Narrative as data: Linguistic and statistical tools for the quantitative study of historical events

Abstract

This paper illustrates some linguistic and statistical tools that can be profitably used by historians and social historians in the study of events (such as strikes, demonstrations and other types of collective conflict). More specifically, the paper shows that "semantic grammars" provide rigorous tools for the collection of rich event narratives. Semantic grammars structure information around the "canonical form" of the language: noun phrase/verb phrase, or subject, action, object and their modifiers (e.g. time, space). The fact that semantic grammars can be easily implemented in a computer environment using relational database systems (RDBMS) makes feasible the practical application of such powerful coding schemes. The data that computer-based semantic grammars make available are richer, more flexible and more reliable than those delivered by more traditional content analysis methods. They are also very well suited for the application of new tools of data analysis such as network models. Both semantic grammars and network models are fundamentally concerned with actors and their actions, with agents and agency. As such, these linguistic and statistical tools should draw sociology closer to history, traditionally much more concerned with issues of agency. I illustrate the power of both the linguistic and statistical tools using data that I collected from some 15,000 newspaper articles on the 1919-1922 period of Italian history, a period characterized by widespread working-class mobilization (1919-1920, the "red years") and fascist counter-mobilization (1921-1922, the "black years"). Data collection for this project was supported, in part, by the National Science Foundation (grant SES-8511632). Data analysis and writing have benefited from another National Science Foundation grant (SBR-9411739). I am indebted to William Bainbridge and Andrea Kline for their help in transferring the NSF grant to the University of Oxford. I would also like to thank Larry Griffin for his comments on a previous draft and Mauro Giorgetti and PierPaolo Mudu for their help with data analysis.