanese internet; Professor Heizo Takenaka, an economic policy expert who later played an active role in the Koizumi Cabinet; and Professor Shigeru Ito of Urban Engineering at the University of Tokyo (son of the writer Sei Ito). Serving as the first Dean of the Faculty of Policy Management, he was the spark for the policy school boom in Japan. Furthermore, he was studying at Harvard Business School in Boston at almost the same time as I was, and the famous political scientist Professor Masao Maruyama of the University of Tokyo was also staying at Harvard as a visiting scholar.

Professor Maruyama later published "Reading 'An Outline of a Theory of Civilization'" (Volumes 1-3, Iwanami Shinsho, 1986). In it, he notes that in the early Meiji period, statistics was generally understood as "expressing all sorts of domestic matters through arithmetic, and being most essential for governing the country and pacifying the people" (Preface by translator Rinsho Mitsukuri to Maurice Block's "Statistique"). Maruyama states that Fukuzawa was the first to grasp "statistics" as "a general method for the objective recognition of social laws, independent of policy theory." This was an important point related to the focus of the debate on the essence of statistics in the 1890s, which I will discuss later.

In 1875, when "Theory of Civilization," considered Yukichi Fukuzawa's masterpiece, was published, Fukuzawa's perception of Japan was as follows. Japan, which had opened its doors after the Edo Shogunate succumbed to pressure from major Western powers demanding the end of the national isolation policy, realized the principle that "those who lag behind in civilization are controlled by those who lead." It had to be said that Japan lagged behind the West in civilization. To ensure its independence, Japan had to understand what civilization is and the state of civilization in leading countries to avoid colonization. The strong enlightening power of Fukuzawa's arguments led to Japan's entry into the ranks of civilized nations from the Meiji era onward.

Fukuzawa first stated that "the theory of civilization is a discussion of the development of the human spirit," and that civilization means "making the human body comfortable and the mind noble," which ultimately is "the progress of human wisdom and virtue." He argued that when considering a nation's civilization, this wisdom and virtue should not be viewed in individuals, but in the nation as a whole. Therefore, one must consider the national wisdom and virtue, or the "total amount of wisdom and virtue distributed generally throughout the country." Since this total amount fluctuates according to changes in the popular mind that create the national spirit, the progress of civilization must be measured by those fluctuations. There are certain rules to such fluctuations in the popular mind, and according to the British historian of civilization Henry Buckle, it is truly remarkable that "if one views the popular mind of a nation as a single entity, there are fixed rules in its workings." Therefore, to discuss civilization, a "method of regarding the popular mind of the world as a single entity, comparing it broadly over long dimensions, and proving what appears in its traces" is necessary, and Fukuzawa's theory of statistics was that this method is statistics.

As an easy-to-understand example of this theory, Fukuzawa took the case of the stock volume of a steamed cake shop in Tokyo. He said that while you cannot tell if an individual will come to buy a cake just by looking at them, if you look at the people as a whole, there is always a fixed rule, and the confectioner considers this to manage stock surprisingly well. Translating this into a modern example, it fits perfectly with the case of inventory management for products like lunch boxes and milk in convenience stores. In my book ("Facing Risk with Statistics," Toyo Keizai Shinpo-sha, 2007), I introduced this as a story titled "Inventory Management of a Tokyo Confectioner," and it was adopted as an entrance exam question for the Faculty of Business and Commerce at the Juku (FY2010). This corresponds to the application of statistics as a "general method for the objective recognition of social laws, independent of policy theory," as Masao Maruyama said.

The provisional government established in Osaka after the pro-imperial, anti-foreigner faction of the Satsuma-Choshu alliance won the Boshin War summoned Fukuzawa, who possessed such high insight, along with Western scholars of the time, Takahira Kanda and Shunzan Yanagawa. However, as the son of Hyakusuke, a Nakatsu clansman who had received a stipend from the Shogunate despite being a low-ranking samurai, Fukuzawa harbored the feelings of a Shogunate vassal who could not bring himself to serve the political enemies who had overthrown the Shogunate. Furthermore, perhaps because his principles were incompatible with the anti-foreigner advocates who had pressured the Shogunate to maintain isolation, he refused, saying, "I absolutely cannot attend due to illness." He also did not respond to repeated summons after the new government moved to Edo. Yet, deeper than these feelings of Fukuzawa was his own powerful sense of mission based on his theory of civilization and theory of learning.

