Chen Yunsong: Empowering Research on Chinese Civilization [1] through Digital-Intelligent Methods [2]
The Chinese nation possesses a long and continuous civilizational lineage spanning over five thousand years; fine traditional Chinese culture is the crystallization of wisdom and the very essence of Chinese civilization. General Secretary Xi Jinping attaches great importance to Chinese civilization and fine traditional Chinese culture, emphasizing the need to "promote the deepening of research into the history of Chinese civilization; encourage the entire Party and society to strengthen historical consciousness and solidify cultural confidence; and unswervingly follow the path of socialism with Chinese characteristics." He further noted, "Only by comprehensively and deeply understanding the history of Chinese civilization can we more effectively promote the creative transformation and innovative development of fine traditional Chinese culture, and more powerfully advance the construction of socialist culture with Chinese characteristics." At present, a new round of technological revolution and industrial transformation is developing in depth. To promote the creative transformation and innovative development of fine traditional Chinese culture, we must not only be adept at inheriting and developing the traditional methods of humanities research, but also be skilled in utilizing contemporary resources and research tools such as big data and artificial intelligence (AI). We must fully leverage the empowering role of digital-intelligence (数智) [1] technologies to deeply grasp the grand trajectory of the development of Chinese civilization, using quantitative means and scientific standards to deeply reveal the structural characteristics of Chinese civilization and interpret the five-thousand-year cultural pulse (文脉) [2] of the Chinese nation.
Scientifically Measuring Civilization and Culture
The investigation of civilization and culture is a systemic project that requires the organic integration of archaeological exploration and documentary research with natural science and technological means. General Secretary Xi Jinping pointed out: "We must strengthen overall planning and scientific layout; persist in multi-disciplinary, multi-perspective, multi-level, and all-round approaches; and foster close joint research between archaeology and history, and between the humanities and natural sciences. We should broaden the temporal and spatial scope and coverage of research to better answer major questions such as the basic picture, internal mechanisms, and regional evolution paths of the origin, formation, and development of Chinese civilization." How to scientifically measure cultural factors such as social consciousness and values has long been a difficult problem in the field of philosophy and social sciences. For example, regarding "soft indicators" like cultural influence, visibility, and reputation, even if methods such as quantitative research and questionnaires are used, the objectivity of the results can still be affected by the researcher's subjective bias. This not only imposes limitations on traditional research methods but also hinders an accurate understanding of the spatial-temporal characteristics of civilizational development and its status in the exchange and mutual learning between civilizations. With the development and application of a new generation of information technology, using methods like big data and machine learning, we can collect, analyze, and organize vast amounts of information—including book text data, global internet search data, news media data, social media text data, and global academic databases. We can then construct multi-dimensional measurement indicators for various cultural symbols—such as historical figures, architecture, and cultural products—in their global development and dissemination, thereby measuring cultural factors more scientifically and accurately.
Take book text data as an example. Books have been the most important carrier of human thought and cultural concepts for thousands of years. The higher the frequency with which a cultural symbol appears in a massive corpus of books, the higher its influence and visibility naturally are within that culture. With the help of book big data, we can perform multi-language word frequency searches and calculations for various Chinese cultural symbols, forming global cultural influence indicators for Chinese cultural heritage, classics, and famous figures over the centuries, allowing for scientific and precise comparisons with other cultural symbols. For instance, by analyzing big data from English-language books, we can generate precise word-frequency graphs to intuitively see whether the Great Wall of China or the Egyptian Pyramids has had greater cultural visibility in the English-speaking world over the last few centuries. Similarly, based on multimodal data such as images, audio, and video, we can establish various measurement indicators for the influence of cultural symbols based on information such as searches, listening, viewing, sharing, comments, and interactions. Who is more famous in the German-speaking world, Li Bai or Du Fu? Between Tang Xianzu and Shakespeare, who has greater influence in East Asian countries? Big data can provide new solutions to such previously unanswerable cultural measurement questions, providing a solid econometric basis for us to further solidify cultural confidence and promote exchange and mutual learning between civilizations.
