The 1619 (Digital Discourse) Project
Objective
Dismantling racial inequalities and constructing educational possibilities requires standing up for the language of historical truth. In the summer of 2023, the state of Florida made headlines for its revised history curriculum including the ahistorical claim that enslavement taught ancestral Black Americans useful skills. In this paper, we investigate the discourses of teaching and representing the history of slavery in the United States, especially in relation to the discourse surrounding the 1619 Project. Using systemic functional linguistics and data modeling, we examine the digital discourse — e.g., online journalism, blog posts, and social media posts including Tik Toks and Instagram posts — composed in response to the three central texts of the 1619 Project:
- The New York Times Magazine’s “1619 Project” issue
- The 1619 Project: A New Origin Story: a book of essays that expands on the magazine’s special issue, “weaving together eighteen essays that explore the legacy of slavery in present-day America with thirty-six poems and works of fiction that illuminate key moments of oppression, struggle, and resistance” (1619books.com)
- The 1619 Project: Born on the Water: a picture book by Nikole Hannah-Jones, Renée Watson, Nikkolas Smith
Frameworks
Often, public discourse that discusses (and denies) the impact of enslavement involves strong emotion. Appraisal analyses are concerned with evaluating the kinds of feelings that are negotiated by speakers, the strength of the feelings that are being negotiated, the intertexts from which feelings are derived, and how listeners and readers are aligned through language (Martin & Rose, 2003: 26). Appraisal analyses highlight the amount of speaker or writer engagement towards a topic, the interactants’ feelings about the topic, and the participants involved. Appraisal analyses also evaluate the degree or gradation of the attitudes and engagement present in spoken or written text. Speakers and writers “use the resources of appraisal for negotiating our social relationships, by telling our listeners or readers how we feel about things and people” (Martin & Rose, 2003: 22).
Results
SFL Results
In the first phase of the project (2022), we focused on digital responses to these three texts in particular based on our interest in researching the role of the 1619 Project in K-16 education. As we delved into the discourse, we sought to trace leitmotifs of imagined slavery — recurring themes in the discourse surrounding the 1619 Project’s central texts. We aim to discover:
What words, phrases, and imagery do the authors of online book reviews, blog posts, social media posts, etc. employ in their responses to the texts of the 1619 Project?
What feelings, judgments, and values do the authors/content creators express in response to the texts?
Our team searched online for sources written in response to the 1619 Project texts and identify representative quotations from the sources that express Affect, Judgment, and Evaluation-Worth/Values. The corpus of selected sources and quotations is meant to be representative of our findings as opposed to comprehensive.
What feelings does the author of the book review, opinion piece, etc. express in response to the 1619 Project texts? What emotional responses do the texts elicit in the author or in the reader?
Examples:
John Duffy, cited in the Washington Post, describes the project as “a compendium of sorrows” that exposes the “everyday brutality and misery” of slavery.
Duffy describes his “reasons for teaching the 1619 Project” as “not entirely intellectual” but “equally visceral.”
What moral or ethical judgments does the author convey about the histories or legacies related in the 1619 Project texts?
Ex.: A Kirkus review describes the way the book The 1619 Project: A New Origin Story “explores the history of slavery in America and its countless toxic consequences.”
What judgments does the author make about the ideas or arguments expressed in the texts?
Ex.: A Kirkus review states, “Hannah-Jones sounds controversial notes at the start: There are no slaves but instead enslaved people, a term that “accurately conveys the condition without stripping the individual of his or her humanity,” while the romantic plantation gives way to the more accurate terms labor camp and forced labor camp.”
What does the author report about others’ judgments of the 1619 Project texts?
Ex.: A Kirkus review states, “As she notes, the accompanying backlash has been vigorous, including attempted laws by the likes of Sen. Tom Cotton to strip federal funds from schools that teach the 1619 Project or critical race theory.”
How does the author evaluate the value, contribution, significance, originality, etc. of the 1619 Project texts?
Ex.: A Kirkus review states, “Those readers open to fresh and startling interpretations of history will find this book a comprehensive education. A much-needed book that stakes a solid place in a battlefield of ideas over America’s past and present.”
Corpus Analysis Results
To conduct large-scale analysis of the public discourse surrounding the 1619 project, we needed to construct a corpus of the public discourse surrounding the project. To do so, we scraped the top 10 articles that mentioned “1619 Project” every week between the article’s first publication and April 4th, 2023.
Each scraped article contains associated metadata, title, author, description provided by publisher, link, hero image, authors, associated movies, and publisher provided keywords.
![](https://jasongodfrey.info/wp-content/uploads/2024/04/Drawing-2024-04-09-16.39.21.excalidraw.png)
Left-leaning publications publish about the 1619 project more
The chart below shows text length by publication bias, but, more importantly, it shows how much more left-leaning publications discuss the 1619 project. This is interesting from a public discourse perspective for several reasons.
The imbalanced dataset also has consequences for continued work. All subsequent analyses will be based on normed percentages.
![](https://jasongodfrey.info/wp-content/uploads/2024/04/textlength-768x497.png)
Left-leaning publications are slightly more positive
The chart below shows article polarity by publication bias after imbalances in the underlying data have been accounted for. A 1 on the x-axis would be a completely positive article; a -1 completely negative. While, on average, all articles are relatively neutral to slightly positive, right-leaning publications more frequently give the project slightly negative coverage.
This analysis included all articles, even though the majority of publications are relatively centrist (although they were bifurcated left or right). Subsequent analyses without centrist articles may be more revealing.
![](https://jasongodfrey.info/wp-content/uploads/2024/04/polarity-768x543.png)
Right-leaning articles are generally shorter
The chart below shows article length by publication bias after imbalances in the underlying data have been accounted for. In general, right-leaning articles are generally shorter. This could be indicative of a disparity for right-leaning publishers to engage deeply with the material on average. Publications on both sides of the aisle both exhibit a long tail; demonstrating that long form articles that do deeply engage are present on both sides.
![](https://jasongodfrey.info/wp-content/uploads/2024/04/textdensity-768x474.png)
Right-leaning articles focus on the NYT. Left, on NHJ.
The chart below shows the frequency of named-entity types by publication bias after imbalances in the underlying data have been accounted for. While this chart reveals many interesting differences based on publication bias, the most prominent difference is that right-leaning articles most frequently mention “ORG” or organizations, which in practice means referencing the project as a NYT publication. Meanwhile, left-leaning publications are more likely to reference the author of the work than the organization that first published it.
![](https://jasongodfrey.info/wp-content/uploads/2024/04/NER-768x328.png)
Left-leaning articles most commonly focus on contemporary issues. Right, on historical topics.
The analysis below is the result of Latent Dirichlet Allocation on the vectorized corpus. It shows the most common topics in the underlying corpus. From the most common set of topics, it is clear that left-leaning publications are likely to frame their discourse on the 1619 project in terms of contemporary impact, referencing president Trump. Meanwhile, right-leaning publications are likely to focus on the historical aspects of the project such as the mayflower, the pilgrims, and the year 1619.