This lesson introduces textual analysis and visualization using Voyant to explore hip hop lyrics.
Workshop (80 minute); faculty collaboration
Teaching faculty in Art & Music History, Literature, African American Studies, Writing
Undergraduate, lower-level, new to digital studies/humanities
Preparation for assignment analyzing songs, albums, or artists’ oeuvres; digital/visual supplement to traditional term paper on a single artist’s or group’s lyrics at different points in their careers. Students’ final papers must include a visualization of some type (one facet of a Voyant Tools analysis, for example) and discussion of the insights into the artists’ work it prompted or provided.
Students will be able to describe a variety of uses for text analysis and visualization to guide their research.
Students will be able to compare and contrast (small-scale) text analysis and visualization approaches and choose appropriate methods for particular research questions.
Students will recognize their needs in procuring a dataset.
Students will be able to prepare their textual data appropriately for an array of analyses.
Students will consider the meanings of the inputs and outputs of textual analysis tools.
Students will consider non-linear and multimodal applications of text analysis in completing their research assignments.
Instructor should prepare and test datasets for a single song, multiple voices, and several song corpora for use in examples and by students during the workshop.
Students will need computers to work through in-class examples and have chosen in advance a song, artist, or album of their own to explore.
Instructor should prepare a website (depending on content instructor intends to include, LibGuides, Google Doc, txt.fyi pages may all be suitable platforms) with links to text analysis tools, downloadable/easy to copy-paste example text files in a variety of forms for experimentation, and links to relevant lyrics sites & YouTube videos.
Librarian instructor introduces concepts of simple visualization and text analysis with relevant examples from digital humanities projects in the subject area of the course. Librarian instructor shows lyrics and plays captioned audio of songs in the dataset.
Librarian instructor & students work through text analysis with Voyant Tools and various small datasets (most prepared in advance by instructor). In each case, if possible, the instructor should play the music being analyzed and remind students close listening in the music and lyrics may lead them to interesting insights or ideas about their analysis.
Librarian instructor demonstrates first order use: copying lyrics from Genius or other source (without any clean-up) for exploration with Voyant Tools for the group, preferably with sufficient complexity (multiple speakers/singers, variant spellings for sung words, text indicating intro/chorus/bridge, and other non-lyrical textual artifacts) to complicate the visualizations’ relationship to literal lyrical content. Concentrating on the Cirrus and word frequency charts, ask students to consider what they want to include in their analysis (song parts, source of lyrics, material from samples).
Comprehension check: As a group, return to Voyant Tools entry point, paste lyrics again, editing out irrelevant words or sections; solicit feedback how the resulting changes might affect analysis.
Librarian instructor demonstrates using multiple prepared files to analyze segments of text separately, sharing separate text files for a song with multiple lyricists for upload and analysis. Instructor should draw attention to the Bubblelines, Summary, and Reader (timeline) features for comparison and exploration. Instructor also invites students to experiment with separate text files representing different songs for comparison.
Comprehension check: Students discuss, in pairs, the types of observations and analyses the multiple files afford and report an insight they found to the group (Think-Pair-Share).
Instructor demonstrates analysis of a larger (20+ track) corpus of an artist’s lyrics, organized chronologically, and considers how the organization of the tracks in the file enable different analyses. Instructor emphasizes Keyword In Context section with this larger dataset.
Comprehension check: Students discuss possible report interpretations of commonly used words and phrases in the dataset (Think-Pair-Share).
Throughout each exploration, instructor should encourage students to think about the different ways in which language enters hip hop (verses, choruses, hooks, adlibs, guests, shouts, toasts, freestyles, samples) and how they would incorporate or exclude such features in their analysis. Additionally, instructor should model resizing/adjusting interface frames to emphasize most relevant portions of Voyant Tools interface for smaller datasets.
Students work with their own chosen songs/artists to create initial datasets and explore with Voyant Tools. During this time instructor checks in with each student; by end of session students should have a small dataset representing their interests to guide their initial exploration.
Formative assessment takes place during comprehension checks and through hands-on work and consultation.
Session assessment may be based on the degree to which students leave the session with a workable starting dataset for analysis they have already auditioned in Voyant Tools.
Summative assessment needs to take place after students have had time to explore and apply what they learned to their term papers. A librarian instructor present for class presentations or given access to submitted papers can perform this assessment work, or more general assessment can come through follow-up with course instructor.
The times I have used this lesson, students have been very receptive —especially if they have been primed in advance to have an artist in mind whose lyrics they would like to analyze. I have also prepared multiple example datasets in advance to give more potential to the conversations among students. Additionally, on the web page I use to support this lesson, I include additional options for text analysis and visualization (databasic.io, Tableau) for student to explore once they have an idea what is possible.