Fukuzawa argued that to advance Japan's lagging civilization, it was necessary to sweep away the spirit that had historically permeated the Japanese popular mind. This was a spirit of "tyrannical oppression by the government and servility and distrust among the people," following the proverb that "a harsh government rules over foolish people, and a good government rules over good people," depending on the virtue or lack thereof of the people. "Gakumon no susume" was written to reform this popular mind. However, the efforts of the government, whose outward appearance had been greatly renewed by the Restoration, were insufficient. It was difficult to reform even by government order or by the persuasion of private individuals. When considering what kind of person could fulfill this task, they were not to be found among the four classes of samurai, farmers, artisans, and merchants, nor among scholars of Japanese and Chinese classics; there were only "a type of Western scholar." Yet those Western scholars "knew the public but not the private... and almost all took government posts," failing to escape the bad habits of the Japanese and Chinese classical scholars. Those in the private sector doing such work were too few to count. This view of scholars held by Fukuzawa is based on his following theory on the origin of learning in "Theory of Civilization": "When learning arises after a period of chaos, in Western countries it arises among the general public, while in our Japan it arises within the government... Learning in Western countries is the business of scholars, and its practice exists solely in the world of scholars, regardless of whether it is public or private. Learning in our country is the so-called learning of the rulers, and is nothing more than a part of the government." "During the 250 years of Tokugawa rule, what were called schools in the country were either established by the central government or by the various domains." Based on such historical considerations, Fukuzawa discussed the advantages and disadvantages of government versus private institutions. Ultimately, as "our duty," he decided to "side with the private sector" through a livelihood as an "unfettered commoner, free and independent," making a living through writing and lecturing. This was the founding spirit of the private Keio University.

In contrast to Yukichi Fukuzawa's theory of statistics as a private citizen described above, Shigenobu Okuma, who entered the Meiji government as a judge of the Foreign Affairs Bureau from the Saga Domain in the first year of Meiji and rose to Minister of Finance in the 6th year, was a keen understander of the importance of statistics within the government. He was consistently involved in the government's statistical system from the earliest stages, to the point that he was later nicknamed the "Earl of Statistics." Furthermore, as they had long been on good terms as proponents of a British-style parliamentary cabinet system, the two hit it off. Fukuzawa supported the establishment of the Board of Statistics planned by Okuma by sending disciples from Keio University such as Fumio Yano, Tsuyoshi Inukai, and Yukio Ozaki. He also supported Okuma when he left the government in the Political Crisis of 1881 and founded the private Tokyo Senmon Gakko (the predecessor of Waseda University). The cooperation of these two great leaders of the Restoration period became a major force, and through the 1880s into the 1890s, the Meiji Constitution was enacted in 1889, and the National Diet was established in 1890.

It was in the 1890s, welcomed in this way, that two famous debates occurred that remain in the history of Japanese statistics. One was the 1889 debate over the translation of the word "statistics" between Rintaro Mori (Ogai), who had studied in Germany as a military doctor and medical scientist, and Takeo Imai, a disciple of Koji Sugi, who was a pioneer in introducing statistics to Japan from a state bureaucratic standpoint and one of the ten founding members of Meirokusha along with Fukuzawa. The other was the 1894 academic debate on statistics between Dr. Rikitaro Fujisawa, a professor of mathematics at Tokyo Imperial University, and Ayatoshi Kure, a disciple of Sugi who, alongside Sugi, was a pioneer of Japanese statistics and considered a leader particularly on the theoretical side. These two debates have been distinguished as a translation debate and an academic debate, but in substance, both were academic debates, particularly involving the important issue of analyzing causal relationships through statistics, which still retains its significance in modern statistics. As described in detail below, one can strongly feel the connection with the theory of statistics in Fukuzawa's "Theory of Civilization," which it is thought all the participants in the debate had read.

From "Facing Risk with Statistics [New Edition]" (Tadao Miyakawa)

First, regarding the translation debate, Mori argued against Imai, who defended Sugi. Sugi had strongly opposed using "統計" (tokei) as the translation for statistics and had even devised a composite character "[character omitted]". Mori contended that the word "tokei," which carries the meaning of "measuring things and governing them," was not unacceptable as a translation. However, by this time, the word "tokei" was already widely used in relevant government departments. Imai himself recognized this and, satisfied with having shown loyalty to his mentor Sugi, did not cling to the translation issue. Instead, he concentrated his rebuttal on Mori's criticism of Sugi and others' theory, which held that statistics exists to prove the relationship between cause and effect and to know the laws of heaven (natural order), calling it like "climbing a tree to seek fish."

Looking at what Mori described in detail during the debate with Imai, although not explicitly stated, it seems likely that Mori had also read Fukuzawa's "Theory of Civilization" thoroughly. One piece of evidence is the following description in Chapter 4 of "Theory of Civilization": "There must always be a cause for the workings of things. These causes are distinguished into two types: immediate causes and remote causes. Immediate causes are easy to see, while remote causes are difficult to discern. Immediate causes are many, while remote causes are few." Fukuzawa says that the key to searching for causes is to process the immediate causes that often clutter and deceive the human eye to reach the remote cause ("discard the immediate cause and proceed to search for where the remote cause lies... encounter the true cause, and see the certain and immovable rules").

