Faculty Contact: John Mohr, Sociology
Abstract: Text analysis in the social sciences and humanities has tended to be limited in scope in the sense that the primary goal has been to extract the basic “surface” meaning of a text and to map it across the space of the textual corpus. But traditional “close readings” of texts by humanistic scholars have tended to have far more ambitious intentions of extracting hidden or basic underlying meanings from a text. With the emergence of ever more sophisticated text mining tools becoming available, the moment has arrived for bringing a more sophisticated theory of “close reading” together with more nuanced and sophisticated tools for automatically coding, parsing and marking texts.
In this IGERT module, the goal is to focus in particular on the nature of rhetoric, its character, shape, form and effect. And the goal is to use state of the art text mining technologies to take the measure of rhetorical speech. We will borrow from literary theorists such as Kenneth Burke to develop a better understanding of the nature of rhetorical forms and from an array of text mining tools as a way to move toward automated coding and formal analysis of rhetorical styles.
Goal: To develop automated procedures for coding free form text so that its rhetorical form and style can be marked and analyzed. Note, each of these is really a goal unto itself: (1) Testing out different ways of marking the text and also, (2) testing out different ways of analyzing this marked text to observe its rhetorical character.
Our strategy will be to:
(#1) Think about what qualities of a text we want to mark. This requires thinking about what features of a text that we want to highlight. What is our “theory of the text”? What is the nature of rhetoric and how should we assess it?
(#2) Survey available text mining tools to assess what techniques have been used for these tasks before and also what are some new text mining tools that could be applied to these issues?
(#3) Test these tools as a way to markup a Test Corpus (the National Security Strategy corpus)
(#4) How can network analysis of this rhetorically marked text allow us to better understand the character and effects of these rhetorical forms?
- Fall 2015: Alex Kulick, Ayme Tomson, Tegan Brennan
- Spring 2017: Devin Cornell