Network Analysis in the Humanities
Letters between scholars, characters in a play, citations among texts — much of culture is relational. Network analysis turns these relations into graphs of nodes and edges, letting humanists measure structure, centrality, and connection in correspondence, narrative, and social worlds.
Definition
The modeling and analysis of humanities materials as networks of related entities — people, texts, characters, places — using graph-theoretic measures and visualization to study relational structure.
Scope
Covers the application of network theory and graph analysis to humanities materials: correspondence and social networks, character networks in narrative, and citation or influence networks. Includes the 'network turn' and methods for modeling, measuring, and visualizing relational structure, along with cautions about inferring meaning from graphs.
Core questions
- What cultural and historical phenomena are usefully modeled as networks?
- How do measures such as centrality illuminate texts and societies?
- What does a network visualization reveal, and what does it obscure?
- How should incomplete and uncertain relational data be handled?
Key concepts
- Node and edge
- Centrality
- Community detection
- Correspondence network
- Character network
Key theories
- Network theory of plot
- Moretti modeled the characters of dramatic and narrative works as networks, using their structure to reinterpret plot and the consequences of removing key figures.
- Quantitative analysis of historical networks
- Ahnert and Ahnert applied network measures to early modern correspondence to identify hidden coordinators and the structure of clandestine communities.
- The network turn
- Ahnert and colleagues argued that network thinking offers the humanities a general framework for relational analysis while warning against treating graphs as self-evident.
History
Borrowing from sociology and graph theory, humanists began applying network analysis to literature and history around 2010. Moretti's 2011 essay on plot networks and the Ahnerts' work on early modern letters were influential; The Network Turn (2021) synthesized the approach and its theoretical stakes.
Debates
- Explanatory power versus visual seduction
- Network diagrams are persuasive and can hide weak data or arbitrary choices, prompting debate over when graph measures genuinely explain and when they merely illustrate.
Key figures
- Franco Moretti
- Ruth Ahnert
- Sebastian Ahnert
- Scott Weingart
Related topics
Seminal works
- moretti2011
- ahnert2015
- ahnert2021
Frequently asked questions
- Does a striking network diagram prove anything?
- Not on its own. A graph reflects choices about what counts as a node and an edge and is only as good as its underlying data. Network measures can reveal real structure, but conclusions need careful validation; an impressive visualization can mislead as easily as it informs.