In response, Mori says, "What in all creation does not have a cause?" Those who say there is no cause simply do not know the cause, and those who say there is no effect do not know the effect; the statistical method assists scholars' research into things where the cause and effect are not yet clear. Here, Mori introduces his unique, eloquent but difficult theory of "Tokusei Tokki / Kakusei Kakki" (Essential Nature/Mechanism vs. Individual Nature/Mechanism). Mori's "Tokusei Tokki" corresponds to Fukuzawa's remote cause, and "Kakusei Kakki" to the immediate cause. Mori says, "The essential is constant, the individual is variable," and "One sees cause and effect in the essential, but not in the individual." He argues that various individual natures and mechanisms are "removed" through large-scale statistical observation, allowing the causal relationship based on essential nature and mechanism to appear. He then stated that it is not statistics, the science of method, but the independent sciences that confirm that causal relationship. The similarities between the theories of both parties mentioned above are essentially the same as the theories of subsequent representative Japanese social statisticians such as Professor Seiji Zaitsu of Kyoto University and Professor Hiromi Arisawa of the University of Tokyo. Furthermore, it is extremely easy to understand when considering the regression analysis using the least squares method in statistics that we use most often today (for details, see Miyakawa's "A History of Statistics in Japan," University of Tokyo Press, 2017).

Next, the second debate between Rikitaro Fujisawa and Ayatoshi Kure began when Fujisawa, in a casual lecture, stated that compared to the long history of his own specialty, mathematics, statistics—with its much shorter history—could not yet be called a science, saying "statistics is a method, not a science." In response, Kure, along with Toshiyasu Kawai, Masao Yokoyama, and other disciples of Sugi, raised their voices in criticism all at once. As a mathematician, Fujisawa was a pioneer of mathematical statistics who emphasized the fundamental importance of large-number observation and the law of large numbers, arguing that mathematical thinking was indispensable for being a statistician. In contrast, Kure argued that simply being proficient in mathematics was not enough to handle "the facts of society as a whole, which are a complex combination of cause and effect," and that for a statistician, there are "subjects essential to a statistician," such as economics, politics, sociology, and ethics.

This academic debate between Fujisawa and Kure has continued as a parallel relationship between the social and mathematical systems in the history of the development of statistics in Japan thereafter. Today, the mathematical system is so overemphasized that many people think of statistics as a part of mathematics. However, I evaluate the significance of statistics as a method for capturing the regularity in human and social structures and behaviors, and I wish to adopt Kure's argument, returning to the starting point of Fukuzawa's thought, who said, "A person without the thought of statistics as a whole is not worthy of speaking about civilization."

次に第２の論争である藤沢利喜太郎と呉文聰との論争は、藤沢が気軽な講演会で自分の専門である数学のような長い歴史に比べるとずっと歴史の短い統計はまだ学問とは言えず、「統計は方法にして科学にあらず」と言ったことに対して呉をはじめ河合利安、横山雅男ら杉門下の人たちが一斉に批判の声をあげたことに端を発したものだった。藤沢は数学者として大数観察、大数の法則の基本的重要性を強調し、数理思想が統計家であるために必要不可欠であるとした数理統計学の先導者だったが、それに対して、数学に習熟しているだけでは「原因結果の複雑な結合である社会全般の事実」をとり扱うことはできず、統計家には経済学、政治学、社会学、倫理学など「統計家には統計家に必須の学科」があると呉は主張した。

この藤沢・呉の学問論争はその後の日本の統計学の発展の歴史における社会系と数理系との並立関係として続いており、今日では統計学というと数学の一部と考える人も多いくらい数理系が偏重されているが、私は人間や社会の構造や行動における定則性をとらえる方法としての統計学の意義を評価し、「統計全体の思想なき人は共に文明の事を語るに足らざるなり」と言った福澤の思想の原点に立ち返って呉の論を採りたいと思う。

The measures currently being taken by governments around the world as countermeasures to the corona crisis, such as lockdowns, bans on gatherings, and movement restrictions, significantly constrain "jinkan kosai (society)." For Fukuzawa, this might be considered a crisis of global civilization itself. Furthermore, Ogai Mori said, "Public health is a major objective of the government; the people have the right to demand of the government, 'Make us healthy,' and the government has the responsibility to vow to the people, 'We shall make you healthy.'" He would likely discuss the corona crisis as a crisis of public health, which is a major role of the government.

