Advancements in machine learning are opening new avenues for the analysis of extensive visual data, particularly within the field of historical political economy. According to Valentine Figuroa from MIT, artworks from museums and private collections hold valuable insights that have yet to be fully explored. To effectively apply computational methods to this rich data source, it is crucial to develop a framework for understanding the information encoded in these paintings and the assumptions that accompany their interpretation.
This article introduces such a framework, addressing key concerns of traditional humanities. Figuroa draws on a database of 25,000 European paintings spanning from 1000 CE to the First World War. The framework outlines three distinct applications that highlight different types of information conveyed through these artworks: depicted content, communicative intent, and incidental information. Each application also reflects a significant cultural transformation during the early-modern period.
Revisiting the Civilizing Process
The first application revisits the concept of a European “civilizing process.” This term describes the internalization of stricter behavioral norms that coincided with the expansion of state power. Figuroa examines paintings depicting meals to determine whether there is evidence of increasingly complex etiquette over time. By analyzing the visual representation of dining customs, this study seeks to understand how societal norms evolved alongside political authority.
Shifts in Elite Representation
The second application focuses on portraits to explore how political elites crafted their public personas. This analysis reveals a long-term transition from chivalric depictions of men to more rational-bureaucratic representations. By examining the evolution of portraiture, Figuroa illustrates how changing societal values influenced the way political figures presented themselves to the public.
The third application documents a sustained trend toward secularization, as indicated by the share of religious paintings in the overall dataset. This process began prior to the Reformation and escalated in the years that followed. By tracing the decline of religious themes in art, this analysis highlights the shifting cultural landscape and the move towards secular values in European society.
The integration of machine learning into art analysis represents a significant advancement in the field of historical political economy. By leveraging large-scale visual data, researchers can uncover new dimensions of cultural and political history that were previously overlooked. Figuroa’s framework not only enriches our understanding of art but also enhances our comprehension of the societal transformations that have shaped modern Europe.
This innovative approach underscores the potential for future research, as the intersection of technology and humanities continues to evolve. The findings from this study could inspire further investigations, revealing deeper connections between art, culture, and political history.
