This repository contains three datasets developed during doctoral research on networks, cultural circulation, and material poetics in twentieth-century art.
The datasets reconstruct different types of relationships in modern art:
- the social network surrounding Marcel Duchamp
- the cultural visibility of individuals within that network using Google Books N-Grams
- the poetic and material structure of Joseph Cornell’s artworks through recurring motifs
All datasets are released for research and educational purposes.
The repository contains three independent datasets.
Each dataset is stored in its own folder and includes its own README file describing the data and methodology.
Each dataset can be used independently.
This dataset reconstructs the ego-centered social network of Marcel Duchamp using documented interactions found in primary historical sources.
The network contains:
- 436 nodes
- 5226 documented interactions
The dataset includes three tables:
- a relationship list documenting interactions between individuals
- a node attribute table with biographical information
- a primary document table describing the archival sources from which relationships were extracted
The data captures the network as documented in Duchamp-related archives and publications, not an exhaustive reconstruction of all historical relationships.
This dataset compiles Google Books N-Gram frequencies for the 435 individuals belonging to Duchamp’s social network.
The dataset measures how often each person's name appears in the Google Books corpus across time.
Data were collected for 13 decades, from 1899 to 2019, producing:
- 5642 observations
The dataset allows researchers to investigate:
- cultural visibility
- historical attention
- long-term fluctuations in the presence of individuals in published books.
The dataset is distributed as an Excel file (.xlsx) containing multiple sheets with documentation and data.
This dataset reconstructs the poetic structure of Joseph Cornell’s artworks through a network of recurring motifs.
The dataset contains:
- 166 artworks
- 29 poetic motifs
Each artwork was coded according to whether it contains specific visual or material elements such as:
- spheres
- birds
- glass objects
- drawers
- astronomical tables
- shells
- wooden fragments
- and other motifs recurring in Cornell’s work.
The resulting matrix allows the construction of a network in which artworks are connected through shared motifs, enabling analysis of:
- recurring poetic structures
- clusters of artworks
- the evolution of Cornell’s visual language across decades.
Each dataset in this repository has its own recommended citation format.
Please consult the README file inside each dataset folder for the appropriate citation.
The datasets are released under a permissive license.
They may be freely used, shared, and adapted provided that the original authorship is properly cited.
Felipe Eduardo Lázaro Braga
Research supported by FAPESP – Fundação de Amparo à Pesquisa do Estado de São Paulo
Grant nº 2021/00645-8