Perspectivizing the Deep Structure of Civilization and Culture
Script and texts, as the core carriers of culture and civilization, carry rich historical information and cultural meaning. Since ancient times, the Chinese nation has attached great importance to preserving historical materials. From the Twenty-Four Histories [3] to poetry and prose, China's rich historical documents have ensured the continuity of Chinese culture. General Secretary Xi Jinping pointed out: "We must use modern technological means to strengthen the protection, restoration, and comprehensive utilization of ancient books and archives, and deeply mine the philosophical ideas, humanistic spirit, value concepts, and moral standards contained within them." Faced with a vast sea of texts, it is difficult for individual researchers alone to form comprehensive, precise, and "bird's-eye" observations and analyses. For example, there are approximately 50,000 extant Tang Dynasty poems, which only the most diligent scholars can read in full; the number of extant Qing Dynasty poems is even more massive and its stylistic categories more complex, with no universally accepted estimate of its total scale yet—one can only imagine its vastness. This means that despite our thousands of years of tradition in "poetic education" (shijiao), as individuals, it is difficult for us to read every poem from every dynasty in Chinese history. The sources of many theories and hypotheses in literary and poetic history can only rely on the scholar's intuitive perception and grasp of a portion of the works. In this sense, the introduction of big data and machine learning methods can help humanity form a more complete and accurate history of literature and culture.
For massive text big data—such as poetry, prose, and even novels—applying unsupervised learning clustering algorithm technology can quickly provide a global perspective on the distribution and structure of thematic content, making up for the limitations of traditional research methods. For example, the LDA (Latent Dirichlet Allocation) topic model, as a text mining tool, can discover latent thematic structures in text and is widely used in information retrieval, text classification, and sentiment analysis. Using this model, one can identify different genres and themes in the Complete Tang Poems, the Complete Qing Poems, or other massive literary collections. By further calculating and aggregating the character-attribute distribution of each work, every poem can be classified into identified genres—such as "frontier fortress poetry" (bianzhanshi) or "fields-and-gardens poetry" (tianyuanli)—and one can even plot the thematic evolution of poetry across historical periods. This method not only helps enhance the ability to read and understand individual texts but also provides an unprecedented tool for a macroscopic "bird's-eye" view of the temporal and generational shifts in cultural lineages. Furthermore, the introduction of deep learning models—especially the application of Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN)—makes the joint analysis of images and text possible. Taking literati painting (wenrenhua) [4] as an example, by analyzing the relationship between images and the poems inscribed upon them, deep learning models can reveal latent connections between the two, such as how imagery in the poem corresponds to visual elements in the painting. This analysis breaks through the limitations of single-mode text data, making the study of texts and images more three-dimensional and multi-dimensional, providing a brand-new perspective for researching traditional art forms. Additionally, Generative AI also demonstrates great potential in the analysis of massive textual content. AI with natural language processing and image generation functions can automatically analyze, generate, and summarize vast amounts of text data. Overall, empowering the study of civilization with digital-intelligence methods can visualize the deep structures within civilizational lineages—especially those internal characteristics that are difficult for individual researchers to discover using only the mind's capacity—thereby providing an important methodology for better understanding the evolution and structural features of the cultural pulse.
Capturing the Beauty of Cultural Heritage
Cultural heritage is a precious wealth left by our ancestors. General Secretary Xi Jinping pointed out: "Protecting and inheriting historical and cultural heritage is being responsible to history and to the people." A city's historical sites, cultural monuments, and humanistic foundations are part of the city's life and an important component of cultural heritage. If the cultural foundation is destroyed, no matter how new or good the city is built, it will lack vitality. During his inspection in Beijing, General Secretary Xi Jinping emphasized the need to "unify the renovation and upgrading of old urban areas with the protection of historical sites and the preservation of the historical cultural pulse, so as to both improve the living environment and protect the historical and cultural foundation, allowing historical culture and modern life to merge into one." During his inspection in Shanghai, he emphasized the need to "focus on inheriting the city's cultural pulse and strengthening the protection of cultural relics and heritage." Digital-intelligence technology provides an important data foundation and monitoring means for cultural heritage protection, enabling the depiction of a "grand vision" and "grand lineage" showing historical and cultural changes. Establishing a data system for comparing the past and present of the contours, textures, and styles of old urban areas is of great significance for protecting and promoting fine traditional Chinese culture and continuing the city's historical cultural pulse.
Traditional methods of cultural heritage protection rely on historical documents, field surveys, and limited site protection means. When faced with massive and complex cultural heritage data and historical evolution, it is difficult to accurately recreate the true appearance of an old city before large-scale urban redevelopment, and equally difficult to achieve precise monitoring and dynamic evaluation. Digital-intelligence technology provides a brand-new approach for data reconstruction in ancient city protection. For example, by using historical satellite imagery as big data, a two-dimensional historical scroll can be established; applying intelligent algorithms like machine learning allows for the analysis and integration of text, images, video, and geographic information in massive historical materials. Supplemented by digital modeling means such as LiDAR and drone scanning, and using high-precision Geographic Information Systems (GIS) and visual enhancement technologies, the distribution and evolution of relics such as ancient city walls, streets, and riverbeds can be presented in a virtual space. Nanjing University's "Digital Atlas for the Protection of Ancient Chinese Cities" has formed such an integrated data innovation. Furthermore, by introducing social development indicators and utilizing time-series and spatial analysis, digital-intelligence technology can study the laws governing changes in cultural heritage and the socio-cultural factors behind them. This helps cultural heritage protection transition from static recording to dynamic monitoring, tracking changes and risks in real-time, which not only improves the accuracy and efficiency of protection but also grants new vitality to cultural heritage.
