Reflections on the Use of Network Analysis and Text-Mining Tools in the Biographical Study of a Historical Figure
Total Page:16
File Type:pdf, Size:1020Kb
December 2016. Vol.5 / pp.55-85 Reflections on the Use of Network Analysis and Text-Mining Tools in the Biographical Study of a Historical Figure Anne S. CHAO*1 Abstract This paper is a preliminary study of the use of computational techniques on the study of a major historical figure. We demonstrate how the life and writings of Chen Duxiu (陳獨秀 , 1879-1942), founder of the Chinese Communist Party (CCP), can be studied with computational methodo- logy in two ways. First, a network visualization of Chen’s ties to the numerous associations in his pre-CCP days reveals Chen’s multiple connections to members of these networks. This is the first step to future analysis of how Chen’s ideas and activities were influenced by the cohorts of these networks, and thereby situating Chen in the political and intellectual historical contexts of the late Qing, early Republic period. Second, by text-mining with R we trace the evolution of some of the major concepts in Chen’s writing over time; in this paper we look at the concept of patriotism. We calculate the word frequency curve of Chen’s writing and place this curve against the natural language curve of Zipf’s law. In particular, we select words such as citizen, race, revolution, youth, democratic, and being patriotic, and call them “umbrella words.” His writing was divided into four periods, and in each period we find other high co-occurrence with these umbrella words, and designate these as “companion words.” We study the high correlation of the um- brella with the companion words, look at fifty words before and after each companion words, and parse the meaning of the umbrella word in * Anne Chao ([email protected]) is Adjunct Lecturer in the History Department at Rice University. 56 _ CONCEPTS AND CONTEXTS IN EAST ASIA (Vol.5 December 2016) each specific context. The shift in meaning of the concept of patriotism is derived as the overall context changed. Keywords Chen Duxiu, network visualization, patriotism, citizen, race, revolution, youth, democratic, being patriotic, text-mining, word frequency Introduction Upon the completion of my dissertation, I spent some time learning and experimenting with the use of computational technology in the study of history. In particular I wanted to know if adding a quantification method to the traditional production of literary text could enhance my biographical study of a major figure in modern Chinese history — Chen Duxiu (陳獨秀, 1879-1942). Chen was the founder of the Chinese Communist Party (CCP) and an intellectual whose iconoclastic battle with tradition helped bring about the New Culture Movement, ca. 1915-1923, in China. I wondered how, as a classically trained scholar and a young man from the hinterland province of Anhui, he became the prime mover of two salient movements in early twentieth century China: the New Culture Movement and the founding of a radical political party. In reading his writings spanning the years 1897 to 1942, I also noticed that the concepts of his youth took on different shades of meaning in his mature years, and these shifts were related somewhat to his political fortune over time. Computational analysis allowed me to answer both of these questions. For the first question: I believe that his accomplishments were shaped in large part by the networking with like-minded individuals. Throughout his life, especially in his formative years, Chen joined many student groups, revolutionary societies, and editorial boards. It was through his interaction with the people he met that encouraged, inspired and pushed him to ever more radical ways of thinking. Thus, the circles in which he moved offered valuable clues to the decisions he ultimately made in his life. I therefore used Gephi to plot his social networks from his trip to Tokyo (1901) up to the founding of the CCP (1920). I connected Chen to the organizations he joined and to other members of these organizations. Visualization clearly indicated points of overlap, not only where Chen overlapped with many Reflections on the Use of Network Analysis and Text-Mining Tools in the ... _ 57 members in multiple organizations, but also where other members overlapped with each other in these organizations. This allowed the detection of a cohort of like-minded rebels and gave us more information about the social forces of the period. For the second question, my colleague used the text-mining technique to help me detect the shift in meaning in Chen’s writing. Chen was a prolific writer throughout his life: the largest extant collection of his writing, Chen Duxiu Zhuzuo Xuanbian (陳獨秀著作選編, henceforth CDXZZXB), consisted of 891 pieces of essays, correspondences, poetry and philological analysis; and of these about 504 were related to politics and nationhood. Chen was a patriot, and he wrote and lectured in an effort to inspire his countrymen to save China. For this paper, we traced the evolution of a major concept in Chen’s writing — the concept