Nanjun Zhou’s “Making Expressionism” reconstructs the social and material networks behind German Expressionist print culture. Drawing on a large dataset of prints from auction houses and museum collections, the study employs social network analysis to map collaborations among artists, printers, and publishers. An object-based approach foregrounds the material processes of production and circulation rather than only interpersonal ties. The paper shows how print techniques, publication venues, and geographic clusters shaped artistic identity and public expectations, underscoring the role of media infrastructures in defining modern visual culture.

Razvan Dumitru’s study of nineteenth-century textile circulation examines how global trade networks reshaped rural taste and cultural identity in Europe. Combining archival research with visual recognition tools, including convolutional neural networks, the paper analyzes extensive corpora of textile imagery to identify recurring motifs and cross-regional influences. This computational approach reveals patterns of exchange difficult to detect through conventional methods. The paper argues that industrial textiles functioned not only as commodities but as agents in the formation of local identities and nation-building projects, linking economic transformation to visual culture.

Lin Du’s “Digital Historical Forensics” develops a methodological framework for analyzing wartime media cultures during the Second Sino-Japanese War and the early People’s Republic of China. Integrating close reading with computer vision and machine learning, the study examines how photojournalism served as a strategic instrument for both the Chinese Communist Party and Japanese forces. Computational analysis exposes patterns of representation and ideological framing that extend beyond the scale of traditional qualitative analysis. The paper proposes digital historical forensics as a model for studying propaganda and visual culture in contexts of political conflict.

Fei Guo’s paper, “Humanistic and Historical Inquiry in the Age of AI,” situates current debates about generative AI within a longer technological and intellectual lineage. Rather than approaching AI solely as a disruptive force, the paper proposes a historically grounded framework informed by the bias-variance paradigm in machine learning. It conceptualizes AI systems as socially embedded and evolving forms of modeling. By placing AI’s inferential logic alongside historical reasoning, the paper reframes AI as both an object of historical study and a lens for reconsidering how historians generalize from evidence and construct interpretation.

Collectively, the session highlights two shared concerns: the material and infrastructural dimensions of cultural production, and the epistemological implications of modeling and pattern detection. By bridging theoretical reflection and empirical case studies across regions and periods, the panel advances discussion about the future of historical practice in an increasingly digital world.