Sunday, February 15, 2026

Jade Harrison and the Architecture of Annotation



By Kenton Rambsy

Digital humanities projects often rely on labor that remains unseen, and the Black Short Fiction dataset depends on that labor, with Jade Harrison serving as Data Ranger Coordinator and lead research assistant for this Mellon-supported phase of the Black Lit Network, structuring the project through annotation templates, data dictionaries, database design, and quality control systems.

I first met Jade when she was an undergraduate at the University of Texas at Arlington. She later returned for her MA in English, where I served as her thesis advisor, and she is now completing her PhD at the University of Kansas. There she worked closely with the History of Black Writing project and with Maryemma Graham before Graham’s retirement, strengthening Jade’s foundation in African American literary scholarship and collaborative archival research.

Jade’s contributions extend beyond coordination. She helped identify, digitize, and organize the corpus of short stories, and she created data dictionaries that define how each column in the dataset functions. She established standards for interpreting different story types, recognizing that annotating speculative fiction requires different decisions than annotating historical or realist fiction. After each story is annotated, she reviews files for consistency and accuracy, ensuring the dataset reflects literary nuance rather than flattening it.

Once datasets are cleaned, Jade produces analytical reports that quantify patterns across writers and stories, developing an expansive knowledge of Black short fiction and identifying differences in how writers structure dialogue, distribute space, and construct character relationships. Moving between annotation and analysis has strengthened her ability to connect themes, forms, and historical contexts across a wide range of authors, demonstrating how literary expertise and digital humanities methods reinforce one another. She moves between close reading, database construction, and quantitative reporting with precision, ensuring that the dataset is not only sizable but intellectually rigorous and historically grounded.

Related:

No comments: