Welcome to Methods of Text Analysis – DHUM72500

This course takes as its guiding questions: “Can there be such a thing as feminist text analysis?” and “What does it mean to do computational text analysis in a humanities context?” Through reading and practice we will examine the degree to which problematic racist, sexist, colonialist, corporate, and gender-normative assumptions that activate algorithmic methods impact humanistic inquiry through text analysis, and how the humanist can formulate effective research questions to explore through methods of text analysis. 

Taking a completely different approach to the topic “methods of text analysis,” this course will consider what it means to “analyze” a “text” with computers within a humanistic context, with an emphasis on shaping effective research questions over programming mastery. How does the language of analysis draw on Western traditions of empiricism in which “the text” occupies a position of authority over other forms of representation? What is the difference between “text analysis” and “philology”? What is being “analyzed” when we count, tokenize, measure, and classify texts with computers? And, importantly, how do the questions we are asking align with the methods we are using?  

The course will be organized according to the stages of the research process as articulated in our first week reading, to be completed in advance of our first class meeting: “How we do things with words: Analyzing text as social and cultural data,” which can be downloaded here:  https://arxiv.org/pdf/1907.01468.pdf. While students will receive materials to help them learn Python and to develop their own text analysis projects, this will not be the objective of the course or the source of evaluation. However, students will be required to develop a literacy in Python and packages frequently used to perform text analysis. Students will be required to complete weekly Jupyter notebook assignments that have significant portions of text analysis activities already completed. Supplementary information about programming and text analysis will be provided to complete in a self directed way using a free DataCamp account. Assignments will include weekly readings and hands-on activities, including up to 14 weekly Jupyter notebooks and DataCamp course modules; up to 2 blog posts and/or presenting as part of a public roundtable; and a final portfolio that includes a five page position paper.

Exploring terms such as “non-consumptive” and “black box algorithms,” this course takes up the affordances and costs of computationally enabled modeling, representation, querying, and interpretation of texts.  We will ask questions such as, “Can you ‘lead a feminist life’ (Ahmed) that is heavily mediated by methods of text analysis?” Readings will include articles by Sarah Ahmed, Mary Beard, Meredith Broussard, Lauren Klein, Wendy Chun, Tanya Clement, Miriam Posner, Liz Losh, Tara McPherson, Johanna Drucker, Andrew Goldstone, Safiya Noble, Bethany Nowviskie, Andrew Piper, Steve Ramsay, Laura Mandell, Susan Brown, Richard Jean So, and Ted Underwood.