#  CBDB Annual Conference 2026: Research in Chinese Studies in the Age of AI 

 



The China Biographical Database (CBDB) project held its annual conference, “Research in Chinese Studies in the Age of AI,” on Friday, October 23, 2026, in the Belfer Case Study Room (S020), CGIS South, at Harvard University. Professor Peter Bol, Director of the China Biographical Database Project, and Dr. Tang, Kwok-Leong, Managing Director of the Harvard Digital China Initiative, opened the conference. Eleven CBDB colleagues then presented their current projects across three panels, each followed by an open discussion. Wenfei Wang, Cambridge Manager of the China Biographical Database Project, offered comments on the papers, and members of the wider Harvard community attended as well. The conference also received generous support from the Institute for Quantitative Social Science (IQSS).

The full program and the abstracts reproduced below are also available on [the conference page](https://cbdb.hsites.harvard.edu/cbdb-annual-conference-2026).

## Panel 1: Between Print Culture, Literature, and Politics

The first panel asked how printing, literary writing, and political conflict shaped the record of who mattered and why.

### Beyond Famous Names: The Rise of Compiler and Selector Credits in Ming Imprint Records

Mei, Yuting 梅愉婷 (CBDB visiting scholar from Zhejiang University)

 ![Mei Yuting speaking at the lectern beside a slide comparing her findings with Kai-wing Chow's, point by point, in a table.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_mei_yuting.jpg)

 

Kai-wing Chow used Du Xinfu’s 杜信孚 earlier catalogue *Mingdai banke zonglu* 明代版刻綜錄 to classify fifteen leading writers by role. He argued that late-Ming editing and commentary reshaped literary production and that publishers matched famous names to genres. Did this extend beyond them, and which roles drove it?

I turn Du’s later *Quan Ming fensheng fenxian keshu kao* 全明分省分縣刻書考 into imprint-level data linked to CBDB. Among 4,114 exactly dated records with Ming contributors, the share crediting a Ming person with an editorial role rises from 22.7% before 1522 to 27.2% during Jiajing, 30.0% during Longqing, and 35.1% during Wanli. Excluding Chow’s fifteen writers leaves the trajectory nearly unchanged. The rise comes overwhelmingly from compiler, editor, and selector credits; commentary remains near 2%. No comparable rise appears for pre-Ming works. It is clearest in private editions from Jiangsu and Zhejiang, and in commercial editions.

Three checks make an extraction artifact unlikely: role labels match extracted responsibility terms in at least 99.6% of rows in every period; contributors per record remain stable; and using contributor occurrences rather than records as the denominator produces the same trajectory. Yet this bibliography records fewer famous-name commentaries than Chow’s source. It can therefore compare timing and geography, but not absolute prevalence. The evidence does not show a rise of commentary. Instead, it reveals a redistribution of bibliographic credit toward compilers and selectors from Jiajing through Wanli. I next test this pattern against title pages, prefaces, and colophons.

### “The South Wants to Shackle the Officials from the North”: Court Politics, Commercial Print, and Local Society in the Southern Ming, 1644

Liu, Zaicheng 劉再騁 (CBDB visiting scholar from Fudan University)

 ![Liu Zaicheng reading from his paper at the lectern, with a slide showing a network graph of the three Southern Ming regimes.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_liu_zaicheng.jpg)

 

How did court strife in late imperial China shape opinion in commercial print, and how did print mediate between the reading public and officialdom? The short-lived Southern Ming offers a test case. With nearly the entire Beijing court captured by Li Zicheng 李自成 (1606–1645) and the north soon lost to the Qing, the Hongguang court in Nanjing was built by officials then in the south. Their colleagues who made their way south from Beijing (beilai zhuchen 北來諸臣) reached Nanjing too late for the founding, were seen as rivals for office, and were pushed aside. The court opened a “collaboration case” against suspected collaborators; publishers went further, branding nearly all northern arrivals as men without integrity; and print opinion echoed “punish the traitors” riots across Jiangnan. As one contemporary put it, “the South (the Nanjing court), wishing to proscribe the officials from the north, ... schemed to catch them all in one net,” while “the petty scholars of the print shops” took up the court’s line and “contrived far-fetched arguments” against them.

Earlier scholarship has read the lists, manifestos, and fiction of 1644 as sources; I treat them as participants, asking how riots, print, and officialdom acted on one another to criminalize the northern arrivals. Using CBDB, I have shown that the Hongguang court was strikingly regional in composition: its officials were overwhelmingly those who happened to be near Nanjing when Beijing fell. Building on this, I ask how far the rift between these southern officials and the northern arrivals overrode the social ties that had once linked them, and how this rift played out in commercial print and public opinion.

