Text Analysis

Text Analysis

Digital and distant reading.

Brief

Computers can't think, at least not yet, but they assist human thinking every day whether we our phone's calculator or answer a random question with a Wikipedia search. What about more complex forms of thought like literary analysis? As it turns out, there are some kinds of software-driven analyses that provide useful tools to the human analyst.

Given, say, a novel, what are the most frequently used adjectives? Can you recognize the novel from its wordcloud? What words are likely to appear near each other in American English versus British English? What is the text "about" on a semantic level? Does the author have certain habits that reveal themselves numerically? How does this version of a text differ from a previous edition or draft? What do those changes suggest about the writing process or the text's position in culture?

These are the sorts of questions that computers can help us answer. They can also raise other questions like, "Why have the Yankees been so popular for so long?" Or "Why does 'Mr' appear more frequently than 'Mrs' in Jane Austen's novels?".

Learning Objectives

  • Learn how computers can assist in analyzing literary texts
  • Practice using different tools for text analysis
  • Learn how to compare and distinguish between different types of software-driven analysis
  • Practice applying insights gained through software-driven analysis within a critical interpretation or analysis of a text

Options and Variations

  • Read one or more of the studies in the bibliography below and write a blog post discussing how those researchers used computational approaches to study some texts. What tools did they use? What questions were they trying to answer? Why was a computer program necessary?
  • Use Voyant Tools to generate different visualizations of a novel you know well, then use it on a novel you haven't read. Compare both results to understand what these visualizations do or do not reveal about these works.
  • Use any of the tools to develop an argument offering new insight about a literary work you know well.
  • Use the "Trend Analysis" in Voyant to understand how a phrase, idea, or character changes across multiple works by the same author.

Resources and Tools

Bibliography

Berensmeyer, Ingo, and Sonja Trurnit. “Post-War British Women Writers and Their Cultural Impact: A Quantitative Approach.” Journal of Cultural Analytics 7, no. 1 (March 31, 2022). https://doi.org/10.22148/001c.33994.
Fyfe, Paul. “How to Not Read a Victorian Novel.” Journal of Victorian Culture, 2011. https://diginole.lib.fsu.edu/islandora/object/fsu%3A207269/.
Intro to DH. “Distant Reading Duffy,” March 5, 2015. https://www.briancroxall.net/s15dh/assignments/distant-reading-duffy/.
Ladd, John R. “Understanding and Using Common Similarity Measures for Text Analysis.” Programming Historian, May 5, 2020. https://programminghistorian.org/en/lessons/common-similarity-measures.
Lavin, Matthew J. “Analyzing Documents with TF-IDF.” Programming Historian, May 13, 2019. https://programminghistorian.org/en/lessons/analyzing-documents-with-tfidf.
Romanello, Matteo, and Simon Hengchen. “Detecting Text Reuse with Passim.” Programming Historian, May 16, 2021. https://programminghistorian.org/en/lessons/detecting-text-reuse-with-passim.