Date created: | Last Updated:

Category: Project

Description: This study examined the topical structure of Hip Hop music and compared male and female Hip Hop lyricism using latent Dirichlet allocation (LDA), a machine learning methodology. Two corpora were constructed (one male and one female) using 103 songs that appeared on the U.S. Billboard charts between 1980 and today. LDA identified seven emergent topics in the male corpus (coherence score = 0.423) and six emergent topics in the female corpus (coherence score = 0.455) that present varying perspectives on Hip Hop culture, lifestyle, and male versus female experiences. Each set of topics is explored independently and then compared to identify similarities and differences. Findings indicated that the male topical structure utilized more active language to demonstrate a greater emphasis on agency, aspiration, and struggle. The female topical structure revealed a higher prevalence of object-oriented language that indicated a greater focus on goals, resources, and aesthetic pleasure. The keywords in the female artist corpus also contained more self-reflective terms that substantiate the idea that female artists endeavor to redefine womanhood in Hip Hop through lyrical prowess and challenging patriarchal norms within the genre. Implications for practice and recommendations for further research are also discussed.