New Foundations in Digital Chinese Studies
On 20-25 June 2022, I had the privilege to participate in the European Summer School in Chinese Digital Humanities followed by the international conference “New Foundations in Digital Chinese Studies”. This major event in our field was hosted by ENP-China ERC project and sponsored by the European Association for Chinese Studies.
My contribution to the Summer School and Conference was threefold:
1. My keynote talk “Alice in Dataland: Stepping Through the Digital Mirror” aimed to deconstruct some common misconceptions regarding Digital Humanities in general, and Digital Chinese Studies in particular.
2. During the four-day summer school, I mentored several workshops and training sessions addressing various methods, including network analysis, geographical information system and geospatial analyses, as well as data mining and computational text analysis with HistText.
3. Finally, the presentation which I gave at the Conference “Mapping the Transnational Public Sphere in Modern China: A Bilingual Topic Modeling of the Republican press (1919-1949)” sought to investigate the formation of a transnational public sphere in China through the joint empirical study of two key institutions – a transnational organization (the Rotary Club) and the Shanghai periodical press, which has been recognized as the first mass medium for disseminating information in modern China. Methodologically, I relied on structural topic modeling (STM) to harness large corpora of digitized newspapers from a bilingual (Chinese/English) and dynamic perspective. We show that STM has the potential to encompass both the discursive and social aspects of the public sphere and to bridge the gap between previous theoretical and empirical studies. Thanks to STM, we are able to precisely classify and quantify the four main domains of the public sphere (meeting, self-organization, forum, community service) and to model their local and global dimensions. Furthermore, we reveal significant convergences between languages and a growing awareness of global issues during the troubled decades of the Republic (1919-1949). Methodologically, the contribution of this paper is threefold. First, we offer a simple yet efficient solution, based on concordance, to the problem of article segmentation in digitized newspapers. Second, this paper constitutes a rare instance of cross-lingual comparison involving a low-resource language, namely, the Chinese language during its transitional stage between classical and modern Chinese. Third, we demonstrate the value of using a multi-model approach to navigate between different levels of granularity, especially when dealing with corpora of different sizes and structures.
This paper will be published in the special issue of which I will be the editor in the Journal of Digital History in September 2023. The data and the full code supporting this research are available here.