### How Did Women in Chinese Opera Enter the Historical Record?

Li, Xiaobei 李小貝 (CBDB visiting scholar from Beijing Union University)

 ![Li Xiaobei at the lectern beside a slide titled “From performing to creating — but only for the women,” with Yuan, Ming and Qing bars.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_li_xiaobei.jpg)

 

Drawing on 1,812 biographical entries in the *Zhongguo quxue dacidian* 中國曲學大辭典, this paper takes 101 women as its primary object of inquiry and uses 1,711 men as a comparative group. It examines why different categories of figures entered the historical record through different forms of information. The results show that women documented for the Yuan dynasty were predominantly performers, whereas the number identified as creators increased from the Ming dynasty onward. Creators were generally recorded through personal attributes such as names, dates of birth and death, native places, and authored works. Performers, by contrast, left historical traces mainly through events and relationships, including performances, lineages of training, theatrical troupes, and artistic evaluations. Comparison with male figures indicates that sparse coverage in standard biographical fields was associated primarily with performer status rather than gender alone. Gender differences nevertheless remained within the same occupational category: women creators were especially likely to have their identities established through kinship ties.

A database is therefore not a neutral container of historical information. Classification, field design, and the encoding of relationships redistribute the visibility of different figures, thereby reshaping research objects, forms of evidence, and the formulation of scholarly questions. Database construction should thus be understood not merely as technical preparation for research, but as an integral part of knowledge production in the study of Chinese opera.

### Recovering Poets from Anthologies: Expanding and Cross-Correcting CBDB with the *Wanqing yi shihui* 晚晴簃詩匯

Li, Meifang 李美芳 (CBDB visiting scholar from Xihua University)

 ![Li Meifang at the lectern beside her title slide, “Recovering Poets from Anthologies: Expanding and Cross-Correcting CBDB with the Wanqing yi shihui.”](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_li_meifang.jpg)

 

The *Wanqing yi shihui* 晚晴簃詩匯 (Wanqing Cottage Poetry Anthology), also known as *Qing shihui* 清詩匯, was compiled under the direction of Xu Shichang 徐世昌, who served as President of the Republic of China (1918–1922). It is a comprehensive anthology of Qing poetry, extending into the late Ming and early Republican periods, and collects works by 6,168 poets of diverse social backgrounds, with each poet’s works prefaced by a short biography. Drawing on large language models, including Claude, this study extracts structured biographical data from these biographies, matches these data against CBDB, and verifies the results through two rounds of manual review.

The study identifies 1,826 poets not recorded in CBDB, as well as 24,435 items of biographical information absent from CBDB, and reveals a number of errors in both CBDB and the anthology. These findings supplement and correct CBDB and provide important data for research on the geographical distribution of Qing poets and on literati networks. They are particularly valuable for the study of lesser-known poets from frontier regions and of women’s literary standing. The study further demonstrates the potential of artificial intelligence and digital humanities methods for research on historical Chinese texts.

## Panel 2: From a Local Perspective

The second panel turned to the provinces, using local gazetteers and regional bibliographies to ask how cultural production, kinship, and officialdom worked outside the capital.

### Mapping Jiangsu’s Cultural Ecology Based on *the Jiangsu yiwen zhi* 江蘇藝文志

Wang, Huimin 王慧敏 (CBDB visiting scholar from Nanjing Xiaozhuang University)

 ![Wang Huimin at the lectern beside a slide asking “Why start with Nanjing and Suzhou?” and tracing writing, editing, printing and later use.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_wang_huimin.jpg)

 

This research uses the 2019 revised edition of *the Jiangsu yiwen zhi* 江蘇藝文志, covering nearly 30,000 people and 85,000 works, and CBDB to examine how cultural production was organized and cultural resources accumulated across generations in Ming and Qing Jiangsu.

Starting with Nanjing and Suzhou, it considers Nanjing’s role as a Ming secondary capital and the Qing division of provincial administration between the cities. It asks how they attracted contributors, organized textual work, and maintained and reshaped their cultural connections. Their approximately 12,000 biographical entries and 37,000 bibliographical records show some editors participating in multiple works, overlapping groups of contributors, and later printing of earlier texts. The research tests whether examination degrees and official service are more closely linked to some bibliographic categories and roles than to others. It asks who connects different authors and texts through compilation and editing. It also examines how later editing and printing extend regional connections beyond the original authors’ contacts with their contemporaries.

Future research will systematically compare works in the same bibliographic categories across the two cities, focusing on their contributors and editorial activities in comparable periods. It will then gradually expand across Jiangsu to explore regional differences and connections.

