Replication Data for "Covering Blue Voices: African American English and Authenticity in Blues Covers"

This repository contains data for a quantitative analysis of blues lyrics performed by artists across time and socio-cultural groups. This analysis is a part of my PhD project on the use of African American English features as indexical expressions of authenticity in blues music. The particular study for which data is shared here examines the use of African American English (AAE) features in blues music, for which a corpus of 270 studio-performed blues songs was compiled from YouTube, consisting of six songs each by 45 artists. These artists were evenly distributed across three social groups (African American; non-African American, US-based; and non-African American, non-US-based) and three time periods (the 1960s, 1980s, and 2010s). Each artist contributed three original songs and three covers (i.e., previously recorded by other performers). Songs were selected to fit broad blues criteria, including structural, melodic, and lyrical patterns, encompassing traditional blues and contemporary blues-rock. All 270 songs were imported into MAXQDA for transcription and annotation of five phonological and three lexico-grammatical AAE features, selected based on established sociolinguistic literature. Each token where a feature could potentially occur was coded in binary fashion (realized or not), with uncertain cases left uncoded. The annotated data were exported from MAXQDA into a structured tabular format for statistical and machine learning analysis in Python.

A peer-reviewed version of this data repository is hosted by TROLLing: https://doi.org/10.18710/DOJXAV. The Python code used to (pre)process and analyze the data are hosted on this GitHub repository: https://github.com/romeodetimmerman/aae-in-blues-slx_and_music.