Lab Meeting

*Note: this week we will meet in Peabody 560*

Joseph Dexter (PhD student in the Systems Biology Program)

Quantitative Criticism and the Cultural Evolution of Literature The digitization of large corpora of natural language texts has resulted in an explosion of interest in the computational analysis of literature. In this talk I will describe "quantitative criticism," an approach to literary study that integrates techniques drawn from machine learning, natural language processing, and bioinformatics with traditional modes of humanistic inquiry (Dexter et al. 2017 PNAS). I will discuss literary adaptation on scales ranging from imitation of individual sub-verbal features to the dynamics of whole traditions, with attention to three case studies: 1) the use of machine learning to characterize subtle instances of text reuse in Roman historiography and to profile the development of Latin prose style over hundreds of years; 2) the high-throughput identification of instances of verbal "intertextuality" using sequence alignment; and 3) the construction of phylogenetic profiles for enhanced visualization of literary reception histories. Each case study will highlight the ability of quantitative criticism both to address specific questions of humanistic interest and to enable large-scale profiling of the cultural evolution of literature. At the end of the talk, I will also discuss an early-stage attempt to quantify cross-cultural variation in the evolutionary dynamics of language regularization.