Remixing “Taste”

Abstract

In this paper, I discuss dimensions of remix, including attribution and authorship, for automated digital music playlists. I use the case study of Spotify Blend, an automated, personalized, mock-collaborative playlist feature that combines up to ten users’ music taste and listening histories and regenerates its content daily. I defend the characterization of Spotify Blend as an example of “remix” (or an example of “mashup,” a related concept), wherein the source material being remixed is user listening data and wherein sampling is the primary remix tactic. In fleshing out this characterization, I discuss how the concepts of “authorship” and “attribution” operate in the context of Spotify Blend, with the important acknowledgment that Spotify’s algorithm remains opaque to users. I also compare Spotify Blend with user-generated, actively collaborative playlists created on the same platform. Observations about Blend and its features of use derive mostly from personal experience with the program.