Introduction to the Special Issue: The Statistical Analysis of Political Text

Text is arguably the most pervasive—and certainly the most persistent—artifact of political behavior. Extensive collections of texts with clearly recognizable political—as distinct from religious—content go back as far as 2500 BCE in the case of Mesopotamia and 1300 BCE for China, and 2400-year-old political discussions dating back to the likes of Plato, Aristotle, and Thucydides are common fare even in the introductory study of political thought. Political tracts were among the earliest productions following the introduction of low-cost printing in Europe—fueling more than a few revolutions and social upheavals—and continuous printed records of legislative debates, such as the British parliament’s Hansard and precursors tracing to 1802, cover centuries of political discussion. The possibility that the analysis of texts could provide insights into the political processes also has a long pedigree. The Italian humanist Lorenzo Valla’s careful philological analysis of the reputed Donation of Constantine in 1439 convincingly demonstrated, using purely textual methods, that the document was a medieval forgery that must have postdated the Emperor Constantine I by at least four centuries. Moving toward our own day, the first modern, theoretically driven content analysis project was Harold Lasswell’s Wartime Communications Project just prior to the outbreak of World War II (Janowitz 1968), and subsequently, content analysis became a standard analytical tool in the West, particularly for the analysis of ‘‘enemy’’ communications, first Nazi and later Communist. Given the tedious nature of human coding, researchers recognized early on that systematic textual analysis might be suited to automation. Harvard’s General Inquirer (Stone et al. 1966) emerged as the first widely used computer program for automated content analysis, and re-written from the original IBM language PL/1 into Java, it persists to this day. However, most of the development of automated tools within the context of the social sciences occurred in Europe (see Alexa and Zuell 1999; Popping 2000), though automated natural language processing in general continued apace in computer science (e.g., Salton 1989; Smith 1990), and occasionally, these efforts would move into the political realm (e.g., ARPA 1993).