### A Multidimensional Analysis Based on *Zhejiang tongzhi: Renwu zhuan* 浙江通志·人物傳

Li, Anrunze 黎安潤澤 (CBDB visiting scholar from Renmin University of China)

 ![Li Anrunze beside a slide titled “Six steps from the PDF to CBDB data,” listing pilot, digitisation, extraction, disambiguation, matching and proofreading.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_li_anrunze_1.jpg)

 

 ![Li Anrunze beside a slide on the source, Zhejiang tongzhi: Renwu zhuan, showing the fixed order in which each biographical entry is written.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_li_anrunze_2.jpg)

 

*Zhejiang tongzhi: Renwu zhuan* 浙江通志·人物傳, compiled in 2022, contains concise accounts of historical and contemporary figures associated with Zhejiang, from imperial times to the present, recording dates of birth and death, place of origin, social status, major life experiences, and interpersonal relationships. This project transforms these records into structured data compatible with CBDB, following its data structure and coding standards for persons, places, official appointments, entry into government, kinship and social relationships, social status, and writings. A human–AI collaborative workflow supports extraction, disambiguation, normalization, coding, and quality control, combining computational assistance with human verification and historical judgment.

The resulting dataset, linked with existing CBDB records, supports three lines of inquiry. First, using places of origin, residence, and office-holding as core variables, it examines dynastic changes in the distribution of talent and patterns of mobility. Second, it builds multilayered networks from kinship, scholarly, and social ties to trace the formation, persistence, and intermarriage of local prominent families. Third, it aggregates the biographical evaluations to identify the dominant evaluative categories and their variation across periods, regions, and social groups.

### Neo-Confucian Learning in Fujian: Networks and Genealogy in Li Qingfu’s 李清馥 *Minzhong lixue yuanyuan kao* 閩中理學淵源考

Wang, Rongjing 王榕靖 (CBDB visiting scholar from Zhejiang University)

 ![Wang Rongjing at the lectern beside her title slide on Neo-Confucian learning in Fujian, illustrated with an old map of the Fujian coast.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_wang_rongjing.jpg)

 

Compiled in the eighteenth century by Li Qingfu 李清馥 (1703–?), *Minzhong lixue yuanyuan kao* 閩中理學淵源考 (A Study of the Origins and Lineages of Lixue in Min) brings together approximately 1,300 scholars associated with Neo-Confucian learning in Fujian during the Song, Yuan, and Ming dynasties. Drawing on earlier works, it assembles a broad range of biographical material and offers a systematic retrospective account of this intellectual tradition. Li organizes these figures into schools through family ties, master-disciple relationships, and other connections, with kinship serving as a notable basis for grouping scholars.

This project extracts information on geographical affiliations, official careers, kinship, and social relationships for inclusion in CBDB. Using these data, it aims to examine the spatial distribution of scholars and their networks of association and intellectual transmission. It will compare the relationships recorded in the biographies with the arrangement of schools across the work, exploring where the two align and where they diverge. Through this comparison, the project seeks to understand how the compiler’s selection, classification, and organization of individuals shaped the genealogy of Neo-Confucian learning in Fujian and the historical narrative built around it.

### Beyond the Center: Official Careers and Imperial Governance in Qing Guangxi

Dang, Qiong 黨瓊 (CBDB visiting scholar from Guangxi University)

 ![Dang Qiong at the lectern beside a slide on Cen Yidong and Tianzhou, asking how Qing military campaigns depended on local power in Guangxi.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_dang_qiong.jpg)

 

How did the Qing imperial bureaucracy operate beyond the political center? Situated on the empire’s southwestern frontier, Guangxi was characterized by ethnic diversity, the coexistence of different administrative institutions, and complex interactions between imperial authority and local societies, making it a valuable setting for examining regional governance.

Drawing on the *Guangxi tongzhi* 廣西通志, this study employs prosopographical analysis and digital humanities methods to examine patterns of official recruitment, appointments, promotions, and geographic mobility. It investigates how examination qualifications, native-place origins, and institutional arrangements shaped officials’ career trajectories and the circulation of administrative personnel.

This paper argues that official careers provide a critical perspective on the relationship between central bureaucratic institutions and local governance. By tracing patterns of appointment and mobility, it examines how personnel administration connected imperial authority with the diverse social and institutional structures of Guangxi. Moving beyond center-focused narratives, this study situates individual careers within broader processes of bureaucratic integration and territorial governance, contributing to scholarly debates on state formation, regional administration, and the exercise of imperial power in Qing China.