Through the empowerment of digital-intelligence methods, the beauty of cultural heritage is presented in a three-dimensional and vivid way. Whether for large human settlements like walled cities and villages or other small-scale immovable cultural relics, this all-round, multi-dimensional presentation method helps break through the limitations of traditional static displays, making every relic and every site "come alive" with stronger narrative attributes. The beauty of cultural heritage will no longer be limited to static displays through text or images; instead, through more vivid presentation methods, it can help trace the development of civilization, show historical evolution, and integrate into social life, thereby reflecting the nation's spiritual essence more profoundly. This mode of presentation can promote cultural heritage as a classic symbol of fine traditional Chinese culture, strengthening its dissemination and turning it into a cultural bond that transcends time and space. Globally, the dissemination of cultural heritage promotes dialogue and understanding between different civilizations, providing a new perspective and impetus for global cultural exchange and mutual learning.
Upholding the Fundamentals and Breaking New Ground to Strengthen Civilizational Research
In advancing the construction of socialist culture with Chinese characteristics, upholding the fundamentals and breaking new ground (守正创新) is an important principle. At the Seminar on Cultural Inheritance and Development, General Secretary Xi Jinping emphasized that "only by upholding the fundamentals can we avoid losing our self and our direction; only by breaking new ground can we grasp and lead the times." At the 17th collective study session of the Political Bureau of the 20th CPC Central Committee, he emphasized the need to "anchor the strategic goal of building a culturally powerful country by 2035; persist in Marxism as the fundamental guiding ideology; take root in the broad and profound Chinese civilization; follow the trend of information technology development; and continuously develop New Era socialist culture with Chinese characteristics that possesses strong ideological leadership, spiritual cohesion, value appeal, and international influence." Empowering the study of Chinese civilization with digital-intelligence methods is a vivid practice of upholding the fundamentals and breaking new ground in cultural construction. Relying on multi-source big data and using AI as a method to perform an all-round "remote sensing" of the process of civilizational and cultural development—from the perspectives of temporal change, spatial difference, and spatial-temporal evolution—complements the traditional research models of philosophy and social sciences, much like the relationship between observing celestial phenomena from afar and inspecting objects up close, or like the relationship between a telescope and a microscope. This digital-intelligence-empowered method of civilizational research is often referred to as "Cultural Distant Reading."
Big data and AI provide "Cultural Distant Reading" with a broader vision and more precise analytical tools; the analysis of cultural phenomena and historical cultural heritage is no longer limited to traditional research based on the accumulation of local experience. By comprehensively utilizing multi-source, heterogeneous, and massive big data to deeply explore the deep structure of Chinese civilization, we can help reveal the influence and dissemination effects of the classic symbols of fine traditional Chinese culture on a global scale, and help correct the biases found in Western databases and Large Language Models. Through intelligent algorithmic analysis, we can present the distributional characteristics, evolutionary paths, dissemination effects, and influencing factors of fine traditional Chinese culture across different historical stages. This research method not only helps promote the inheritance and innovative development of Chinese civilization but also provides a new perspective for understanding the diversity and uniqueness of global cultures.
While recognizing the immense potential of "Cultural Distant Reading," we must also note its limitations, such as high technical thresholds and dependence on massive data sources and complex algorithms. It must still be combined with traditional cultural research methods to help researchers understand cultural meanings more comprehensively and deeply. Empowering the study of Chinese civilization with digital-intelligence methods can promote interdisciplinary development in this field, accelerate the construction of a Chinese independent knowledge system for philosophy and social sciences, and provide "Chinese ideas" for the cultural governance practices of other countries. At the same time, it helps enhance the discourse power of Chinese civilization within the global civilizational system, laying a foundational knowledge base for truly understanding the historical necessity, cultural connotation, and unique advantages of the Chinese path, thus opening a scientific path for consolidating cultural subjectivity and solidifying cultural confidence.
(The author is the Deputy Secretary of the CPC Committee of Nanjing University) Source: People’s Daily (February 6, 2025, Page 09) Web Editor: Hui Hui