of patriotism, over the period from 1897 to 1942. Using the computer language R, we generated a list of keywords and another list of words contextualizing the first list of keywords, and searched for differences in nuance. Assuredly, the detection of shifting meaning is a qualitative assessment that is best performed by human interpretation. The paper will consist of two parts: the first uses network visualization to trace Chen’s networks; and the second uses text-mining to track the evolution in Chen’s understanding of the concept of “patriotism.” The Social Networks of Chen Duxiu I compiled the names of the 340-plus members of the 21 organizations connected directly or indirectly to Chen (the indirect connection being one where Chen helped organize the group but did not join). The chronology spans the first two decades of the twentieth century, from Chen’s first trip to Japan in 1901 to the founding of the first cell of the CCP in 1920. Gephi generated a graph that visualized Chen’s entire network of members (Figure 1). I then filtered out most of those who had less than three ties with Chen (Figure 2). This approach involves several caveats: (a) All the ties are assumed to be of equal value, which is not true in life. (b) Those who share only one or two ties with Chen may not be less influential than those with more ties. (c) It is possible that Chen made close friends who never joined any of the associations that he was involved in. Proceeding with these precautions, the network analysis is nevertheless quite revealing: just by 58 _ CONCEPTS AND CONTEXTS IN EAST ASIA (Vol.5 December 2016) following the series of social organizations with which Chen chose to be involved, it is clear that his life took a “radical turn.” <Figure 1> Twenty-One Social Networks of Chen Duxiu Reflections on the Use of Network Analysis and Text-Mining Tools in the ... _ 59 <Figure 2> Filtering out Names with Less than Three Ties with Chen Duxiu Determination Society, 1900 Leftist Newspapers and Youth Society, early CCP Cell in Anhui Tokyo, 1902 China Education Society, Education Conference in Anhui, 1921 Shanghai, 1902-3 Pan Zanhua Shanghai Anhui Shanghai Assassination Reform Society, 1920 Wu Zhihui Zhang Ji Squad, Shanghai, 1904 Xie Wuliang Beijing Anhui Lin Xie Reform Society, 1920 Zhang Shizhao Tao Chengzhang Subao, CCP Shanghai Shanghai, 1903 Cell, 1920 Su Manshu Yu Ziyi Li Dazhao Cai Yuanpei Huang Xing Gao Yuhan Chen Duxiu Citizen's Daily, Shanghai, 1903 Hu Shi Meizhou Pinglun, Liu Shipei 1918-19 Zhang Binglin Zhao Sheng Anhui Patriotic Gao Yihan Society, 1903 Li Xinbai Anhui Vernacular Xin Qingnian, Paper, 1904-1905 1915-1926 Bo Wenwei Anhui Public Sun Yujun Jiayin, School, 1905 1914-15 Li Guangjiong Yue Fei Loyalist Anhui Provisional Society, 1905 Govt, 1911-12 Asia Friendship Society, Tokyo, 1907 Society to Study Socialism, Tokyo, 1907 60 _ CONCEPTS AND CONTEXTS IN EAST ASIA (Vol.5 December 2016) Figure 1 shows twenty-one of Chen’s personal networks clockwise in chronological order. It reveals lots of overlapping membership and the fact that many of these organizations are populated by the same group of people. However, it is hard to discover significant connections with Chen due to the number of lines. I filtered out members who had less than three ties with Chen, assuming that those who had multiple ties with Chen potentially enjoyed a more significant and long-lasting relationship with him, see Figure 2. Visualizing the data in this way has to be combined with careful analysis: although all ties appear equally weighted, in reality they are not. A quick glance reveals that Chen’s political, intellectual and social development can be roughly divided chronologically into four quadrants. These are categorized as: “revolutionary awakening: 1901-1904,” “provin- cial stirring and introduction to socialism: 1903-1907,” “leading the New Culture Movement: 1911-1919,” and “a Marxist turn and the founding of the CCP: 1920-1922.” The four quadrants are characterized by the following characteristics: Revolutionary Awakening: 1901-1904― The upper right quadrant plots his introduction to the politically liberal community of overseas students in Tokyo, his membership in the Determination Society (1901), and Youth Society (1902). It continues with his return to Shanghai and his introduction to the nationally prominent radical members of the China Education Society (1902-1903), and finally to his editorship of the radical Shanghai paper, Citizen’s Daily (1903). Provincial Stirring and Introduction to Socialism: 1903-1907― The lower right quadrant traces Chen’s return to his home province of Anhui and his introduction to socialism on his subsequent trip to Tokyo. This includes his membership in the Anhui Patriotic Society (1903), his editorship of the Anhui Vernacular Paper (1904-1905), his founding of the Yue Fei Loyalist Society (1905), his role in the Anhui Public School (1905), his joining of the Shanghai Assassination Squad (1904), his subsequent memberships in Society to Study Socialism (1907) and in the Asia Friendship Society (1907) respectively in Tokyo.