Until just before the corona crisis, there were even whispers of a major financial and economic shock exceeding the Lehman Shock due to concerns about the precariousness of the economic upturn caused by global hyper-monetary easing. However, what appeared was not a mere financial and economic shock, but a global complex shock to which problems of civilization, society, and public health were added—truly a crisis of the arena of "jinkan kosai (society)" as Fukuzawa called it. To use a play on words, it is a crisis involving the Wealth of Nations (Adam Smith) and the Health of Nations. Fukuzawa, who said he would remain a private citizen, was a person who worried about the independence of the country and preached the promotion of civilization. The great writer Ogai Mori, who was also a medical scientist, was interested in public health, which has a strong political nature within medical sciences. Therefore, they would surely have had statements worth listening to regarding the current corona crisis.

I would like to add one more thing here. The academic achievements of the late Professor Emeritus Akira Hayami, a proud member of the Juku, founder of historical demography in our country, and recipient of the Order of Culture, are also closely related to the history of Japanese statistics to which Fukuzawa contributed. His major work, "The Spanish Influenza that Hit Japan: The First World War between Humanity and Viruses" (Fujiwara Shoten, 2006), appealed that we have learned nothing from its lessons. It is deeply regrettable that we can no longer obtain comments from him, as he passed away on December 4 last year.

Finally, in connection with the corona crisis, I would like all Keio University alumni and readers to recognize once again the greatness of Shibasaburo Kitasato, who was involved in the founding of the Juku's School of Medicine in response to the favors received from Fukuzawa. While studying at the fledgling University of Tokyo School of Medicine, Kitasato became convinced that "the mission of a doctor lies in preventing disease," decided to make preventive medicine his life's work, and joined the Sanitary Bureau of the Ministry of Home Affairs (under Officer Nagayo Sensai). From 1886 to 1891, he studied under Robert Koch, a world authority on experimental medicine, at the University of Berlin, achieving great feats such as the successful pure cultivation of the tetanus bacillus. However, when he was a student, Professor Masanori Ogata of the University of Tokyo (then a lecturer), who had guided him in handling bacteria, announced the discovery of the "beriberi bacillus." Kitasato conducted experiments on that beriberi bacillus during his studies abroad and published the conclusion that it was unrelated to beriberi. This caused a storm of criticism at the University of Tokyo, labeling him an "ungrateful person." Upon returning from his studies, Kitasato was not accepted as a researcher. Officer Nagayo mediated and sought help from Fukuzawa. With support from Ichizaemon Morimura as well, Fukuzawa established Japan's first Institute for Study of Infectious Diseases. Yet Kitasato was by no means an "ungrateful person"; he later served as the first Dean of the Juku's School of Medicine, accepting the post without salary. How does the current corona crisis, where public health has become the greatest focus, appear to Kitasato, who will appear on the front of the next new 1,000-yen bill, replacing Hideyo Noguchi, following Eiichi Shibusawa who replaces Fukuzawa on the current 10,000-yen bill?

さらにここでもう１つつけ加えておきたい。塾の誇るべき、わが国歴史人口学の創始者であり文化勲章受章者の故速水融名誉教授の学問的業績も福澤が貢献した日本の統計学の歴史に密接に関係したものであるが、彼の大著『日本を襲ったスペイン・インフルエンザ ── 人類とウイルスの第一次世界戦争』（藤原書店、平成18年）は、その教訓をわれわれは何も学んでこなかったと訴えているものであった。昨年12月４日に他界されてしまった同氏のコメントも得られなくなってしまったのは残念でならない。

最後に、コロナ危機にも関連して塾員、読者の皆さんに改めて認識して頂きたいのは福澤から受けた恩顧に応えて塾医学部の創設に関わった北里柴三郎の偉大さである。北里は草創期の東大医学部の在学中に「医者の使命は病気を予防することにある」と確信し、予防医学を生涯の仕事とすると決心し、内務省衛生局（長与専斎局長）に入局した。そして明治19年から24年までベルリン大学で実験医学の世界的権威のロベルト・コッホに師事し、破傷風菌の純粋培養に成功するなど大きな業績をあげた。しかし、彼は学生の時、細菌の扱い方など指導を受けた東大の緒方正規教授（当時は講師）が「脚気菌」を発見したと発表したのに対して、留学中にその脚気菌についての実験を行い、脚気とは無関係という結論を発表した。これを東大では「忘恩の輩」として非難の嵐が吹き荒れ、留学から帰った北里は研究者として受け入れられず、長与局長が仲介して福澤に助けを求め、福澤が森村市左衛門の支援も受け、日本初の伝染病研究所が設立された。しかし北里は決して「忘恩の輩」ではなく、後に塾医学部の初代学部長として無給でその職務を引き受けていたのである。公衆衛生が最大の焦点となっている現在のコロナ危機は、福澤の現１万円札の後の渋沢栄一、そして野口英世に替わって次の新1000円札の表面に登場する北里にとってどのように映っているのであろうか。