
This dataset contains a multi-layered prosodic, phonetic, musicological, and rhetorical analysis of Maya Angelou’s performance of her poem “Caged Bird.” The data were generated using the Rhetoric Analysis Toolkit, which integrates forced phonetic alignment (Montreal Forced Aligner), acoustic feature extraction (Praat/Parselmouth), musicological modeling (MIDI and MusicXML), and AI-assisted rhetorical-scheme detection.
It includes word- and phoneme-level acoustic measurements (pitch, intensity, spectral features), time-aligned TextGrid files, full JSON-based representations of prosody and phonetics, AI-generated annotations of rhetorical patterns (alliteration, assonance, anaphora, isocolon variants, rhyme patterns, and ploce), and musicological encodings of melodic contours derived from speech. These components allow researchers to examine how vocal performance—through pitch movement, timing, intensity, articulation, and rhythmic patterning—interacts with rhetorical structure to produce aesthetic, affective, and political meaning.
The dataset is designed for work in digital humanities, rhetorical studies, African American literary studies, linguistics, music-speech analysis, and computational poetics. It supports fine-grained analysis of activation contours, stylistic patterning, and the materiality of voice in African American performance traditions.