Investigating African American Writing and Thought by comparison of two corpora via Distant Reading

Investigating African American Writing and Thought by comparison of two corpora via Distant Reading
Michael Achmann
Aenne Knierim
Christian Wolff


In critical race theory, researchers have introduced the notion of a unique voice of color [3]. More exactly, minorities share different histories and experiences with oppression and have a competence to speak about race and racism [3]. As Delgado and Stefanic state, “society constructs the social world through a series of tacit agreements mediated by images, pictures, tales, tweets, blog postings, social media, and other scripts […] Describing the reality of black and brown lives can help readers understand what life is like for others and may begin a process of correction on our system of beliefs and categories by calling attention to neglected evidence […] [3].” Until the beginning of the 21st century, African American writers and thinkers have expressed their voice through both works of fiction as well as non-fiction works such as “interviews, journal articles, speeches, essays, pamphlets and letters”. But digital media has changed the way African Americans and other minorities can make their voices audible. For instance, in the form of New Social Movements like #Blacklivesmatter, social media platforms like Twitter or Facebook offer space for activism. Today, many African Americans memorize historical events and share current experiences of racist discrimination under the hashtag #BlackHistoryMonth on Twitter.

Zielsetzung der Arbeit

In my thesis, the voices of African Americans are investigated by comparison of a current Twitter corpus and a historic corpus via Distant Reading, using an explorative mixed-method approach. I have three research interests: Firstly, what topics of African American history are spoken about on Twitter today? Secondly, how are these topics displayed within the historical time period? Thirdly, how does the expression of African American thought differ due to the medium? The form of language will be compared using statistical measures for text analysis. The content will be explored via topic modeling.

Konkrete Aufgaben

  • Create #blackhistorymonths-corpus.
  • Get access to the historical Corpus on African American Writing and Thought, containing 22000 documents.
  • Execute topic modeling for #BlackHistoryMonth corpus.
    1. Based on the result, creation of subcorpora from the Corpus on African American Writing and Thought. This guarantees comparability from the American studies perspective. Also, like this, I make use of the macroanalytical perspective following Jockers instead of limiting my research object a priori.
    2. Use statistical measures for text analysis to investigate the form of writing of both corpora.
    3. Execute topic modeling for the subcorpora of the historical corpus on African American Writing and Thought.
    4. Comparison, interpretation, analysis

    Erwartete Vorkenntnisse

  • Kenntnisse in Python
  • Kenntnisse in Natural Language Processing
  • Kenntnisse in Arbeit mit Social Media Sprache
  • Kenntnisse amerikanischer Geschichte und Critical Race Theory

Weiterführende Quellen

[1] Caliandro, A. & Graham, J. (2020). Studying Instagram Beyond Selfies. Social Media + Society, 6(2), 205630512092477.

[2] Da, N. Z. (2019). The Computational Case against Computational Literary Studies. Critical Inquiry, 45(3), 601–639.

[3] Delgado and Stefanic

[4] Earle, J. H. (2000). Routledge Atlas of African American History (Pap). Routledge.

[5] Franklin, V. P. (1992). Black self-determination: A cultural history of african-american resistance (2nd ed.). Lawrence Hill Books.

[6] Jockers, M. L. (2013). Macroanalysis: Digital Methods and Literary History. University of Illinois Press.

[7] Moretti, F. (2016). Distant Reading. Konstanz University Press.

[8] Mundt, M., Ross, K. & Burnett, C. M. (2018). Scaling Social Movements Through Social Media: The Case of Black Lives Matter. Social Media + Society, 4(4), 205630511880791.