ABSTRACT In this project, the investigators use computational techniques and multivariate statistical techniques to trace linguistic changes in English over the last four centuries. The study analyzes a computerized corpus of 1,500 texts (3,000,000 words), representing 17 speech-based and written registers. The computational techniques and large corpus allow analyzing change along multiple parameters of variation and uncovering patterns of greater scope, accuracy, and reliability than would be possible otherwise. Specific undertakings will be: 1. descriptions of the linguistic characteristics of speech-based and written registers in English for each of the last four centuries; 2. diachronic analyses of the changes within and among these registers; 3. comparison of developments across British and American English; and 4. 'explanations' of the observed developments in terms of changes in social attitudes, demographics, and purposes of communication (including the expanded range of uses for written language). Although numerous studies have analyzed differences between speech and writing in English, apart from the earlier work by the PIs no study has analyzed differences in earlier periods or traced the linguistic evolution of written styles. The research advances the multi-feature/multi-dimensional model of variation developed in their earlier studies. This model provides an empirical basis for analyzing the evolution of registers in relation to the expansion of literacy uses in English; and it lays a foundation for cross-linguistic comparisons, thus enabling investigation of universal tendencies of register development over time.