## Panel 3: Another Look at Historiography

The final panel returned to the historical record itself, asking how genealogies, standard histories, and administrative compilations distribute prestige and visibility.

### The Place of Prestige: The Geography and Social Capital of Choronyms in Tang China

Zhang, Zhongyang 張忠楊 (CBDB visiting scholar from Zhejiang University)

 ![Zhang Zhongyang reading at the lectern beside a slide on literary-type choronyms in Shandong, Jiangnan and the Jing-Xiang region.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_zhang_zhongyang.jpg)

 

How did choronyms (*junwang* 郡望), as state-certified markers of regional identity, function as social capital in Tang China? This study encodes the *Yuanhe xingzuan* 元和姓纂 (Genealogical Compilation of Surnames in the Yuanhe Reign) into a biographical dataset of 8,434 individuals from the Han to the late Tang and links it with Tang epitaph data. Drawing on GIS spatial analysis, a layered comparison of choronyms, time-varying-coefficient fixed-effects regressions, and correlation analysis at the choronym level, it shows that in genealogies, choronyms converged on the Chang’an–Luoyang corridor as political centers shifted, while new choronyms kept emerging through migration and office-holding. In epitaphs, by contrast, they converged on a few renowned old choronyms; detached from actual places, they became claims to ancient lineage.

Overall, compared with the examinations and political factions, choronyms conferred no significant advantage in office. They served mostly as prestige labels for those of modest official standing, while retaining a general advantage in the marriage market. Prestige-type choronyms, concentrated in the Chang’an–Luoyang corridor, sustained status through their location and marriage networks. Literary-type choronyms, scattered across Jiangnan, Hebei, and the Jing–Xiang region, rose through the examinations and literary achievement. The old Shandong choronyms combined both, converting their prestige into official advantage through the examinations and literature.

### Protagonists and Supporting Players: Two-Way Narrative Salience of Persons in the *Songshi jishi benmo* 宋史紀事本末 and the Standard Histories of the Song, Liao, Jin, and Yuan

Lin, Litao 林立濤 (CBDB visiting scholar from Nanjing University)

 ![Lin Litao standing before a slide titled “Person in Episode: Narrative Centrality,” showing a star-shaped network of figures around one person.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_lin_litao.jpg)

 

Chinese historians have long distinguished persons remembered through an event from events remembered through a person. The distinction concerns how histories divide narrative between persons and events, but it has remained impressionistic. *Songshi jishi benmo* 宋史紀事本末 (Historical Events in their Entirety in Song History) reorganizes Song history by event, and much of its wording corresponds to the biography-based standard histories of the Song, Liao, Jin, and Yuan; largely the same material thus survives under two organizing principles, allowing their effect on a figure’s prominence to be examined.

Locating passages shared by the two kinds of text, this study identifies which parts of each chapter correspond to these histories and finds the matching passages in individual biographies. It then measures narrative salience, the share of a text devoted to a person, in two directions: a person’s share of an event’s account, and an event’s share of that person’s biography. Crossing the two directions sorts figures into four types: central to both the event and their own biography; leading an event that weighs little in their biography; entering the record through an event in which they were minor; and peripheral in both. Figures prominent in events but without a biography or annal of their own in these four histories are also identified. The framework turns the two traditional relations into testable ones, making it possible to ask whether changing the organizing principle redistributes salience, whether the two relations exclude each other, and whether their distribution varies across polity, status, and biographical category.

### The Networks of the Officials under Tang Provincial Governors, and the Relationship between Regional Military Regimes and the Central Government

Nie, Wanchun 聶琬純 (CBDB visiting scholar from Renmin University of China), presenting online from China

This project involves the systematic compilation and organization of data on officials serving under Tang provincial governors, based primarily on *Tang fangzhen wenzhi liaozuo kao* 唐方鎮文職僚佐考. It also builds the dataset necessary to examine the career networks and marriage networks of these officials in order to reconstruct patterns of personnel mobility and social connection. Particular attention is given to the relationships among different regional military regimes, as well as to the varying degrees of political and institutional connection between these regional authorities and the Tang central government. Through this analysis, the project seeks to provide a clearer picture of the personnel structure and broader political networks surrounding Tang provincial administrations.

Professor Peter Bol gave the concluding remarks, and the conference closed with a group photo.

 ![Group photo of about twenty conference participants standing together in the Belfer Case Study Room beneath the conference title slide.](/sites/g/files/omnuum3101/files/2026-11/cbdb_conf2026_group_photo.jpg)

 

*Blog post by Mei, Yuting 梅愉婷 (CBDB visiting scholar from Zhejiang University). Edited by Wang, Wenfei 王文菲 (CBDB project manager).*